ExtendSim Tutorial

Dr David Brown April 28, 2010

Innovative Decisions, Inc. 1945 Old Gallows Road Suite 207 Vienna, VA 22182 www.innovativedecisions.com Phone:(703) 854-1130

Agenda
ExtendSim Overview Continuous and Discrete Models Model Optimization Custom Model Blocks Animation Building Animation Objects Integration with Bayesian Networks.

What is ExtendSim?
GUI based simulation package Continuous, discrete event, discrete rate, and agent-based modeling is supported Robust set of modeling and data analysis components. Allows for development of custom components. Links to external data sources and also includes built-in databases Supports animation and advanced graphic for enhanced presentation

Extend Libraries

Internal Block Settings

Activity Block

Double click to open block and set internal options

Typical Applications
Manufacturing
Assembly lines, distribution, shipping

Facilities
Customer flow, operational efficiency, logistics

Industrial Operations
Construction, mining

Military Operations
Troop readiness, military logistics

Process Improvement
Business operations

Output
Built-in graphing and report generation Export data to spreadsheets or databases
0.00025 0.0002

0.00015
Probability

Baseline Rule-based Bayesian Network 0.0001

0.00005

0 20000

22000

24000

26000

28000

30000

32000

34000

Weekly Net Revenue ($)

Hierarchical Blocks
1 Assembly Start BinsIn
Bins In Type Bin In Bins Out Bins In Bins Out Bins In Bins Out Bins In Bins Out Bins In Bins Out Bins In Bins Out

Assembly Station
Sta#

2 Station #
Sta#

3 Assembly Station Station #

Sta#

4 Assembly Station Station #

Sta#

5 Assembly Station Station #

Sta#

Assembly Station

Station #

Rand

1 2

3 6 Station #
Sta#

Assembly Station

7 Station #
Sta#

8 Assembly Station Station #

Sta#

9 Assembly Station Station #

Sta#

10 Assembly Station Station #

Sta#

Assembly Station

BinsOut
Bins In Bins Out Bins In Bins Out Bins In Bins Out Bins In Bins Out Bins In Bins Out

a BinsIn ? b BinTy pe Get A T U D F b a BinsOut

select

D Bin Load Delay

L W Conv ey or Queue

Eqn A Receiv e R C T # Stations B N a>b

Rand

Y 1 Constant

Receiv e

1 2 3 Delay Time

R C T Load Delay Mean

3 Constant Receiv e R C T Load Delay Dev

StationIn

Activity-Based Costing

Types of Models
Continuous Models
Values are always present Examples: Temperature in a room, Accounts, Fuel in your car

Discrete Models
Things happen at specific times Examples: People enter a building, Cars in a parking lot, Files in an inbox

Many models contain elements of each

Example: Deposit to a savings account

Inputs and Outputs

Continuous Block Outputs must be connected to Inputs
y =f (x)
Discrete Input
D F

Discrete Block
Discrete Output Continuous Output

Continuous Input

Continuous Output

Continuous Input Drag Down Tabs

D C

F U L P NB SD SH PT BT DT AS

Continuous Models
1.25 ### Hare Births get Hare Population

Hare Pop

Buttons

Lynx Pop

Hare Birth Rate

Lynx Pop Hare Pop

Hares Killed

0.25 ### 100 ### Hare Density Ly nx Birth Rate Ecosy stem Area Value Ly nx Births want get Ly nx Population

Lynx Pop

Ly nx Mortality

Lynx Pop

Ly nx Deaths

If only the value blocks are used, there are no special requirements

Discrete Models
T= 60 4000 C CO Recy cle 100 800 0 0 % Q Valv e 2 ID C CO I Indicator R Q Valv e 3 0 12000

Executive Block
C CO Source R

Q Valv e 1

CO I Indicator

C Sink

CO

Reception

Processing y =f (x)

Stage

y =f (x)

Stage

Stage

y =f (x)

Indicator 1
B N 1

A A=2 Y

Indicator 1
B N demand 0

A A=0 Y

Indicator 2
B N demand 0 0

A A=0 Y

demand

start

Start
F L

L Gate 1 Stage 1

AD

L Gate 2 Stage 2

AD

Start L1 L2 L3 L4

L1
max con

L2

Stage 3

L3

Gate 3 Stage 4

AD

L4

Stage

Stage 1: Stage 2: Stage 3: Stage 4:

Open Valve 1 and release flow into Reception Open Valve 2 and release flow into Processing Process the flow Open Valve 3 and release flow into Sink

An executive block must be used if any discrete blocks are used in a model. The executive block must be to the left of all other blocks.

