The purpose of this thesis is to use an agent-based
simulation to make a quantitative assessment of the effectiveness of a team of
firefighters when responding to a fire-on-board situation.� The agent-based part
of the simulation utilizes the METutor package.� A stochastic model simulates
the fire spread; it simulates the effects of fire in different compartments
with different inflammable contents and anti-fire features when a team of
firefighters attacks it.� The simulation enables performance comparison of
teams of firefighters with different skill levels.
An important issue in a damage-control situation
involving fire is how fast that fire may spread inside a compartment with or
without anti-fire devices, based on the time and skill of reaction of a team.�
Part of that skill is situational awareness.� How quickly might a team with
different skill levels put out or control a fire in a compartment?� What
measures of performance and effectiveness best summarize their performance (for
example, the expected or median time until a fire is out, expected or median
damage to compartment contents, risk/probability of total inflammation/burn up,
and probability of injury or death of a team member)?
All these measures depend on factors such as time and
correctness of response.� What should be the composition of a team to
accomplish all the tasks needed to put out a fire?� How effective is a sequence
of actions that comprise a standard procedure like setting boundaries for
possible fire spread, use of anti-fire devices such as hose, foam, halon, and
CO2, verification that a fire is out, testing of O2
level, removal of smoke from the area, removal of water from the area?� What
happens if incorrect or improper orders are given by unskilled team members, or
orders are not followed correctly, or are late?� How do combinations of such
mistakes affect the extinguishing of the fire?
The scope and purpose of the thesis is to develop
and test a multi-agent simulation program module of a firefighting team which
integrates unique real-world elements and subjective characteristics of human
beings such as knowledge, experience, communication and coordination.
Since a physical experiment is very difficult to
reproduce and test without any risk to human lives and expensive resources, an
agent-based simulation is used to study the effectiveness of a firefighting
team.� The simulation models the fire and the effects of sequences of actions
required to put out a fire.� These actions vary with the size of compartment,
how the fire spreads, the type of the fire, communications between the team
members, and the skill levels of the team members.
The agents in this multi-agent simulation are models
of members of the team, the command center, and the fire itself.� Agents are
defined as software modules with more autonomous capabilities than most
software components, typically including reasoning, communication and
decision-making. �In this simulation, agents react to external events produced
by other agents including the fire, but, unlike adaptive agents, do not change
their rules in response to new environments.�
Actions
(such as approaching the fire, extinguishing the fire, testing the gases, and
removing the smoke) require specific preconditions be achieved so that they can
be carried out.� The skills of the executor influence the duration of each
action and the overall time to complete the mission.� Agents use orders and
reports to communicate and coordinate actions.
The Prolog programming language and METutor package
are used to provide computerized artificial intelligence input to the model of
the situation.� The artificial-intelligence means-ends algorithm is used to
build action sequences from a given start state to a goal condition by a
sequence of state transitions.� The algorithm permits the modeler to specify
the possibility of random events such as casualties, availability of a doctor,
faulty (possibly broken) lines of communication, and equipment malfunctions
that can influence the outcome of the simulation.
Analysis of firefighting approaches requires
assessment of teams with members of different skill levels and different
numbers of members who must interact and follow specified policies.� Simulation
output analysis leads to conclusions about the successful composition and basic
tactics of a team.� Measures of effectiveness for firefighting include the mean
and median fire duration as a function of firefighter skill level, and the mean
and median material burned as a function of firefighter skill level (the median
is less susceptible than the mean to long durations and extensive
destruction).� Histograms, as in Figures 4-1 to 4-6, indicate that fire
durations can have exceedingly variable, and occasional long, durations. The
median does not reflect these.
THIS PAGE IS INTENTIONALLY LEFT BLANK
The subject of this
thesis is a simulation model of a shipboard emergency response team fighting a
fire.� In this chapter we describe the initiation and progression of a fire,
the methods used to fight a fire, and the organization of a team responding to
a fire.� The first section of this chapter summarizes how a fire spreads, how it
is classified, and what kind of devices and personnel are necessary in
different situations.� The second section describes the basic composition and
organization of a team of firefighters, as well as a sequence of firefighting
actions and the personnel needed to carry them out.� The third section
describes the importance of a well-skilled team, and how this can be achieved.�
The fourth section describes artificial-intelligence methods that can be used
to make a simulation mimic human behavior.� The fifth section describes the
agent-based approach that we use for simulation.
Once started the fire may go out if its combustibles
are isolated [Ref. 1:pp.1-9].� But usually the fire spreads to other material,
depending on their mutual configuration.� Ignition rates (rates with which
combustibles catch fire) differ for solid fuels and, for example, are greater
for dust or shavings than for bulky materials (so small wood chips will burn
faster than a solid wood beam because more vapors are available for ignition).�
�Ignition rates are measured in kilowatts (kW); one kW is approximately 0.95
BTU/min�.� While the fire is burning, the material that has been on fire
gradually becomes no longer inflammable; the rate of this is called the
burn-out rate (a function of the fire size and the kind of inflammable
material).� An anti-fire action can affect both the ignition rate and the
burnout rate.� As a fire is extinguished, the temperature and the ignition rate
decrease.� During extinguishment, the burn-out rate usually increases due to
materials that become unable to burn.
�There are four classifications of fires: class A,
class B, class C and class D. Class A fires involve wood and wood products,
cloth, textiles and fibrous materials, and paper.� Class A fires should be
extinguished with water in a straight (stream) or fog pattern (spray).� If the
fire is deep-seated, aqueous film foam (AFFF) is more effective than seawater
and can be used as a wetting agent to rapidly penetrate and extinguish the
fire.� Class B fires involve flammable liquids such as gasoline, diesel fuels,
jet fuels, hydraulic fluid, and lubricating oil.� These fires are normally
extinguished with AFFF, Halon 1211, Halon 1301, or potassium bicarbonate
(PKP).� Class C fires are energized electrical fires that are attacked at
prescribed distances using nonconductive agents such as CO2 , Halon
1211, or water spray.� The most effective tactic for a class C fire is to
deenergize the compartment and handle the fire as a class A fire.� Class D
fires involve inflammable metals such as magnesium and titanium.� Application
of water in quantity, using fog patterns, is the recommended tactic.� When
water is applied to burning class D materials, there may be small explosions;
the firefighter should apply water from a safe distance, or from behind,
shelter.� Metal fires on board ship are commonly associated with aircraft wheel
structures� [Ref. 1:pp.1-13].
�There are many materials that may be used as
firefighting agents.� Examples are water, aqueous film-forming foam (AFFF),
carbon dioxide (CO2), and halon.� Water is a cooling agent and, on
board a ship, the sea provides an inexhaustible supply.� If the surface
temperature of a fire can be lowered below the fuel's ignition temperature, the
fire will be extinguished.� Water is most efficient when it absorbs enough heat
to change to steam; the steam carries away the heat, which lowers the surface
temperature.� Water in the form of a straight stream is used to reach into
smoke-filled spaces or areas at a distance from the firefighter.� It should be
directed into the seat of fire, and for maximum cooling the water must come in
direct contact with the burning material.� Water fog or spray is very effective
for firefighting.� However, it must be applied directly to fire if its benefits
are to be realized.� Additionally, a team that uses a hose with a fog nozzle
can protect firefighters from both convective and radiant heat during the
extinguishing action.
AFFF is composed of synthetically produced materials
similar to liquid detergents.� These film-forming agents form water-solution
films on the surface of flammable liquids.� AFFF diluted with water provides
three fire extinguishing advantages.� First, an aqueous film is formed on the
surface of the fuel, which prevents the escape of the fuel vapors.� Second, the
layer of foam effectively excludes oxygen from the fuel surface.� Third, the
water content of foam provides a cooling effect.
Extinguishing fires by smothering is possible with
the inert gas CO2.� It is 1.5 times heavier than air, so it tends to
settle and blanket the fire.� It is a dry, non-corrosive gas which is inert
when in contact with most substances, and it will not leave a residue and
damage machinery or electrical equipment.� In both the gaseous state and the
finely divided solid (snow) state, it is a nonconductor of electricity
regardless of voltage, and can be safely used in fighting fires that would
present the hazard of electric shock.
Halon is a halogenated hydrocarbon with
nonflammability and flame-extinguishing properties.� Both Halon 1211 and 1301
chemically inhibit the flame front.� Halon 1301 is used in fixed flooding
systems for extinguishing flammable liquid fires.� The short discharge time of
Halon 1301 (10 seconds maximum) helps keep the thermal decomposition products
well below lethal concentrations.� Personnel should not remain in a space where
Halon 1301 has been released to extinguish a fire without a breathing
apparatus.� When released in engine enclosures, approximately 15 minutes is required
before reentry [Ref. 1:pp. 1-22, 23, 24].
Most ship compartments do
not have installed automatic fire-extinguishing systems such as active
sprinklers or CO2 flooding.� When such compartments are on fire,
mobile anti-fire apparatus operated by a team of firefighters is needed.�
However, special rooms such as an ammunition magazine must have active
sprinklers to prevent chain reactions, while others such as the Main Machinery
Room and the Pump Room must have halon, CO2, and AFFF installed
inside or nearby�.
The most common shipboard emergency is fire.� US
Navy Ships organize their primary emergency-response team as a fire party made
up of people with the different skills needed to combat emergencies.� A ship must
determine the availability of personnel and materials and design an
organization for the employment of personnel and equipment to combat fires and
respond to other emergencies.� Specific responsibilities, duties, and
employment of equipment must be assigned to individuals, divisions, or
departments.� This information is put into a comprehensive form called a Fire
Bill and is made available to all personnel addressed in the document.� The
purpose of the Fire Bill is to establish a firefighting organization and
specify responsibilities for individuals and departments to ensure that fires
and other related emergencies are effectively and quickly handled.
