Simulation and Computational Red Teaming for Problem Solving
Description
An authoritative guide to computer simulation grounded in a multi-disciplinary approach for solving complex problems
Simulation and Computational Red Teaming for Problem Solving offers a review of computer simulation that is grounded in a multi-disciplinary approach. The authors present the theoretical foundations of simulation and modeling paradigms from the perspective of an analyst. The book provides the fundamental background information needed for designing and developing consistent and useful simulations. In addition to this basic information, the authors explore several advanced topics.
The book's advanced topics demonstrate how modern artificial intelligence and computational intelligence concepts and techniques can be combined with various simulation paradigms for solving complex and critical problems. Authors examine the concept of Computational Red Teaming to reveal how the combined fundamentals and advanced techniques are used successfully for solving and testing complex real-world problems. This important book:
. Demonstrates how computer simulation and Computational Red Teaming support each other for solving complex problems
. Describes the main approaches to modeling real-world phenomena and embedding these models into computer simulations
. Explores how a number of advanced artificial intelligence and computational intelligence concepts are used in conjunction with the fundamental aspects of simulation
Written for researchers and students in the computational modelling and data analysis fields, Simulation and Computational Red Teaming for Problem Solving covers the foundation and the standard elements of the process of building a simulation and explores the simulation topic with a modern research approach.
More details
Other editions
Additional editions
Persons
JIANGJUN TANG, PHD, is a Lecturer at the School of Engineering and Information Technology at the University of New South Wales Canberra, Australia.
GEORGE LEU, PHD, is a Senior Research Associate at the School of Engineering and Information Technology at the University of New South Wales Canberra, Australia.
HUSSEIN A. ABBASS, PHD, is a Professor at the School of Engineering and Information Technology at the University of New South Wales Canberra, Australia.
Content
Preface xi
List of Figures xv
List of Tables xxv
Part I On Problem Solving, Computational Red Teaming, and Simulation 1
1. Problem Solving, Simulation, and Computational Red Teaming 3
1.1 Introduction 3
1.2 Problem Solving 4
1.3 Computational Red Teaming and Self-'Verification and Validation' 8
2. Introduction to Fundamentals of Simulation 11
2.1 Introduction 11
2.2 System 14
2.3 Concepts in Simulation 17
2.4 Simulation Types 21
2.5 Tools for Simulation 23
2.6 Conclusion 24
Part II Before Simulation Starts 25
3. The Simulation Process 27
3.1 Introduction 27
3.2 Define the System and its Environment 27
3.3 Build a Model 29
3.4 Encode a Simulator 30
3.5 Design Sampling Mechanisms 32
3.6 Run Simulator Under Different Samples 33
3.7 Summarise Results 33
3.8 Make a Recommendation 34
3.9 An Evolutionary Approach 35
3.10 A Battle Simulation by Lanchester Square Law 35
4. Simulation Worldview and Conflict Resolution 57
4.1 Simulation Worldview 57
4.2 Simultaneous Events and Conflicts in Simulation 64
4.3 Priority Queue and Binary Heap 68
4.4 Conclusion 72
5. The Language of Abstraction and Representation 73
5.1 Introduction 73
5.2 Informal Representation 75
5.3 Semi-formal Representation 76
5.4 Formal Representation 82
5.5 Finite-state Machine 86
5.6 Ant in Maze Modelled by Finite-state Machine 89
5.7 Conclusion 99
6. Experimental Design 101
6.1 Introduction 101
6.2 Factor Screening 103
6.3 Metamodel and Response Surface 113
6.4 Input Sampling 116
6.5 Output Analysis 117
6.6 Conclusion 120
Part III Simulation Methodologies 121
7. Discrete Event Simulation 123
7.1 Discrete Event Systems 123
7.2 Discrete Event Simulation 126
7.3 Conclusion 142
8. Discrete Time Simulation 143
8.1 Introduction 143
8.2 Discrete Time System and Modelling 145
8.3 Sample Path 148
8.4 Discrete Time Simulation and Discrete Event Simulation 149
8.5 A Case Study: Car-following Model 151
8.6 Conclusion 154
9. Continuous Simulation 157