Example Problem
Work station for evaluation of incoming paperwork
1000 forms as input
100 per hour enter system

Start with 3 workers One worker can review one form at a time Variable time to complete review
Mean 3 Std Dev 1

Work Flow Functions

Incoming Files 1 file 1000 100 per hour Inbox 1 file File Review Mean 3 Std Dev 1 mins 1 worker 3 1 worker Completed Files

Workers

Work Flow Model

Incoming Files 1000 100 per hour Workers 3 1 worker Inbox 1 file File Review Mean 3 Std Dev 1 mins 1 worker Completed Files 1 file

24

2 0 0 D F

13 13 L

TR

Run ExtendSim

Marine Corps Fuel and Water Logistics

Modeled availability of complex fuel and water systems based on repair and transport of components

Marine Corps Fuel and Water Logistics

Included use, repair, procurement, transport, and use-life guidelines for multiple system components with different attributes.

Output was system availability data for multiple battalions with different system use profiles.

Navy Ship and Aircraft Operations

Ship and aircraft loading, movement, and debarkation modeling.

Naval Ship Design

Model movement of troops and vehicles within a new ship design to see how long it takes to outfit and deploy a landing force with their vehicles.

Model Optimization
Extend can be used to optimize problems Mathematical solutions require equal numbers of variables and equations to solve
These are not always available

Extend uses genetic algorithms to optimize multiple variables in one equation

Optimization Problem
An unprocessed file results in an average loss of $300 A processed file results in a gain of $1000 A worker costs $3000 per 600 time units What is the optimal number of people to assign to this job?

Work Flow Optimization

L Inbox

0 0 D F

585 r L

Completed

TR

Workers

Run ExtendSim

Set Optimization

Complex Optimization
Fire
Full FireCalls

Dispatch Operations Simulation

Dispatchers F
# FEL In progress Completed 40

E L

Processing Input Data

F
# FE

In progress Completed 0

Performance Stats

EMS
Full

# of processors
# FL

F L

In progress Completed 0

EMS Calls

Call Processor(s)
#

E L
EL

In progress Completed 0

Pct of Calls < 15 secs

Full

Target 0.95
LE Calls

Pct of calls < 40 secs Target 0.99

# F

In progress Completed 23933 In progress

# E Full Non-emergency Calls

Completed 28699 In progress

L
# L

Completed 31143

Full

Dispatch Optimization

Modified Blocks
The blocks that come with ExtendSim can be modified to Meet unique customer needs
a b c a b c demand a b c a b c demand a b c

Standard Block

Modified Block

One IDI customer needed a custom batching block that could create custom orders instead of batching the same items every time

Other Custom Blocks

a b c demand a b c demand a b c
use # use # I u change

F L W

D L W

F
u change

Buffer with variable input Variable unbatch

Bin with variable starting items

use

# u

use

# u

# Gate

# Gate sensor

i change change

sensor

Labor pool with variable start laborers

Gate with variable number Of items within the gate area

Animation

Advanced 3D visualization is supported.