Fires or emergencies that occur during combat or
while the ship is at General Quarters should be handled as battle
casualties by the Repair Party organization in that section of the ship which
reports to their General Quarters stations on Fire Call.� While the ship is in
port, the ship's Fire Bill may designate the In-port Fire Team as the primary
firefighting team.� The In-port Fire Team is composed primarily of personnel in
the regular damage control repair parties, so each duty section must have an
effective firefighting force.� The number of people in the fire party assigned
to a fire will vary depending on the nature of fire and the number of people
available.� There can be a large turnover of personnel assigned to repair
parties, so repair party personnel may not be familiar with the location of
damage-control equipment.
The member of a team in charge of firefighting at
the scene of a fire is the scene leader.� He directs the attack against
the fire and communicates with the repair party leader via messenger.� He must
be capable of making correct decisions during a changing situation based on his
assessment of the current state.� The scene leader is located a short distance
from the fire, and his assessment must be derived from the reports he
receives.� Scene reports can contradict previous reports due to unexpected
events such as human errors, personnel casualties, fires reflashing after being
previously extinguished, and equipment malfunction.� Nevertheless, a scene
leader must continually reassess the needed response as a result of the
reports.� He must always investigate unsatisfactory or incomplete reports from
the scene; for instance, if an unskilled member fails to report the end of a
task or does not know what to do, the scene leader may order that action again
to ensure its satisfactory completion.� The scene leader relays information to
the damage control assistant (at a place that we refer as the Command Center
in our simulation) to keep it updated and to receive further instructions.
The fire attack team consists of an optional attack
scene leader, and one or two hoses manned by nozzlemen and hosemen.�
The attack scene leader directs the nozzleman in employment of the hose,
directs the hosemen, directs that the hose be charged or secured, selects the
water pattern to be used, and directs the rotation of attack team personnel.�
Generally, the nozzleman works the hardest and will need to be relieved first;
rotating the nozzlemen and the hosemen can extend the endurance of the attack
team.� The attack team can be supported by other emergency response team
members such as an electrician (responsible for deenergizing, reenergizing, and
desmoking the area), investigators (responsible for testing for gases and
oxygen after fire is out and for monitoring the oxygen-mask control time of
nozzlemen and hosemen for rotation purposes), hospital corpman, and dewatering
and desmoking teams when personnel are available.� Communications are often
disrupted, so voice amplifiers should be provided to the members of a fire
party.
There is a preferred sequence to completing major
actions while combating a fire.� For example, if only 90% of a fire is
extinguished and an order to desmoke is given, the high rate of airflow from
the desmoking fans will cause the fire to increase in intensity.� Fire team
members must understand the general order of actions or strategy of combating a
fire and how to effectively concentrate available resources to carry out these
actions.� In addition to strategic knowledge, the scene leader must understand
the details of every other member's job and equipment to differentiate between
satisfactory and unsatisfactory reports.� For example, he must know that 16%
oxygen is inadequate to sustain human breathing, whereas 21% is satisfactory
indefinitely.� The factual knowledge of a scene leader must cover a wide range
of functions and equipment.
A sequence of actions and
reports takes place during firefighting.� Reports announce the class of the
fire, actions taken to isolate and combat the fire, that the fire is contained,
that the fire is out, that the reflash watch is set, that the compartment is ventilated,
and that the compartment is tested for oxygen, flammable gases, or toxic
gases.� Some actions must be coordinated with reports from the fire scene.�
Generally, the first step is to go to the repair locker and assemble
equipment.� Once at the scene of the fire, the leader appraises the state of
fire and determines its precise location using men with breathing apparatus and
thermal imagers for large fires involving much smoke.� The leader must select
the appropriate tactics to attack the fire and choose appropriate tools.� Fire
boundaries must be established, and the area must be deenergized to avoid
potential injuries.� Now team members with firefighting equipment can approach
the fire.� If during extinguishment the team leader perceives that the fire is
not decreasing, he may order a change of tools.� If the fire situation becomes
worse and the fire becomes uncontrollable he may decide to isolate the
compartment.� However, if the fire decreases and eventually goes out, desmoking
and dewatering takes place, and the area is reenergized and ventilated.� Then
the leader orders a watch for a reflash (means fire can start again); if
nothing else happens, the team can return to the repair locker, store the
equipment, and receive a debrief of the action.
The best organization and equipment are useless
without trained personnel.� Properly drilled crewmen will minimize confusion
during fires, increase the probability that proper actions are taken, and
enhance the predictability of responses and tactics.� People assigned to a
firefighting party should retain that position even if other shipboard duties
change.� In the U.S. Navy, they are required to meet the damage control
training requirements specified in the General Damage Control Qualification
Standard within 6 months of reporting aboard.� The time to train an individual
for both General Damage Control and for a specific job on a fire team can take
up to 8 months depending on the individual's motivation and learning ability. �In
a shipboard environment, duty assignments rotate frequently and maintaining a
large group trained and fully qualified is a problem.� So, all members of a
fire party should be cross-trained for at least one other position in the fire
party to compensate for frequent rotation [Ref. 2:p.15], and everyone on the
ship, particularly on small ships, should be trained to serve on a fire party
[Ref. 1:pp.9-1].
U.S. Navy Ships today typically conduct a daily
drill or formal instruction for inport fire teams.� Effective fire teams gain
experience through numerous hours of realistic drill supplied by a
knowledgeable drill team.� The drill team can supply realistic details (e.g.
smoke, flames, personnel injuries), challenge fire-team members with difficult
jobs, provide a realistic shipwide casualty environment, and conduct an
effective debriefing of each person's actions.� However, a knowledgeable drill
team to execute this kind of training aboard is not always available.
Intelligent Computer-Aided Instruction programs
could be a cost-effective substitute for some firefighting instruction although
missing the physical aspects of the task.� �Procedural skills in events like
firefighting require a sequence of actions to achieve some desired result, and
"learning by doing" by using computer simulations is often a good way
for students to learn and practice these skills.� Technical organizations such
as the US Navy, demand that students learn a wide range of reactive and
technical procedures.� Learning these skills becomes especially challenging
when it is necessary to choose between actions with context-dependent effects
[Ref. 3]�.
Intelligent
Computer-Aided Instruction programs like FIRE [Ref. 2], and authoring systems
like MEBuilder and METutor [Ref. 3] are tools for constructing intelligent
simulation-based tutors for procedural skills.� The tools use planning methods,
consistency enforcement, objects, and structured menus to make writing tutors
easier than with conventional authoring software.� An instructor can create
lessons and simulations for hands-on training without many resources.� Students
can learn at their own pace, with minimal direct supervision from a teacher.�
Instructional computer programs can describe realistic problems; they can
challenge knowledge of job responsibilities and equipment operation; they can
describe a realistic shipboard environment; keep records of student
performances; and tutor when an incorrect action is chosen. �Therefore,
simulation-based tutors for procedural skills could improve the quality of
training for damage control and emergency procedures, and could sharpen
"almost-right" skills of people who need "refresher" or
"checkup" validation periodically, a frequent circumstance in the
military [Ref. 4]�.�����
In this section we describe the computer modeling of
sequences of actions used in firefighting.� Firefighting knowledge can be
envisioned as a tree of tasks where nodes represent a sequence of linked
actions and subactions which team members must execute to recover from a fire
(see Appendix C).� The tree results from a top-down stepwise refinement
design.� An expandable-action node corresponds to an action which can be broken
down into subactions.� There is a preferred order for most major firefighting actions
if we are to model the way an expert scene leader would handle a fire episode.�
If a subaction is not completed, a negative report is received, and that
subaction must be redone.
A data structure needs to represent possible states
during a fire episode.� Two states are special: a given start state (set of
initial facts) including a fire, its size, type and location (although this is
initially unknown to the fire team), and a given goal (state in which
objectives have been achieved), such as "all fires are out and the ship is
restored to normal operation".� Searching through the possible states
results in a path of states from start state to goal which can be followed by
the simulation.� The procedure used to find the path is means-ends analysis
[Ref. 5:pp. 263-270], a search method using top-down recursive decomposition of
a search problem into simpler subproblems.� This structure uses a
"difference table" showing the recommended major action for any
search problem.� It is specified by assertions of the form "recommended
(<difference list>, <action>)" which gives conditions for
recommending an action based on the facts different between the current state
and the goal.� The operator will not necessarily be able to apply the recommended
action immediately, but the action should be the most important one necessary
to solve its search problem at any moment.
Each action has
preconditions (facts that must be present in a state before we can do the
action); these are specified by facts of the form "precondition (<action>,
<precondition-facts>)".� Once an action is done, the state is
changed: some facts become true and some facts become false.� For those
conditions that become true, an "addpostcondition (<action>,
<added-facts>)" fact is used, and for the facts that become false, a
"deletepostcondition (<action>, <deleted-facts>)" fact is
used.� Addpostcondition and deletepostcondition facts convey both intended and
side effects of an action, such as "fire is out" and "area is watery"
(addpostconditions) and "it is no longer true that boundaries are
set" (deletepostconditions) for the "extinguish" action.� They
also convey state-dependent effects, such as "there is damage to the
floor" when the user forgets to turn the power off before extinguishing
the fire.