9.1 Continuous System 157
9.2 Continuous Simulation 159
9.3 Numerical Solution Techniques for Continuous Simulation 164
9.4 System Dynamics Approach 172
9.5 Combined Discrete-continuous Simulation 174
9.6 Conclusion 176
10. Agent-based Simulation 179
10.1 Introduction 179
10.2 Agent-based Simulation 181
10.3 Examples of Agent-based Simulation 185
10.4 Conclusion 194
Part IV Simulation and Computational Red Teaming Systems 197
11. Knowledge Acquisition 199
11.1 Introduction 199
11.2 Agent-enabled Knowledge Acquisition: Core Processes 202
11.3 Human Agents 203
11.4 Human-inspired Agents 208
11.5 Machine Agents 211
11.6 Summary Discussion and Perspectives on Knowledge Acquisition 215
12. Computational Intelligence 219
12.1 Introduction 219
12.2 Evolutionary Computation 223
12.3 Artificial Neural Networks 232
12.4 Conclusion 239
13. Computational Red Teaming 241
13.1 Introduction 241
13.2 Computational Red Teaming: The Challenge Loop 242
13.3 Computational Red Teaming Objects 243
13.4 Computational Red Teaming Purposes 244
13.5 Objectives of Red Teaming Exercises in Computational Red Teaming Purposes 245
13.6 Discovering Biases 246
13.7 Computational Red Teaming Lifecycle: A Systematic Approach to Red Teaming Exercises 247
13.8 Conclusion 251
Part V Simulation and Computational Red Teaming Applications 253
14. Computational Red Teaming for Battlefield Management 255
14.1 Introduction 255
14.2 Battlefield Management Simulation 256
14.3 Conclusion 261
15. Computational Red Teaming for Air Traffic Management 263
15.1 Introduction 263
15.2 Air Traffic Simulation 263
15.3 A Human-in-the-loop Application 270
15.4 Conclusion 271
16. Computational Red Teaming Application for Skill-based Performance Assessment 273
16.1 Introduction 273
16.2 Cognitive Task Analysis-based Skill Modelling and Assessment Methodology 274
16.3 Sudoku and Human Players 276
16.4 Sudoku and Computational Solvers 280
16.5 The Proposed Skill-based Computational Solver 283
16.6 Discussion of Simulation Results 293
16.7 Conclusions 300
17. Computational Red Teaming for Driver Assessment 301
17.1 Introduction 301
17.2 Background on Cognitive Agents 303
17.3 The Society of Mind Agent 306
17.4 Society of Mind Agents in an Artificial Environment 312
17.5 Case Study 325
17.6 Conclusion 330
18. Computational Red Teaming for Trusted Autonomous Systems 333
18.1 Introduction 333
18.2 Trust for Influence and Shaping 334
18.3 The Model 335
18.4 Experiment Design and Parameter Settings 342
18.5 Results and Discussion 344
18.6 Conclusion 347
A. Probability and Statistics in Simulation 349
A.1 Foundation of Probability and Statistics 349
A.2 Useful Distributions 369
A.3 Mathematical Characteristics of Random Variables 390
A.4 Conclusion 396
B Sampling and Random Numbers 397
B.1 Introduction 397
B.2 Random Number Generator 400
B.3 Testing Random Number Generators 408
B.4 Approaches to Generating Random Variates 413
B.5 Generating Random Variates 416
B.6 Monte Carlo Method 423
B.7 Conclusion 432
Bibliography 435
Index 459
LIST OF FIGURES
FIGURE 2.1 The system of a vehicle, the sub-system of an engine, and their corresponding boundaries and entities (not all entities are shown here). FIGURE 2.2 Relationship between resolution, abstraction, and fidelity (adapted from [3]). FIGURE 2.3 An example of elevator modelling and simulation. FIGURE 3.1 The steps to develop a simulation. FIGURE 3.2 Time-advancing mechanism: next-event time advancing and fixed-increment time advancing. FIGURE 3.3 An example plotting a differential equation and difference equation. FIGURE 3.4 Blue force level calculated by difference equations with fighting effectiveness in different magnitudes with the same ratio for blue to red. Fighting effectiveness: and . FIGURE 3.5 Blue force level calculated by difference equations with fighting effectiveness in different magnitudes with the same ratio for blue to red. Fighting effectiveness: and . FIGURE 3.6 The flowchart of a battle simulator - simulator advanced by time. FIGURE 3.7 The flowchart of a battle simulator - simulator advanced by fire rounds. FIGURE 3.8 The parameters of LSL declared in a spreadsheet. FIGURE 3.9 The spreadsheet implementation of LSL based on time. FIGURE 3.10 The spreadsheet implementation of LSL based on