Example Animation

Run ExtendSim

Integrating Netica Models with ExtendSim

Hostile Neutral Friendly None Intel 18.4 15.1 16.5 50.0 EW F14Radar 27.0 CommRadar1 11.7 CommRadar2 13.2 Mig29Radar 21.5 MissileRadar 7.75 NoRadar 18.8 F14 Mig29 Boeing Airbus Other None RSM 30.8 24.3 9.25 7.25 5.40 23.0

IFF ValidReply 5.60 NoReply 94.4

Identity Hostile 42.0 Neutral 25.5 Friendly 32.5

Classification Mig29 30.0 Missile 5.00 Boeing 15.0 Airbus 10.0 F14 40.0

Comms True 33.9 False 66.1

Position AirRoute 20.2 ReturnToForce 11.5 Neither 68.3

Kinematics GT700kts 40.3 s500to700Kts 41.5 LT500kts 18.3

Netica Models
Netica is software package for the creation of Bayesian Networks and influence diagrams by Norsys Software Corp Comes in a graphical user interface version and a set of Application Programmer Interface (API) functions

Sample Network
Visit To Asia visit 1.00 no visit 99.0 Tuberculosis present 1.04 absent 99.0 Smoking smoker 50.0 non smoker 50.0 Lung Cancer present 5.50 absent 94.5 Bronchitis present 45.0 absent 55.0

Tuberculosis or Cancer true 6.48 false 93.5 XRay Result abnormal 11.0 normal 89.0
Distributed by Norsys Software Corp.

Chest Clinic
Contributing Factors Diseases Symptoms

Dyspnea present 43.6 absent 56.4

Example Inference
Visit To Asia visit 1.22 no visit 98.8 Tuberculosis present 6.72 absent 93.3 Smoking smoker 100 non smoker 0 Lung Cancer present 64.6 absent 35.4 Bronchitis present 60.0 absent 40.0

Tuberculosis or Cancer true 70.6 false 29.4 XRay Result abnormal 100 normal 0
Distributed by Norsys Software Corp.

Chest Clinic
Contributing Factors Diseases Symptoms

Dyspnea present 73.2 absent 26.8

Adding known information that the person is a smoker and that their chest XRay was abnormal cause the probability of cancer to jump from 5% to 65%

Influence Diagrams
Umbrella
Distributed by Norsys Software Corp.
Weather no rain 70.0 rain 30.0

Forecast sunny 53.5 cloudy 21.5 rainy 25.0

Decide Umbrella take it 35.0000 leave at home 70.0000

Satisfaction

A n Influence Diagram is a Bayesian Network that also includes Decision and Utility nodes

Library Description
Get Belief

State#

Belief

Input desired state number

Output belief

The Netica library for Extend is a set of modeling blocks that allows you perform any function on a Netica Bayesian Network or Influence Diagram that can be done using the Netica graphical user interface. Inputs an outputs allow for dynamic interaction with the rest of the Extend model.

Why is this Unique?

Extend

Qu'est que c'est le croyance que le noeud de rayons X est sur l'etat un?

What?

Netica API

ModL (Pascal-based)

C++ (C-based)

Extend and Netica can not communicate with each other directly

Custom Translator Allows Communications

Extend Qu'est que c'est le croyance que le noeud de rayons X est sur l'etat un? What is the belief that node XRay is in state one? Netica API

.675

ModL (Pascal-based)

Translator (C++)

C++ (C-based)

Applications for Netica Users

Test a network you have created in its simulated operating environment Generate probabilities and utilities for CPT tables by having the network learn from a simulation Develop networks that can change over time Batch process a large number of cases by reading and writing to an Excel worksheet Reduce complexity by building multiple, smaller networks and then using Extend to have them operate as one

Applications for Extend Users

Have networks make complex, multi-attribute decisions during the simulation
Network can control the simulation

Model a complex phenomenon that can not be described mathematically

Netica Library
File Functions
Initialize/Open/ Compile Network Initialize/Open Network Open/ Compile Network Sav e Network Close Network

Compile Network

Fade Node

Retract All Findings

Network Functions
Retract Node Findings Rev ise Probabilities Select Network

degree

Enter/Set Functions
Enter Finding Enter Finding Not Enter Likelihood Set Node Value Set State

degree

Set State Not

Set Utility Value

1 2 3 4 5
Value State State Value

Get Functions
Get Belief Get Belief Get Decision Get Likelihood Most Likely State Most Likely Value Get Statistics Get Value