Firefighting simulations fit well the paradigm of
intelligent multi-agent simulations.� An intelligent agent is software with autonomy,
adaptation and cooperation [Ref. 6:pp. 1-5].� Multi-agent systems are
computational systems in which several agents interact to achieve their goals.�
In the human world, complex jobs are usually performed by several individuals
because individual capabilities are limited to some extent.� Analogously,
agents in the multi-agent system communicate, coordinate, and negotiate with
one another to achieve goals.� There are three aspects of agents:
�
reactivity:� to what degree agents� have an internal
representation of the world.
�
deliberation:� to what degree agents have a symbolic
representation of the world using beliefs, goals, and intentions, and possess
logical inference mechanisms to make decisions based on their representation.
�
sociability:� to what degree agents coordinate their activities
with those of the other agents through communication.
Agents used in the firefighting simulation have
these properties except the ability to adapt themselves through learning.
Agents' models for real
applications often address complex, dynamic, and nondeterministic, uncertain,
environments.� Thus they need to represent worlds with exogenous events, other
agents, uncertain effects, and social interactions.� The interaction and
synchronization between agents can be accomplished by protocols for
communication.� Protocols govern the sending of messages between social agents
and monitoring of execution.� An agent has knowledge of its internal control
mechanisms, or in other words, it knows how to make behaviors, problem-solving
methods, and strategies work together in order to achieve its goals.�
Means-ends analysis accomplishes this for the firefighting simulation.� Agents
can also have purely reactive behaviors, a necessary ingredient for the design
of real-time multi-agent systems.� A fire agent is an example of reactive agent
since its behavior is based on a mathematical model: it responds to other
agents' actions (such as extinguishing) or events (a possible flashover).
The firefighting models proposed here are
agent-based with some random behavior.� There are six agents in our simulation:
command center, team scene leader, nozzleman, hoseman, electrician and fire.�
The agents nozzleman and hoseman could represent several people.
The fire agent received special attention in our
simulation.� The fire will only be noticed if it grows above a certain limit
inside a compartment.� The kind of inflammable will determine the type of fire
and thus the necessary tools and tactics.� Then the fire will spread inside the
compartment until the team of firefighters starts to extinguish it. �We used a
simple stochastic epidemic-like model for the fire agent [Ref. 7.]� The
model assumes that the growth of a fire is a function of the amount of unburned
inflammables and the rate with which inflammables catch fire (the ignition
rate).� The model includes the feature that items on fire can burn out (the
combustibility of each elementary item), and once a unit is burned out, it
cannot burn again.� Both ignition and burn-out rates have random fluctuations
during a fire and from fire to fire.� The number of unburned combustibles at
any time is a function of the pattern of fire growth.� When extinguishing is
being carried out, the ignition rate generally decreases and the burn-out rate
increases, and both depend on the skill of the team member in charge of
extinguishing.� If the extinguish action is successful, the fire will
eventually go out (so fire size will equal 0).
The fire model allows partial flashovers (when a
fire grows unexpectedly but ignites only part of the intact (unburned)
inflammables), and total flashovers as well.� Both may happen with a certain
probability.� A total flashover ignites all intact combustibles, increasing the
size of fire and making the number of unburned combustibles go to zero very
quickly.� The probability of occurrence of a flashover is a function of the
size of fire, the temperature in the region, and the number of inflammables on
fire.� The model for firespread and flashover is speculative and illustrative,
and currently not calibrated to actual fire data or to high-resolution models
based on physical principles.� It is planned to carry out such steps in future.
Each member of the firefighting team has a skill level.�
This parameter influences the duration of the actions �equip�, �extinguish�,
�desmoke�, �dewater� and �watch for reflashing�, because these are the most
important firefighting actions.� Actions have durations which are assumed to be
independent random variables whose mean and range are a function of the action,
the agent, and the current state.� The duration of the extinguishing action,
however, depends on the behavior of the fire, as will be explained.
The stochastic model for fire spread is as follows.�
Suppose the shipboard compartment contains a quantity C of inflammable items
(in the implementation C(0), the quantity at time 0, is 100).� Let us assume
that they are all equally inflammable for the present; also, let us ignore
spatial considerations for the present.� Let X(t) denote the number of
inflammable units on fire at time t; this number is a continuous
variable with 0� X(t) � C(0).� The
fire starts with X(0)=X, and then spreads until items are on fire
simultaneously. Items on fire eventually burn out (the combustibility of the
item becomes exhausted).
The fire spreading described resembles a SIR
(Susceptible, Infectious, and Removed) epidemic model used in
epidemiology and public health; see e.g. Gani and Daley (1999).� Such a model
can be relatively easily studied deterministically using differential or difference
equations, and also stochastically, possibly using Markov chains if the
"population" size, C, is not too big.� It is proposed� here that fire
progression be modeled in this way.� Epidemic models for such processes do not
normally consider a flashover phenomenon, so we will postpone its
consideration.
Let X(t) be the number of inflammable/combustible
elements on fire at t, and let� C(t) be the number of
inflammable/combustible units unburned at t. In epidemic theory X(t)
represents infectives and C(t) represents susceptibles.�
An inflammable element burns for awhile and then becomes burned out, but
while is burning it may ignite more inflammable elements.� According to this
model:
������������������������������ X(t + h) = X(t) + lhX(t)C(t)/C(0) -
mhX(t) ����������������(3.1)
�������������������������������������������������������
New ignitions in (t,t+h)� Burnout in (t,t+h)�
���������������������������������������������������������������������
for ignition rate l����������
for� burn-out rate m
Since C(t) inflammables are immediately reduced to
C(t+h) with fire spread to X(t),� we have
������������������������������ C(t + h) =� C(t)� - lhX
(t)C(t)/C(0)����������������������������� (3.2)
�������������������������������������������������������������
����������
�����������������������������������������������������������������
In the limit for small h we obtain the differential
equations:
����������������������������������������������������
dX(t) = lX(t)C(t)/C(0) - mX(t)������������������� (3.3)�
�� ����������������������������������������������������������dt
����������������������������������������������������
dC(t) = -lX(t)C(t)/C(0)������������������������������� (3.4)
������������������������������������������������������������
dt
Note: C(0) � C(t) � X(t) is the number of inflammable
units burned out at t, where C(0) is the initial number of inflammable
units; this burn-out increases as the fire burns. If l is small compared to m,
then fires can go out on their own accord.
Fire spread can be more
realistically modeled as a random process.� Suppose that the competition
between fire growth and decay can be represented as a deterministic term
("drift") plus Gaussian /normal increments.� Then:
���� X(t+h)
= X(t) + lhX (t)(C(t)/C(0)) - mhX(t)
������� + σl [(X(t), C(t)] Zl(t) � σm [(X(t),C(t)] Zm(t)�����������������������������������������
(3.5)
where here:
lhX(t)C(t)/C(0) represents the mean incremental fire
growth������������������������� (3.6)
mhX(t) represents the mean
incremental burn-out growth�������������������������������� (3.7)
σl [X(t),C(t)] ��lhX(t)C(t)/C(0) represents the standard deviation of the
incremental random fire
growth�������������������������������������������������������������� ����������������������(3.8)
σm� [X(t),C(t)] ��mhX(t)�������� represents the standard deviation of the
incremental random
burn-out������������������������������������������������������������������������������������������������������������
(3.9)
and Zl(t) and Zm(t) are normally distributed with mean 0 and
variance 1, and {Zl(t) } and { Zm(t) }
are mutually independent sequences of independent and identically distributed
Gaussian random variables.
The decline of the intact unenflamed items can be
similarly modified to:
����� C(t+h)
= C(t) � lhX(t)C(t)/C(0)
� σl[X(t),C(t)]Zl(t)������������������������������ (3.10)
Note if a random fluctuation in X(t) is positive, represented by a positive value of Zl(t), this must mean a negative fluctuation in the
unburned material, C(t). Note that the
last term has a negative sign although Z(t)
can be of either sign.
It is necessary to make sure that boundary overshoots do
not occur when using the difference equations; if overshoot occurs, we replace X(t)
or C(t) by the boundary value 0 or C.
Equations
3.5 and 3.10 can be modified to allow for total flashover.� We use Equation 3.5
and Equation 3.10 with probability 1 � j
[(X(t),C(t)], when there is no flashover, and X(t+h)=X(t)+C(t) and C(t+h) = 0 with probability j
[X(t),C(t)], when there is a total flashover .� The function j increases monotonically with X so that there is an
increasing probability of flashover. Examples
of j are:
j(t) = [1�
exp(-gh(X(t)(C�C(t))/C))],where
the flashover constant parameter g is
1.� This is an illustrative function, not one that has been physically
validated; and
�j =
FireSize/(100C)���������������������������������������������������������������������������������������
(3.11)
The fire model when undergoing
extinguishing is similar to the normal spread model except for the values of
ignition rate and the burn-out rate.� The new ignition rate is a function of
the old ignition rate and the skills S of the team member in charge of
extinguishing such that the greater the skills, the lower the ignition rate.�
We use:�
lNew =
(lOld)
exp (-
10S)��������������������������������������������������������������������������
(3.12)
The new burn-out rate is similarly affected by the
skill level :
m New =
(m Old)
exp
(S)�������������������������������������������������������������������������������
(3.13)
To model the human behavior in firefighting we use
planning methods from artificial intelligence to determine the sequence of
actions performed.� A list of facts describes each planning state.� The initial
conditions for our model comprise the locations where fire may start (such as
the engine room or ammunition magazine), the ignition and burn-out rates for
the material inside the compartment, the quantity of inflammables (which was
100 for our test runs), the type of fire (e.g. type a, b, c, or d), and the
initial size of fire and smoke (e.g. one and zero).� Fire will not be noticed
until its size is big enough (more than 1.5 in our test runs) to activate an
alarm inside the compartment.�
Each animate agent (fire scene leader, nozzlemen,
hosemen, electrician, or command center), has a specific goal to be achieved,
so they have goal-directed behavior.� These agents use means-ends
analysis to figure what to do when they are unoccupied.� Each agent then plans
a way to achieve its goals, and chooses one action from the plan to execute
first.� An agent can be active (doing a task), idle (if its goals
are achieved), or waiting (if its goals are not achieved but it has
nothing it can do).� At each step in the simulation, the active agent that has
been least recently updated gets a chance to plan. When an agent is idle it can
be awakened by specified triggers such as orders given it or reports received
by it.