rounds. FIGURE 3.11 The bell curves of troop sizes: and . FIGURE 3.12 The bell curves of fighting effectiveness: and . FIGURE 3.13 The CDF of troop sizes: and . FIGURE 3.14 The bell curves of fighting effectiveness: and . FIGURE 3.15 An implementation of data sampling. FIGURE 3.16 An implementation of data sampling. FIGURE 3.17 The force levels of both sides in a battle, where , , , and . FIGURE 4.1 Three traditional simulation worldviews of DES. FIGURE 4.2 Time advancing in event-scheduling worldview. FIGURE 4.3 Time advancing in activity-scan worldview. FIGURE 4.4 The process flow of a grocery shop register. FIGURE 4.5 Time advancing in process-interaction worldview. FIGURE 4.6 Time advancing in three-phase worldview. FIGURE 4.7 Time advancing in agent-oriented worldview. FIGURE 4.8 Horizontal event list versus vertical event list. FIGURE 4.9 The effects from FEL data structure and event storing order for the bank account example. FIGURE 4.10 The binary heap constructed based on the event priorities. FIGURE 4.11 Obtain and delete the event with the highest priority (lowest key) from the root location of a binary heap, and update node positions accordingly. FIGURE 4.12 Inserting a new event into a binary heap and reordering the node positions. FIGURE 5.1 Abstraction - the modelling process. FIGURE 5.2 An abstract example of a data flow diagram. FIGURE 5.3 An abstract example of an entity-relationship diagram. FIGURE 5.4 An abstract example of a class diagram. FIGURE 5.5 Simple FSM model for a traffic light. FIGURE 5.6 State transition diagram for traffic light. FIGURE 5.7 Flowchart building blocks and their meanings. FIGURE 5.8 Flowchart modelling the determination of a car insurance premium. FIGURE 5.9 Decision tree modelling for determining a car insurance premium. FIGURE 5.10 Improved FSM model for a traffic light. FIGURE 5.11 An ant in a maze. FIGURE 5.12 'LOST' state - an ant walks forward until hitting a wall. FIGURE 5.13 'ENCW' (encountering a wall) state - an ant keeps turning left until neither antenna touches a wall. FIGURE 5.14 An ant walks along the wall by checking if the wall is on its right. FIGURE 5.15 An ant encountering a wall break or turning corner when walking along the wall on the right side. FIGURE 5.16 An optimised FSM for modelling ant behaviour to find the way out of a maze. FIGURE 5.17 A NetLogo implementation of the FSM simulator for modelling an ant's behaviours to find the way out of a maze. FIGURE 5.18 Simulation results of an ant's behaviours to find the way out of a maze. FIGURE 6.1 Flowchart of the OFAT process for factors in a model. FIGURE 6.2 The main effect calculation by design matrix. FIGURE 6.3 Calculation of two-factor interaction between Factors 1 and 2 using design matrix. FIGURE 6.4 Calculation of three-factor interaction between Factors 1, 2, and 3 using design matrix. FIGURE 6.5 Response surface and contour plot for the metamodel. FIGURE 6.6 Hypothesis test methods for two or more groups of variables (adapted from[372]). FIGURE 6.7 Transient and steady-state behaviour of a simulation model. FIGURE 6.8 A long series of outputs divided into batches. FIGURE 7.1 A simple queueing system with a single queue and single server. FIGURE 7.2 Next-event time-advance approach for the single-server queueing system of the mechanical centre. FIGURE 7.3 Flowchart for a next-event time-advance approach and logical relationships among the components and routines of the simulator for DES. FIGURE 7.4 FSM of the mechanical support centre driven by client arrival and client departure events. FIGURE 7.5 Intervals for and , inverse transformation for exponential distribution. FIGURE 7.6 Example of generating , interval between arrivals (), and arrival times () by Excel. FIGURE 7.7 Flowchart for arrival event routine. FIGURE 7.8 Flowchart for departure event routine. FIGURE 8.1 Discrete time system - input, output, and state. FIGURE 8.2 Examples of sample path of sheep population. FIGURE 8.3 A DES approach for the vending machine by FSM. FIGURE 8.4 A DTS approach for the vending machine represented by flowchart. FIGURE 8.5 A DTS based on the GM car-following model. FIGURE 8.6 Speed and acceleration of the following vehicle in the first 50 s produced by DTS based on the GM...