Belief

State#

Belief

Decision#

State# Likelihood

State

Value

Mean

Std Dev

Value

Integrated Simulation Example

Aircraft Motion Equation Model ExtendSim Radar Cross Section Bayesian Network Netica

Radar Equation Model ExtendSim

Shoot Decision Influence Diagram Netica

42

Aircraft Radar Cross Section

Bayesian Network RCS Model

dBsm 35 30 25 20 15 10 5 0 Test Data Model Prediction

Bayesian network provides a good representation of the RCS

Radar Target Track

(F-16 FMS Radar vs. B-29)
45 40

35
YRange (NM)

No Detection Detection 30

25

20 60 70 80 90 100 XRange (NM) 110 120 130

Model displays target scintillation (fades in and out) like a real engagement

Combat ID Network
Hostile Neutral Friendly None Intel 18.4 15.1 16.5 50.0 EW F14Radar 27.0 CommRadar1 11.7 CommRadar2 13.2 Mig29Radar 21.5 MissileRadar 7.75 NoRadar 18.8 F14 Mig29 Boeing Airbus Other None RSM 30.8 24.3 9.25 7.25 5.40 23.0

IFF ValidReply 5.60 NoReply 94.4

Identity Hostile 42.0 Neutral 25.5 Friendly 32.5

Classification Mig29 30.0 Missile 5.00 Boeing 15.0 Airbus 10.0 F14 40.0

Comms True 33.9 False 66.1

Position AirRoute 20.2 ReturnToForce 11.5 Neither 68.3

Kinematics GT700kts 40.3 s500to700Kts 41.5 LT500kts 18.3

Courtesy of CDR George Laskey

Evaluation of Utility
Action/Hypothesis

Friendly
- Destroy friendly aircraft. - Damage US prestige. - Tens of million of dollars in damage and loss of pilot - Possible loss of job

Commercial
- Destroy commercial airliner and kill possibly hundreds of civilians and cause tens of million of dollars in damage - No harm to US forces but significant damage to US prestige. - No harm to US forces or prestige

Hostile
- Destroy enemy aircraft. - Reaffirms US forces are capable of defending themselves.

Shoot

Dont Shoot

- No harm to US forces or prestige.

- Significant damage to US fleet and prestige on the order of hundreds to thousands of lives lost and hundreds of millions to billions $ in damage
47

Expert Elicitation Network Test

250 200

150

Engagements

Shoot No_Shoot 100

50

0 MIG-29 Missile Boeing Aircraft Airbus F-14

Network Modification
Network was then modified by adding arcs from node Identity to nodes EW and Kinematics. A new policy that a MIG-29 flying at the radar at speeds greater than 500 knots and with their radar on is considered to be hostile. Otherwise, these aircraft are neutral. The influence diagram was cleared of all probabilities and utility values. The equation-based model generated random aircraft engagements, but was modified to randomly make the shoot decision and then determine the utility of that decision. Once all node states for the influence diagram were set, the probabilities and utility values were updated

49

Simulation Learning Network Test

250

200

150

Engagements

Shoot No_Shoot 100

50

0 MIG-29 Missile Boeing Aircraft Airbus F-14

Genetic Agents Teach an Influence Diagram

OpEMCSS

CSS Priority
count Agent1 Mov e

Current Game Results

Reward1

Agent2 Mov e

Reward2

Set Game Parameters

Initialize

Initialize/Open/ Compile Network Set A Local Agent1 Best Game A degree

First Move
Local Agent2 Best

Set A

B
Def ine Rewards

Set A

Multiple Games to Find Best Strategy

Get Ev olv e Agent1 Ev olv e Ev olv e A Agent2 Ev olv e Agent1 Mov e Get Split A Agent2 Mov e Local

Logic

Set A

Agent1

Agent1 Reward Local Agent2 Reward Assemble

?
A

Agent1

Set A

Agent2 Agent2
A

Results

Mov e Statistics

Set Node Value

Set Node Value

Set Utility Value

Rev ise Probabilities

Sav e Network

Close Network

Statistics

E
b
Value Value Value

Def ine MOEs

Agent1

degree

Agent2

Utility

degree

Game2