Each action has a specified set of possible agent
"actors"; the agent waiting the longest is assigned the action if
more than one agent is available.� Some complicated actions also require the
presence of assistants to the actor, such as the hoseman to hold the hose while
the nozzleman is extinguishing the fire; such actions require that the
assistant be waiting as well as the actor for the action to proceed.
There are two different ways to choose an action.�
With "first doable operator" the agent executes the first action in
the plan (generated as explained below) that is possible, whereas with
"random doable operator" the agent chooses a random possible action
in the plan.� A possible action must be permitted by preconditions, recommended
at the given time, and must not be in progress already.� The implementation
uses the �random doable operator�.��
The simulation of actions also adjusts the states
with results of terminating concurrent actions and ensures no two actions end
at the same time to prevent confusion in state reasoning.� The simulation
permits actions to be aborted by other actions in high-priority circumstances.�
For instance, a fire may reignite at random when a crewman is ventilating a
compartment; the crewman then ceases ventilation and initiates immediate fire
extinguishment steps.� Actions are also aborted when their preconditions become
false during their execution.
Means-ends analysis (see chapter 2) is responsible
for the planning for each animate agent.� It decomposes a complicated task into
simpler subtasks until all subtasks can be accomplished by single actions.�
Appendix C describes the tasks an agent must carry out, and Appendix D shows
the sequence of tasks for an example simulation run.
A list of actions needed
to achieve particular facts, not necessarily actions that can be performed
immediately, is used as recommendation conditions.� For example, the
"ventilate" action is recommended whenever the gases or oxygen are to
be made safe, and they are known to be unsafe.� On the other hand, the
preconditions of "ventilate" (facts that must be true beforehand) are
that the fire team must be at the location of the fire, the team must be
equipped, and the fire is out.
The state of the system will be affected by the
"ventilate" action as some facts become false and will be deleted
from the previous state ("deletepostconditions") and other facts will
be true and will be added to the previous state
("addpostconditions").� The postconditions of "ventilate"
are that gases and oxygen are safe, gases and oxygen are no longer tested, and
smoke is removed.� Random postconditions model uncertainties and accidents in
the simulation.� For instance, when an agent has to test gases, there is a
probability (for example, of 0.3) that gases are unsafe.� Random changes are
applied to the state after regular postconditions are applied.
Each action has a duration which is a random
variable whose mean and range can depend on other parameters, such as the skill
of the assigned agent, the fire size, the kind of tool used, smoke intensity,
and water magnitude.� To each team member we assign a number between 0 and 1
that represents a subjective measure of their skill level, where the higher the
number the better the skill.� Uniform distributions are assumed for durations.�
For instance, order and report actions have a mean of 0.25 minutes and a range
of 0.1(0.15,0.35), and the action of watching for a reflash has a mean of 15
minutes and range of 5 minutes (10,20).� The time to "dewater" has a
mean of the product of 0.15 and (WaterQuantity/Skill) and a range of the
product of 0.12 and (WaterQuantity/Skill), and "desmoke" has a mean
of the product of 0.8 and (Smoke/Skill) and a range of the product of 0.6 and
(Smoke/Skill). Other actions have a mean of 1 minute and a range of 0.5.� These
are just rough estimates for illustrative purposes; determination of
trustworthy submodels for the entire process remains a significant practical
task.
The fire model discussed earlier is implemented with
a fire agent.� A one-minute time step is used by the fire agent to update size
of fire and number of intact inflammables.� The fire is monitored at time steps
of 5 minutes.� If fire does not seem
to decrease during the extinguishment process, the team realizes this and
changes anti-fire tools with a probability of tool failure (FSize/(FSize+(100*Skill))), where Fsize is the size
of fire and Skill is the skill level of the team member in charge of the
extinguishment.� The model does not allow the number of inflammables to be more
than the initial number of inflammables or less than 0.001.
This simulation has been
implemented by a program written in Quintus Prolog, taking advantage of several
Quintus Prolog library modules such as �math� and �random� libraries.� The
simulation is implemented as two files: a problem-dependent part
("fireagents") and problem-independent machinery for running multiple
agents in means-ends simulations ("meagent").� The latter is complex
and includes general control of the agents' actions with code from METUTOR
system [Ref. 3] for procedural modeling of goal-directed behavior.
THIS PAGE IS INTENTIONALLY
LEFT BLANK
Simulation trials were carried out to
assess teams with different skill levels in firefighting.� One hundred trials
were done for each of several sets of parameters. Skill levels were
illustratively chosen as follows: 0.9 for a high skill level, 0.5 for an
average skill level and 0.1 for a poor skill level.� There are 27 possible
skill combinations for a team of four members if the nozzlemen and the hosemen
have the same skill levels.� Six cases of special interest were: an ideal team
(all members with 0.9 for skill level); an average team (all members with 0.5
for skill level); a poorly-skilled team (all members with 0.1 for skill level);
a team with highly-skilled scene leader, average-skilled hosemen and nozzlemen
and poorly-skilled electrician; a team with average-skilled scene leader,
poorly-skilled hosemen and nozzlemen, and highly-skilled electrician; and a
team with poorly-skilled scene leader, highly-skilled hosemen and nozzlemen,
and average-skilled electrician.� The average skills in all of the above
non-homogeneous teams are the same.
Two kinds of fire locations were used: a
highly inflammable compartment with pre-extinguishing ignition and burn-out
rates of 1 and 0.5 respectively; and a compartment with pre-extinguishing
ignition and burn-out rates of 0.5 and 0.25 respectively (the rates are
modified when the extinguishing action starts using equations 3.12 and 3.13);
both locations had the same flashover probabilities.� For each six combinations
of skills, one hundred replicate runs were executed, and summary data were
recorded.
We assume that the total time for
firefighting is the time elapsed between the first appearance of the fire (time
0) and the reporting of the completion of both extinguishment and debriefing.�
Also, an upper limit on simulation duration of 400 minutes is enforced in cases
when the fire burns out but recurs sporadically, never exceeding the alarm
threshold.� Otherwise, the mean time for the set of one hundred runs could be
unfairly high, though the median is a useful alternative summary estimate.
Our hypothesis is that� the better the
scene leader, the better is the performance of the team.� The following
representative results are obtained (see Tables 4-1 to 4-4):
Table 4-1.������� Mean and Median
Summaries for 100 Runs for Ignition Rate = 0.5 and Burn-out Rate = 0.25 Ordered
by Median Total Time for Firefighting.
Table 4-2.������� Mean and Median
Summaries for 100 Runs for Ignition Rate = 0.5 and Burn-out Rate = 0.25 Ordered
by Mean Total Time for Firefighting.
Table 4-3.������� Mean and Median
Summaries for 100 Runs for Ignition Rate = 0.5 and Burn-out Rate = 0.25 Ordered
by Mean Intact Inflammables.
Table 4-4.������� Mean and Median
Summaries for 100 Runs for Ignition Rate = 0.5 and Burn-out Rate = 0.25 Ordered
by Median Intact Inflammables.
Figure 4-1.������ Histogram for Total
Time for Firefighting for Ignition Rate = 0.5, Burn-out Rate = 0.25 and all
Members with Skill Level 0.9.
Figure 4-2.������ Histogram for Total
Time for Firefighting for Ignition Rate=0.5, Burn-out Rate=0.25 and all Members
with Skill Level 0.5.
Figure 4-3.������ Histogram for
Total Time for Firefighting for IgnitionRate=0.5, Burn-out Rate=0.25 and all
Members with Skill Level 0.1.
The histograms
show that the 400-minute cases for total time for firefighting are more likely
to occur when the skill level is 0.1.� The tables also show the most important
influence on total time for firefighting is low skill (0.1) of the hosemen and
nozzlemen.� The amount of intact inflammables (unburned material) appears to
correlate well with average team skill.
We next double
the pre-extinguishing ignition rate to 1, and burn-out rate to 0.5 (see Tables
4-5 to 4-8).� Results are similar, except mean and median times are smaller
because the fire progresses more quickly.
Table 4-5.����������� Mean and Median Summaries for
100 Runs for Ignition Rates = 1, Burn-out Rate = 0.5 Ordered by Mean Total Time
for Firefighting.
Table 4-6.������� Mean and
Median Summaries for 100 Runs for Ignition Rate = 1, Burn-out Rate = 0.5
Ordered by Median Total Time for Firefighting.
Table 4-7.������� Mean and Median
Summaries 100 Runs for Ignition Rate = 1, Burn-out Rate = 0.5 Ordered by Mean
Number of Intact Inflammables.
Table 4-8.������� Mean and Median
Summaries for 100 Runs for Ignition Rate = 1,Burn-out Rate = 0.5 Ordered by
Median Number of Intact Inflammables.
Figure 4-4.������ Histogram for
Total Time for Firefighting for Ignition Rate=1,Burn-out Rate=0.5 and all
Members with Skill Level 0.9.
Figure 4-5.������ Histogram for Total
Time for Firefighting for Ignition Rate=1,Burn-out Rate=0.5 and all Members
with Skill Level 0.5.
Figure 4-6.������ Histogram for Total
Time for Firefighting for Ignition Rate=1,Burn-out Rate=0.5 and all Members
with Skill Level 0.1.
.One hundred trials with a constant 35%
probability of flashover (added to the original model, and independent of fire
state), all members with skill levels 0.5, were also done.� The objective of
the experiment is to show that a constant probability of flashover increases
the number of flashover cases, and decreases the amount of unburned
inflammables.� The experiment had 39 flashover cases (note that the simulation
tests for flashover when the state of fire changes).� When compared to a
similar experiment (for same skill levels and rates) where the probability of
flashover is not constant, the mean intact inflammables decreases
significantly.�
In order to analyze the performance of
teams when the fire does not have a random component, runs with a near
deterministic fire model were carried out for homogeneous teams (all with the
same skill level) (see Tables 4-9 and 4-10). The fire model is near
deterministic, because there is intrinsic randomness in some programming rules
that can not be changed.
|
|
|
|
|
|
|
Mean
|
|
Median
|
|
Skill
|
Mean
|
Standard
|
Median
|
Intact
|
Standard
|
Intact
|
|
Levels
|
Time
|
Error
|
Time
|
Inflammables
|
Error
|
Inflammables
|
|
all 0.9
|
54.32
|
0.36
|
53.50
|
58.39
|
1.65
|
65.17
|
|
all 0.5
|
67.15
|
0.19
|
67.00
|
48.32
|
0.81
|
52.59
|
|
all 0.1
|
105.86
|
0.52
|
106.12
|
36.19
|
1.06
|
38.28
|
Table 4-9.������� Near
Deterministic Cases for Ignition
Rate 0.5 and Burn-out Rate 0.25
|
|
|
|
|
|
|
Mean
|
|
Median
|
|
Skill
|
Mean
|
Standard
|
Median
|
Intact
|
Standard
|
Intact
|
|
Levels
|
Time
|
Error
|
Time
|
Inflammables
|
Error
|
Inflammables
|
|
all 0.9
|
39.39
|
0.30
|
39.00
|
26.12
|
0.60
|
25.35
|
|
all 0.5
|
46.54
|
0.18
|
46.00
|
19.49
|
0.10
|
20.17
|
|
all 0.1
|
55.85
|
0.16
|
56.00
|
19.89
|
0.28
|
19.16
|
Table 4-10.����� Near
Deterministic Cases for Ignition
Rate 1 and Burn-out Rate 0.5.
The results
above show that the near deterministic cases have lower mean time than the non-deterministic
stochastic cases.� The durations of actions for the near deterministic cases
have fixed means that depend on the skills, and have no variance.� The final
amount of unburned material is larger (about the same for case where all skill
levels are 0.9 and ignition rate 1 and burn-out rate 0.5) when there is no
random fire spread, flashover, and reflash.� The effects of randomness on the
mean and median times increases more substantially as the skill levels decrease
(compare to table 4-8).
The performance of a team of firefighters
with the same skills seems to be different for cases with different
pre-extinguishing rates in the three different experiments. The total time to
complete all actions is the most sensitive to team members skill levels.� When�
the fire spreads faster, and the amount of material that burns out is very
high, the percentage of intact inflammables at the end of action is going to be
low, no matter what the composition of a team.� On the other hand, when fire spreads
more slowly, different teams produce different results; the fire seems more difficult to extinguish. �Flashover occurs in almost every run
with one hundred trials at least once for all cases.� The simulation shows that
the skills of the electrician do not much affect the overall team performance
since an electrician skill level 0.5 and the rest of the team skill level 0.9
(added to the second experiment) do not show a great difference compared to a
team with all members having skill level 0.9.� Two simulations also added to
the second experiment (one for the scene leader skill level 0.5 and the rest of
the team 0.9, and other for the hosemen/nozzlemen skill level 0.5 and the rest
of the team 0.9) show that when the skill levels for the hosemen/nozzlemen
decrease, the performance of the team also decreases. A skilled scene leader is
important for the coordination of actions, and for the readiness before
extinguishment, but skilled hosemen and nozzlemen, who are responsible for the
majority of actions, are even more important.�
This thesis can be used as a basis for study of
extended and enhanced models and control tactics.� New team members and new
actions and procedures to be carried out in a fire environment can be added.�
Parameters for ignition rate, burn-out rate and type of fire can be easily
changed; new anti-fire technology can affect the performance of the team, such
as thermal imagery devices, and alternative, enhanced, mathematical models for
the spread of fire and smoke, can be used to simulate behavior of fire,
including the spread of fire to other compartments, and the dynamic allocation of firefighters to counter and control
that spread.
+
The fire agent is
specified by a mathematical model that handles the normal Gaussian stochastic
case of a growing fire.� This mathematical model, implemented in Prolog,
computes and updates the conditions of fire size, amount of smoke, and number
of intact inflammables inside a compartment for all ongoing fires.� It also
handles the cases when a fire can suddenly flare and ignite all inflammables
inside (flashovers) (Figure A-1).
|
/* The fifteen lines below show how the normal case of a� Gaussian
fire spread with total� flashover possible (see equations 3.5-3.11) is handled. */
�special_agent3(Loc,S,T,NS)
:-
��
member(fire(FSize,Loc),S), member(inflammables(C,Loc),S),
��
member(smoke(SSize,Loc),S),
/*Next three lines define
the parameters for the fire spread*/
��
start_state(SS),member(inflammables(CO,Loc),SS),
�� C>0.001, Phi is
FSize/(100*C), randnum2(UU), U is UU+0.5, steptime(ST),
�� ignition_rate(Loc,Irate),
burnout_rate(Loc,Brate),
/*Next six lines show the
fire spread when there is no flashover*/
/*Fire growth is computed*/
�� ((U>Phi, FGrowth is
Irate*FSize*(C/CO)*ST, sqrt(FGrowth,Si),
/*Burn out growth is
computed*/
�� BGrowth is
Brate*FSize*ST,
�� sqrt(BGrowth,Sm),
normalrandnum2(Ui), normalrandnum2(Um),
/*Random components are
computed*/
�� RGrowth is Ui*Si,
RBurnout is Um*Sm,
/*New size of fire and
percentage of inflammables are computed*/
�� NSize is
FSize+FGrowth-BGrowth+RGrowth-RBurnout,
�� NC is C-FGrowth-�
RGrowth,
/*The size of smoke is also� computed here.*/
�� NSSize is SSize+((2.4-SSize)*0.1),
/*Diode expressions do not let the size of fire and the number of
inflammables go below zero or above 100.*/
�� diode1(NSize,XNSize),
diode(XNSize,XXNSize),
��
diode1(NC,XNC),diode(XNC,XXNC) );
/*A flashover may happen and
ignite all inflammables inside*/
��� (U=<Phi, XXNSize is
FSize+C, XNC=0, NSSize is SSize + ((2.4-Ssize)*0.1)) ),
/*The changed state of fire
is calculated */
��
delete(fire(FSize,Loc),S,S2),�
��
delete(inflammables(C,Loc),S2,S3),
��
delete(smoke(SSize,Loc),S3,S4), fire_deletions(Loc,S4,S5),
/*Next fact deals with
concurrent alarms*/
�(((FSize>1.5;
concurrent_alarm_raising_act(T,Loc)),
�� \+
member(alarmed(fire,Loc),S), S6=[alarmed(fire,Loc)|S5] ); S6=S5),
�� NS=[fire(XXNSize,Loc), smoke(NSSize,Loc),
��� inflammables(XXNC,Loc)|S6],
��� ((\+ member(alarmed(fire,Loc),NS);
member(alarmed(fire,Loc),S));
��� alarm_aborts(Loc,NT,NS) ), !.
|
Figure A-1.����� Normal Case of
a Growing Fire with Possible Total Flashover.
|
/*A fire can randomly
reflash on rare occasions; if an action is being done there, the reflash will be noticed immediately, else wait for
alarm.� This rule also handles side effects of reflash of a fire.*/
�special_agent3(Loc,S,T,NS)
:- \+ member(fire(_,Loc),S), randnum(R),
�� mod(R,113,RR), RR = 1,
randnum(R2), mod(R2,100,RR2),
�� FSize is� 0.005*(1+RR2),
��
fire_deletions(Loc,S,XS1), delete(firetype(Loc,_),XS1,S1),
�� randnum(R3),
mod(R3,100,RR3),
�� ((RR3>80,
PT=firetype(Loc,'c')); PT=firetype(Loc,'a')),
��
((concurrrent_alarm_raising_act(T,Loc),
���� NS=[fire(FSize,Loc),
smoke(0,Loc), alarmed(fire,Loc), PT|S1],
���� alarm_aborts(Loc,NT,NS)
);
���� NS=[fire(FSize,Loc),
smoke(0,Loc),PT|S1] ),
���� write('%%%% Fire flared
up in '), write(Loc), write('.'), nl, !.
|
Figure A-2.����� Reflash Case.
The fire can randomly reflash on rare occasions; if
an action is being done in the same compartment, this will be noticed
immediately, else the fire must grow enough to trigger the alarm (Figure A-2).�
Extinguishment is implemented similarly except the ignition and burn-out rates
are computed as in equations 3.12 and 3.13 (Figure A-3).
/* The lines below show how the case where the fire is being extinguished
is handled. Note that both ignition and
burn-out rates are recalculated.� The duration of the action is also computed
in real time.
special_agent3(Loc,S,T,NS)
:- \+ member(wrong_tool(Loc),S),
� member(fire(FSize,Loc),S),
member(inflammables(C,Loc),S),
�
member(smoke(SSize,Loc),S), start_state(SS),�
�
member(inflammables(CO,Loc),SS),
� C>0.001, act
(A,extinguish(Loc,Tool),TS,TE), TS =< T, T < TE,
� steptime(ST),
/*Ignition and burn-out
rates are recalculated */
�
ignition_rate(Loc,IR),burnout_rate(Loc,BR),skill(_,SL),F1 is (-10)*SL,
� exp(F1,F2),Extinguishrate
is IR*F2,exp(SL,F3),Newburnoutrate is BR*F3,
� FGrowth is Extinguishrate*FSize*(C/CO)*ST,
� sqrt(FGrowth,Si),BGrowth
is Newburnoutrate*FSize*ST,
� sqrt(BGrowth,Sm),
normalrandnum2(Ui), normalrandnum2(Um),
� RGrowth is Ui*Si, RBurnout
is Um*Sm,
� NSize is
FSize+FGrowth-BGrowth+RGrowth-RBurnout,
� NC is C-FGrowth- RGrowth,
� NSSize is
SSize+((2.4-SSize)*0.1),
� diode1(NSize,XNSize),
diode(XNSize,XXNSize),
�
diode1(NC,XNC),diode(XNC,XXNC),
�
delete(fire(FSize,Loc),S,S2), delete(inflammables(C,Loc),S2,S3),
�
delete(smoke(SSize,Loc),S3,S4), fire_deletions(Loc,S4,S5), NT is T+1,
� NS=[fire(XXNSize,Loc),
�
smoke(NSSize,Loc),inflammables(XXNC,Loc)|S5],
� ((TE < NT,
retract(act(A,extinguish(Loc,Tool),TS,TE)),
� write('%% Extinguish act
is being extended by one minute.'), nl,
�� NTE is TE+1,
assertz(act(A,extinguish(Loc,Tool),TS,NTE)),
�� retract(state(TE,SE)),
asserta(state(NTE,SE)), update_status_times(NTE) );
�� true), !.
|
Figure A-3.����� Extinguishing
Action Rule.
A spreadsheet
simulation of the stochastic fire model was used to help visualize how the fire
spread model works before the implementation of the Prolog simulation.� The
simulation was written in Visual Basic with an Excel interface by Professor
Jacobs.� One part modeled the normal spread of fire, and one part modeled the behavior
of fire during extinguishment.�
The worksheet "Parameters"
includes the parameters used.� The fire spread is� governed by the parameters
lambda (l)
and mu (m), determined by the inflammable material inside a
compartment.� Lambda is the rate that combustibles catch fire (ignition rate)
while mu is the rate that inflammables on fire are burned out (burn-out rate)
before firefighting action (see Figure 1).� The model equations are (3.5) and
(3.10).� The skill level of the team member in charge of the extinguishment is
a parameter that ranges between 0 and 1.� It is used to calculate the new
values of lambda and mu after the beginning of the firefighting action (see
Figure B-1).� The values of mu and lambda appear in equations (3.12) and (3.13)
for different skill levels.� The fire can be subject to flashovers as a
function of the fire size and the amount of inflammables intact.� Other
parameters are the time step for the simulation and a mean time for team to
arrive; the latter is exponentially distributed unlike the Prolog simulation.�
Other actions, such as "equip", "go to compartment" and
"deenergize", are not modeled in the spreadsheet simulation.
|
|
nothing done
|
firefighters action
|
|
Ignition rate l
|
1
|
0.2
|
|
Burn-out rate m
|
0.1
|
1
|
Amount of material not on fire: C(0) = 100
Amount of material on fire at time 0: X(0) = 1
Step size h :� 0.05 minutes
Mean time until team arrives (exponential time): 3
minutes
Probability of flashover (at time t) =
c*(1-exp(-gamma*h*
(X(t)/C(0))*(C(t)/C(0))))
Flashover constant gamma : 1
Flashover binary variable c: 0 (no flashover) or 1
(flashover)
Figure B-1.������ "Parameters"Worksheet.
The size of fire X(t) and the number of
intact inflammables C(t) are updated according to Gaussian fire spread model in equations 3.5, 3.8, 3.9, and 3.10.
Figure B-2
shows an example run.� The case displayed shows that a fire quickly spread over
the compartment.� Within six minutes all inflammables were on fire, and after
eighteen minutes more, all inflammables burned out despite the intervention of
a team of firefighters.
Figure B-2.������ Example
of the State of Material in a Compartment On Fire.
Such data comes from a worksheet
"Results" that also shows the time that the team of firefighters
arrive, the time the fire is out, the amount of material burned, the amount of
material not burned, and the size of fire when the team arrives (see Figure
B-3).
Figure B-3.������ Example
� Results� Output.
This part is the work of
Professor Neil Rowe and students.� It describes the possible actions for team
members.
Equip�
Preconditions: team must
be at repair locker
Agent in charge of the
action: scene leader
����������� Go from repair locker to location of
fire
Preconditions: team must
be equipped at repair locker
Agent in charge of the
action: scene leader
Order to report the
size of fire
Preconditions: team must
be at location of fire
Agent in charge of the
action: Command Center
Report the size of
fire
Preconditions: agent must
be ordered to report the size of fire
Agent in charge of the
action: scene leader
Record the size of
fire
Preconditions: scene
leader must report the size of fire
Agent in charge of the
action: Command Center
Order deenergize
Preconditions: team must
be at location of fire
Agent in charge of the
action: scene leader
Deenergize
Preconditions: team must
be at location of fire
Agent in charge of the
action: electrician
Report deenergized
Preconditions: agent must
be ordered to deenergize
Agent in charge of the
action: scene leader
Set boundaries of fire
Preconditions: team must
be at location of fire, fire must have been recorded
Agent in charge of the
action: scene leader
Order tend hose
Preconditions: team must be at location of fire,
equipped, deenergized reported,
boundaries set
Agent in charge of the
action: nozzleman or hoseman
Report hose tended
Preconditions: agent must have been ordered to
tend hose
Agent in charge of the
action: hoseman or nozzleman
Approach fire
Preconditions: team must be at location of fire,
equipped, deenergized reported,
boundaries set, and fire has not already been
approached
Agent in charge of the
action: nozzleman or hoseman
Extinguish
Preconditions: team must be at location of fire,
equipped, deenergized reported,
boundaries set, fire approached, reported tended
hose
Agent in charge of the
action: nozzleman or hoseman
Order verify fire out
Preconditions: team must be at location of fire
and equipped
Agent in charge of the
action: scene leader
Verify fire out
Preconditions: team must be at location of fire,
agent must have been ordered to
verify if the fire is out
Agent in charge of the
action: nozzleman
Report fire out
verified
Preconditions: agent must have been ordered to
verify if fire is out
Agent in charge of the
action: nozzleman
Estimate water
Preconditions: team must be at location of fire
and fire is out
Agent in charge of the
action: scene leader
Order desmoke
Preconditions: fire is out
Agent in charge of the
action: scene leader
Desmoke
Preconditions: team must be at location of fire,
fire is out, and there is smoke in
the compartment
Agent in charge of the
action: electrician
Report desmoked
Preconditions: agent must have been ordered to
desmoke
Agent in charge of the
action: electrician
Order dewater
Preconditions: water has been estimated
Agent in charge of the
action: scene leader
Dewater
Preconditions: team must be at location of fire,
fire is out, and there is water in
the compartment, and water has been estimated
Agent in charge of the
action: nozzleman or hoseman
Report dewatered
Preconditions: agent must have been ordered to
dewater
Agent in charge of the
action: nozzleman or hoseman
Order test for safe
oxygen/ gases
Preconditions: fire is out
Agent in charge of the
action: scene leader
Test oxygen tester
Preconditions: team equipped
Agent in charge of the
action: nozzleman or hoseman
Test oxygen/gases
Preconditions: team equipped and at location of
fire, fire is out, oxygen tester is
OK, oxygen/gases level is not safe
Agent in charge of the
action: nozzleman or hoseman
Report oxygen/gases
safe
Preconditions: agent must have been ordered to
test for safe oxygen/gases
Agent in charge of the
action: nozzleman or hoseman
Ventilate
Preconditions: team equipped and at location of
fire, fire is out
Agent in charge of the
action: nozzleman or hoseman
Order watch for
reflash
Preconditions: fire is out
Agent in charge of the
action: scene leader
Watch for reflash
Preconditions: fire is out, oxygen/gases are
safe, and there is no watch for reflash
Agent in charge of the
action: nozzleman or hoseman
Report watched for
reflash
Preconditions: agent must have been ordered to
watch for reflash
Agent in charge of the
action: nozzleman or hoseman
����������� Go from location of fire to repair
locker
Precondition: if watch for reflash reported:
desmoked reported, dewatered reported, safe oxygen/gases reported, team is not
at repair locker; if watch for reflash not reported: desmoked reported,
dewatered reported, safe oxygen/gases reported, team is not at repair locker
and ordered watch for reflash
Agent in charge of the
action: scene leader
Store equipment
Precondition: team must
be equipped at repair locker
Agent in charge of the
action: scene leader
Debrief team
Precondition: team must not be equipped at repair
locker, fire is out, watch for reflash reported, desmoked reported, dewatered
reported, safe oxygen/gases reported
Agent in charge of the
action: scene leader
The simulation used
probabilities for random changes.� When approaching or extinguishing the fire,
there is a probability of 15% that a casualty is present; when desmoking,
dewatering, or extinguishing the fire, there is a probability of 8% that a team
member accidentally turns the power on; and when desmoking, dewatering or
storing, there is a probability of 8% that a casualty is present; when testing
gases or oxygen, there is a probability of 30% that gases or oxygen are not
safe.� These are illustrative values that can be changed at the will of an
analyst.
Here is an example of a
possible sequence of actions.
As soon as the alarm goes
off, the Command Center orders the scene leader to put out the fire.� The scene
leader orders all team members to their repair locker to get equipped.� Then
the scene leader takes the team to the compartment on fire and reports the size
of fire to the Command Center.� The scene leader tells the electrician to
deenergize the area.� Once the area is reported deenergized and the size of
fire is recorded, the scene leader has to set the boundaries of the fire.� Now
it is safe for the nozzleman to order the hoseman to tend the hose.� After the
report of tended hose, the nozzleman and the hoseman are ready to approach
fire, and then begin extinguishment using the appropriate tool.� If the tool
fails to extinguish the fire within a certain time (with a given probability of
failure), they can change tools, and continue the action.� After verifying the
fire is out, the nozzleman reports it to the scene leader.� The scene leader
then estimates the amount of water in the compartment, gives orders to desmoke
the area, gives orders to test for safe gases and oxygen, and gives orders to
dewater.� Either the nozzleman or the hoseman can carry out the testing of
oxygen and gases and the removal of water; the electrician removes smoke.� If
it is still unsafe for gases or oxygen, the scene leader orders the electrician
to ventilate the area.� When these actions are completed, they are reported to
the scene leader.� The scene leader has to order a watch for a possible reflash
before leaving the area.� The scene leader takes the team from the compartment
back to the repair locker where they store the equipment.� When the reflash
watch terminates, the scene leader debriefs the team and reports that the team
was debriefed to the Command Center.� The Command Center records that the team
was debriefed, and the procedure is finished.
The fire agent behavior
is displayed in the following flowchart (fig. C-1).
Here is an
example of an �actout� file showing the tasks carried out by each agent in a
full simulation run.� The parameters used for this run are: ignition rate 0.5,
burn-out rate 0.25, all team members with skill level 0.9.
Agent
fire did "burn([fire(1,352b),smoke(0,352b),inflammables(99,352b)],[fire(0.944749,352b),
inflammables(98.6915,352b),smoke(0.24,352b)])" from 0.0 to 1.0 minutes
The action
"burn" done by the fire agent between time 0.0 and 1.0 minutes shows
a decrease in the fire size inside compartment 352b (from 1 to 0.944749), an
increase in the size of smoke inside the compartment (from 0 to 0.24), and a
decrease in the percentage of intact inflammables inside the compartment (from
99 to 98.6915 percent).
Agent
fire did "burn([fire(0.944749,352b),inflammables(98.6915,352b),smoke(0.24,352b)],[fire(1.01586,352b),
inflammables(97.7957,352b),smoke(0.456,352b)])" from 1.0 to 2.0 minutes
Agent
fire did
"burn([fire(1.01586,352b),inflammables(97.7957,352b),smoke(0.456,352b)],[fire(1.55998,352b),
inflammables(97.2838,352b),smoke(0.6504,352b)])" from 2.0 to 3.0 minutes
Agent
fire did
"burn([fire(1.55998,352b),inflammables(97.2838,352b),smoke(0.6504,352b)],[alarmed(fire,352b),
fire(2.02247,352b),inflammables(95.9187,352b),smoke(0.82536,352b)])" from
3.0 to 4.0 minutes
Agent
command center did "order (debriefed(team,352b),command center, scene
leader)" from 4.0 to 4.28 minutes
Agent
fire did
"burn([fire(2.02247,352b),inflammables(95.9187,352b),smoke(0.82536,352b)],[fire(3.416,352b),
inflammables(94.3547,352b),smoke(0.982824,352b)])" from 4.0 to 5.0 minutes
Agent
scene leader did "equip" from 4.28 to 6.85 minutes.
�This
means the scene leader helps the team get equipped between minutes 4.28 and
6.85.
Agent
fire did
"burn([fire(3.41641,352b),inflammables(94.3547,352b),smoke(0.982824,352b)],[fire(5.92,352b),
inflammables(89.965,352b),smoke(1.12454,352b)])" from 5.0 to 6.0 minutes
Agent
fire did
"burn([fire(5.9261,352b),inflammables(89.965,352b),smoke(1.12454,352b)],[fire(7.75211,352b),
inflammables(84.7063,352b),smoke(1.25209,352b)])" from 6.0 to 7.0 minutes
Agent scene leader
did "go(repairlocker,352b)" from 6.85 to 7.97 minutes.
�It means that between
minutes 6.85 and 7.97 the scene leader took the team from the repair locker to
compartment 352b.
Agent
fire did "burn([fire(7.75211,352b),inflammables(84.7063,352b),smoke(1.25209,352b)],[fire(7.66,352b),
inflammables(80.5647,352b),smoke(1.36688,352b)])" from 7.0 to 8.0 minutes
Agent
command center did "order (fire(_454788,352b),command center, scene
leader)" from 7.97 to 8.24 minutes.
Here
the Command Center orders the scene leader to put the fire out in compartment
352b.
Agent
scene leader did "order(deenergized(352b),scene leader,
electrician)" from 7.97 to 8.23 minutes.
Here
the electrician receives an order from the scene leader to deenergize the area
of compartment 352b.
Agent
fire did
"burn([fire(7.66114,352b),inflammables(80.5647,352b),smoke(1.36688,352b)],[fire(6.556,352b),
inflammables(77.0206,352b),smoke(1.47019,352b)])" from 8.0 to 9.0 minutes
Agent
electrician did "deenergize(352b)" from 8.23 to 9.41 minutes
Agent
scene leader did "order (verified(fire_out,352b),scene leader,
nozzleman)" from 8.23 to 8.45 minutes.
Agent
scene leader did "report (fire(7.66114,352b),command center, scene
leader)" from 8.45 to 8.67 minutes.
The
scene leader reports the size of fire in compartment 352b to the Command
Center.
Agent
command center did "record fire(7.66114,352b))" from 9.0 to 10.02
minutes
Agent
fire did
"burn([fire(6.55156,352b),inflammables(77.0206,352b),smoke(1.47019,352b)],[fire(8.276,352b),
inflammables(72.1762,352b),smoke(1.56317,352b)])" from 9.0 to 10.0 minutes
����������� Agent
electrician did "report (deenergized(352b),scene leader,
electrician)" from 9.41 to 9.69 minutes
Agent
fire did "burn([fire(8.27526,352b),inflammables(72.1762,352b),smoke(1.56317,352b)],[fire(11.357,352b),
inflammables(65.8635,352b),smoke(1.64685,352b)])" from 10.0 to 11.0
minutes
Agent
scene leader did "set (boundaries, 352b)" from 10.02 to 10.9
minutes
Agent
nozzleman did "approach (fire, 352b)" from 11.0 to 11.81 minutes
Agent
fire did
"burn([fire(11.3457,352b),inflammables(65.8635,352b),smoke(1.64685,352b)],[fire(13.114,352b),
inflammables(58.9284,352b),smoke(1.72217,352b)])" from 11.0 to 12.0
minutes
Agent
nozzleman did "order (tended(hose,352b),nozzleman, hoseman)" from
11.81 to 12.07 minutes.
The
nozzleman orders the hoseman to tend the hose.
Agent
fire did
"burn([fire(13.1148,352b),inflammables(58.9284,352b),smoke(1.72217,352b)],[fire(16.084,352b),
inflammables(50.7363,352b),smoke(1.78995,352b)])" from 12.0 to 13.0
minutes
Agent
hoseman did "tend(hose,352b)" from 12.07 to 13.14 minutes
Agent
fire did
"burn([fire(16.0846,352b),inflammables(50.7363,352b),smoke(1.78995,352b)],[fire(13.037,352b),
inflammables(42.4855,352b),smoke(1.85096,352b)])" from 13.0 to 14.0
minutes
Agent
hoseman did "report(tended(hose,352b),nozzleman, hoseman)" from
13.14 to 13.36 minutes.
The
hoseman reports that the hose is tended in compartment 352b.
Agent
fire did
"burn([fire(13.0379,352b),inflammables(42.4855,352b),smoke(1.85096,352b)],[fire(13.228,352b),
inflammables(35.6229,352b),smoke(1.90586,352b)])" from 14.0 to 15.0
minutes
Agent
fire did
"burn([fire(13.2281,352b),inflammables(35.6229,352b),smoke(1.90586,352b)],[fire(10.409,352b),
inflammables(32.1959,352b),smoke(1.95528,352b)])" from 15.0 to 16.0
minutes
Agent
fire did
"burn([fire(10.4092,352b),inflammables(32.1959,352b),smoke(1.95528,352b)],[fire(9.5596,352b),
inflammables(29.1744,352b),smoke(1.99975,352b)])" from 16.0 to 17.0
minutes
Agent
fire did "burn([fire(9.55963,352b),inflammables(29.1744,352b),smoke(1.99975,352b)],[fire(8.4067,352b),
inflammables(25.4959,352b),smoke(2.03977,352b)])" from 17.0 to 18.0
minutes
Agent
fire did
"burn([fire(8.40675,352b),inflammables(25.4959,352b),smoke(2.03977,352b)],[fire(7.8989,352b),
inflammables(22.7267,352b),smoke(2.0758,352b)])" from 18.0 to 19.0 minutes
Agent
fire did
"burn([fire(7.89891,352b),inflammables(22.7267,352b),smoke(2.0758,352b)],[fire(6.8429,352b),
inflammables(19.7633,352b),smoke(2.10822,352b)])" from 19.0 to 20.0 minutes
Agent
nozzleman did "extinguish (352b,stream)" from 14.0 to 20.0
minutes.
The
nozzleman uses stream to carry out the extinguishment in compartment 352b.
Agent
fire did
"burn([wrong_tool(352b),fire(6.84294,352b),inflammables(19.7633,352b),smoke(2.10822,352b)],
[failed(352b,stream),fire(5.38766,352b),inflammables(17.8817,352b),smoke(2.13739,352b)])"
from 20.0 to 21.0 minutes
Agent
fire did
"burn([fire(5.38766,352b),inflammables(17.8817,352b),smoke(2.13739,352b)],[fire(1.0587,352b),
inflammables(17.8602,352b),smoke(2.16365,352b)])" from 21.0 to 22.0
minutes
Agent
fire did
"burn([fire(1.05878,352b),inflammables(17.8602,352b),smoke(2.16365,352b)],[fire(0,352b),
inflammables(17.866,352b),smoke(2.18729,352b)])" from 22.0 to 23.0 minutes
Agent
nozzleman did "extinguish(352b,foam)" from 21.0 to 23.0 minutes.
The
nozzleman uses foam to carry out the extinguishment in compartment 352b.
Agent
fire did
"burn([alarmed(fire,352b),confronted(fire,352b),fire(0,352b),indicated(boundaries,352b),
tended(hose,352b),reported(tended(hose,352b),nozzleman,hoseman)],[formerfire(0,352b),
water(0,352b)])" from 23.0 to 24.0 minutes
Agent
nozzleman did "verify (fire_out,352b)" from 24.0 to 24.88
minutes.
Agent
scene leader did "order (safe(oxygen,352b),scene leader,hoseman)"
from 24.0 to 24.29 minutes.
�The
scene leader orders the hoseman to test for safe oxygen.
Agent
scene leader did "estimate(water,352b)" from 24.29 to 25.14
minutes
Agent
hoseman did "test(oxygen_tester)" from 24.88 to 25.66 minutes
Agent
nozzleman did "report(verified(fire_out,352b),scene leader ,
nozzleman)"� from 24.88 to 25.15 minutes.
The
nozzleman reports to the scene leader that the fire is out.
Agent
scene leader did "order(not(water(_454790,352b)),scene leader,
nozzleman)" from 25.14 to 25.42 minutes.
The scene
leader orders the nozzleman to remove water from compartment 352b.
Agent
nozzleman did "dewater(352b)" from 25.42 to 25.45 minutes.
Water
is removed by the nozzleman in compartment 352b.
Agent
scene leader did "order(safe(gases,352b),scene leader, nozzleman)"
from 25.42 to 25.66 minutes.
The
scene leader orders the nozzleman to test for safe gases.
Agent
nozzleman did "report(not(water(_454790,352b)),scene leader,
nozzleman)" from 25.45 to 25.68 minutes.
�The
nozzleman reports that water was removed from compartment 352b.
Agent
scene leader did "order(watched(reflash,352b),scene
leader,hoseman)" from 25.66 to 25.88 minutes.
�The
scene leader orders the hoseman to set a watch for reflash.
Agent
hoseman did "test(oxygen,352b)" from 25.66 to 26.45 minutes
Agent
nozzleman did "test(gases,352b)" from 25.68 to 26.56 minutes
Agent
scene leader did "order(not(smoke(_454790,352b)),scene leader,
electrician)" from 25.88 to 26.09 minutes.
The
scene leader orders the electrician to remove the smoke from compartment 352b.
Agent
electrician did "desmoke(352b)" from 26.09 to 30.57 minutes
Agent
hoseman did "report(safe(oxygen,352b),scene leader,hoseman)" from
26.45 to 26.69 minutes.
The
hoseman reports that the oxygen test in compartment 352b is safe.
Agent
nozzleman did "report(safe(gases,352b),scene leader,nozzleman)"
from 26.56 to 26.81 minutes.
The
nozzleman reports that the gases test in compartment 352b is safe.
Agent
hoseman did "watch(reflash,352b)" from 26.69 to 42.94 minutes.
The
hoseman stays in compartment 352b to watch for a reflash.
Agent
electrician did "report(not(smoke(_454790,352b)),scene
leader,electrician)" from 30.57 to 30.83 minute.
The
electrician reports to the scene leader that the smoke was removed.
Agent
scene leader did "go(352b,repairlocker)" from 31.0 to 31.79
minutes.
The
scene leader takes the team from compartment 352b back to the repair locker.
Agent
scene leader did "store(equipment)" from 31.79 to 32.64 minutes.
The
scene leader and the members of the team store the equipment.
Agent
hoseman did "report(watched(reflash,352b),scene leader,hoseman)"
from 42.94 to 43.16 minutes.
The
hoseman reports the end of the watch for reflash to the scene leader.
Agent
scene leader did "debrief(team,352b)" from 43.16 to 44.09
minutes.
The
scene leader debriefs the team about the fire in compartment 352b.
Agent
scene leader did "report(debriefed(team,352b),command center,scene
leader)" from 44.09 to 44.36 minutes.
The
scene leader reports to the Command Center that the team was debriefed about
the fire in compartment 352b.
Agent
command center did "record(debriefed(team,352b))" from 45.0 to
46.01 minutes.
The
Command Center records that team was debriefed about fire in compartment 352b.
this page
Is intentionally left blank
1.�������� Naval Ships' Technical Manual Chapter
555 � VOLUME 1-Surface Ship Firefighting, pp. 1-9, 1-13, 1-22, 1-23, 1-24,
9-1, 1998.�
2.�������� Weingart,
Stephen G., "Development of a Shipboard Damage Control Fire Team Leader
Intelligent Computer Aided Instructional Tutoring System." �Master's
Thesis, Computer Science Department, Naval Postgraduate School, 1986.
3.�������� Rowe,
N. and Galvin,T., An Authoring System for Intelligent
Procedural-Skill Tutors, 1998.
4.�������� Rowe,
N., Instructions for use of the Metutor means-ends tutoring system,
February 1990.
5.�������� Rowe,
N., Artificial Intelligence Through Prolog, pp.263-270, Prentice-Hall,
Englewood Cliffs, NJ, 1988.
6.�������� Yuan,
Soe-Tsyr, MAS Building Environments with Product-Line-Architecture
Awareness, pp. 1-5, Lecture Notes in Artificial Intelligence, First
Pacific Rim International Workshop on Multi-Agents, Singapore, November 1998.
7.�������� Gaver, Donald P. and Jacobs, Patricia A., �Fire Spread
Modeling: Deterministic and Stochastic�, unpublished working paper, 2000.
this page
Is intentionally left blank
Ishida,
Toru, Multiagent Platforms, Lecture Notes in Artificial Intelligence,
First Pacific Rim International Workshop on Multi-Agents, Singapore,
November 1998.
Zhang,
Chengqi, and Lukos, Dickson, Multiagent Systems Methodologies and
Applications, Lecture Notes in Artificial Intelligence, Second Australian
Workshop on Distributed Artificial Intelligence, Cairns Australia, August 1996.
J.
Rumbaugh, M. Blaha, W. Premerlani, F. Eddy, and W. Lorensen, Object-Oriented
Modeling and Design. Prentice-Hall, Englewood Cliffs, NJ, 1991.
Law,
Averill and Kelton, David, Simulation Modeling and Analysis, Second
Edition
McGraw-Hill,
1991.
Galvin,
Thomas P. "MEBUILDER: An Object-Oriented Lesson Authoring System for
Procedural Skills."� Master's Thesis, Naval Postgraduate School, 1994.
Web site http://ait.nrl.navy.mil/DamageControl/Shadwell.html
Developing
Effective Standard Operating Procedures for Fire & EMS Departments, Federal
Emergency Management Agency, United States Fire Administration.
J.A. Swartz, R.F. Fahy, E.M. Connelly and D.P.
Demers, Final Technical Report on Building Fire Simulation Model - Volume
I, revised edition, National Fire Protection Association, Quincy MA, May 1983.
Daley, P.J. and Gani, J, Epidemic Modelling, An
Introduction, Cambridge University Press,
1999.
this page
IS intentionally left blank
1.
Defense Technical Information
Center........................................................................ 2
8725 John J. Kingman
Road, Suite 0944
Ft. Belvoir, VA�
22060-6218
2.
Dudley Knox Library................................................................................................. 2
Naval Postgraduate
School
411 Dyer Road
Monterey, CA�
93943-5101
3.
Chairman, Code
OR������������������������.2������
Operations Research
Department
Naval Postgraduate
School
����������� Monterey, CA 93943-5101
4.
Professor Neil C. Rowe, Code
CS/RP�����������������.2
Computer Science
Department
Naval Postgraduate
School
����������� Monterey, CA 93943-5101
5.
Distinguished Professor Donald P.
Gaver Jr., Code OR/DG��������..2
Operations Research
Department
Naval Postgraduate
School
����������� Monterey, CA 93943-5101
6.
Professor Patricia A. Jacobs,
Code OR/PJ���������������..1
Operations Research
Department
Naval Postgraduate
School
����������� Monterey, CA 93943-5101
7.
�����Lieutenant Commander Sylvio
Flavio Andrade, Code OR/SA�.������..2
����������
Rua Coronel Paulo Malta Rezende 135/907
���������� Barra da Tijuca, Rio de Janeiro, RJ Brazil, CEP
22631-040
8.
�����Centro de Analise de
Sistemas Navais�����������������.2
Via Brazilian Naval
Commission
����������� 5130 MacArthur Blvd., N.W.
Washington. DC
20016-3344
