Modeling and Simulation

• Modeling represents a model, including its design and function
• Modeling helps analysts foresee changes in the real system
• Modeling involves creating a system representation with properties
• In essence, it embodies the act of constructing a model

Simulation

• The act of running a system model over a period to analyse its behavior
• Simulation is a process to to examine the performance in existing or proposed systems
• Examples of simulation include; Weather forecasting, flight simulator, car crash modeling, and traffic systems

History of Simulation

• 1940: Monte Carlo method was developed
• Researchers (John von Neumann, Stanislaw Ulan, Edward Teller, Herman Kahn) and physicists from the Manhattan project developed Monte Carlo to study neutron scattering
• Monte Carlo is a model to predict the probability of a variety of outcomes with random variables
• 1960: Special-purpose simulation languages were developed, such as SIMSCRIPT by Harry Markowitz at RAND
• 1970: Research was initiated on mathematical foundations of simulation
• 1980: PC-based simulation software, GUIs, and object-oriented programming
• 1990: Web-based simulation, animated graphics, simulation-based optimization, and Markov-chain Monte Carlo Methods

Simulation Model Components

• System entities, input variables, performance measures, and functional relationships
• Step 1: Identify issues with or requirements of the investigated system
• Step 2: Design the problem while considering relevant system factors and limitations
• Step 3: Gather and process system data, observe performance and results
• Step 4: Develop and verify the model using network diagrams and verification techniques
• Step 5: Validate the model by comparing performance under real conditions
• Step 6: Document the model for future use, including objectives, assumptions, input variables, and performance details
• Step 7: Select an experimental design appropriate to requirements
• Step 8: Induce experimental conditions on the model and observe results
• Step 9: Gather real-life system data for simulation input
• Step 10: Generate a flowchart showing simulation progress
• Step 11: select suitable simulation software
• Step 12: Verify the simulation model by comparing it with real-time system results
• Step 13: Experiment with the model with variable manipulation to identify the best solution
• Step 14: Implement solutions in the real-time system

Simulation analysis

• Step 1: Prepare a problem statement
• Step 2: Determine input variables and entities, including decision variables controlled by the programmer and uncontrollable (random) variables
• Step 3: Set constraints on decision variables by assigning them to the simulation process
• Step 4: Determine the output variables
• Step 5: Collect data from the real-life system to input into the simulation
• Step 6: Develop a flowchart mapping the simulation process
• Step 7: Select suitable simulation software
• Step 8: Verify the simulation model against the real-time system results
• Step 9: Experiment by altering the model's variable values to find the optimal solution
• Step 10: Apply findings to the real-time system

Modelling & Simulation Advantages

• Easy to understand: Allows understanding of a system's operation without real-time work
• Easy to test: Allows system changes and to see their effect without real-time work
• Easy to upgrade: Evaluates system requirements across various configurations
• Easy to identify constraints: Identifies bottlenecks causing delays in processes
• Easy to diagnose problems: Helps understand complex system interactions

Modelling & Simulation Disadvantages

• Model design requires domain knowledge, training, and relevant experience
• Operations use random numbers to derive results, making outcome prediction difficult
• It requires manpower and it is a time-consuming process
• Results translation is complex and requires experts and it is an expensive process

Application Areas

• Military, Training & Support, Designing semiconductors, telecommunications and civil engineering designs & presentations, and E-business models
• Used to study the internal structure of complex system such as the biological system
• Optimizes designs like routing algorithms and assembly lines
• Used to test new designs and policies
• Verifies analytical solutions

Computer Simulation (Module 1 Part 2)

• It is a step-by-step computer program that explores approximate behavior in mathematical models
• A model of a real-world (or imaginary/hypothetical) system
• Computer program = computer simulation model

Simulation Types

• Equation-based: Common in physical sciences, uses governing theory to construct mathematical models based on differential equations
• Simulates global equations in physical theories
• Examples of Equation-Based Simulation include physics, engineering, medical, finance
• Agent-based: Common in social and behavioral sciences, artificial life, epidemiology, and ecology
• Used in disciplines studying networked interaction
• Examples of Agent-Based Simulation include traffic, epidemic spread, games, market
• Multiscale: Hybrid models mix different modeling methods and couple modeling elements from different description scales
• Monte Carlo: Computer algorithms using randomness to determine properties of a mathematical model

Purposes of Simulation

• It can be for heuristic purposes
• Predicts unavailable data
• Generates understanding of existing data

Simulation of Complex System (Module 1 Part 3)

• Experimentation with a simplified operations system imitation on a computer for the purpose of understanding and/or improving that system better
• System: Simulates the behavior of another system over time
• System: Inter-related component collection working towards a common objective
• System components: Other systems, components, entities, resources, jobs, events, or variables describing system states
• Systems are boxes packed into successively bigger boxes
• Components can be systems/parts of a larger system
• Systems gain complexity: Additional components/relationships
• Complexity increases: System variability/variation of components or their interrelationships
• System Complexity classification depends on number/nature of logical and temporal interrelationships among internal components/the interrelationship between the system and its super systems

Complexity Types

• Combinatorial: Results from number of interrelationships between components
• Computational: Derives from algorithms to express logic and temporal interrelationships which generate mathematical complexity
• Mathematical: Reproduced in computer programs; increases the complexity of design, algorithms, and data structure
• Dynamic: Results from number/nature of components' interrelationships over time
• Communication: Often bidirectional, e.g., physical flow creating financial counterpart in a financial system

Simulation Development Stage

• Simulation modeling and analysis is a learning process iterating through stages for system behavior understanding

The stages in simulation development are:

• Conceptual Model Building: Captures key features of the real/hypothetical ones for a base model
• Computer Implementation: Logic translation of the lumped model into a computer one
• Validation: Checks computer model against experimental frame/real system to suit study purpose
• Experimentation: Repeatedly runs the model with different inputs, analyzes output data
• Implementation: Applies recommendations from the simulation

Discrete Event Simulation (DES)

• Studies system behavior with state changes at distinct time points
• System state described by variables (customer number, waiting list length)
• Continuous system: Important activities complete smoothly without delays
• Continuous simulation: State variables change continuously

Application of Simulation

• Civil Engineering: Dam embankment and tunnel construction
• Military: Missile trajectory simulation
• Logistics: Toll plaza design, airport passenger flow analysis, flight schedule evaluation
• Business Development: Product development planning, staff management, and market study analysis

Dynamic Simulation Models (Module 2)

• Virtual representations of the real world that simulate how individuals and communities act, react, develop health conditions and use health services
• Use research, expert knowledge, datasets to represent complex human behaviors
• Tool which we can test various policy scenarios over time
• Traffic Management: Helps optimize traffic flow by modeling road layouts, signals, and driver behavior
• Manufacturing Processes: Optimizes production lines and logistics by simulating material movement, machine operation, and task scheduling

Other Applications

• Used to identify bottlenecks, minimize downtime, and increase efficiency
• Supply Chain Management: Helps in optimization by modeling inventory levels, transportation routes, and demand fluctuations
• Used to identify risks, streamline operations, and improve responsiveness
• Environmental Modeling: Models ecosystems, climate patterns, and pollution dispersion for human impact assessment, prediction, and conservation strategies

Finite State Machine (FSM)

• Simulation based on information processing with signals
• Finite, discrete, sequential, and deterministic actions are in the FSM model, which is applied within the following; mechanical and digital electronic systems, pneumatic or hydraulic systems, and chemical systems
• Used as an information processing procedure and for encoding/decoding messages and syntax analysis for computer programs
• Finite state machines describe states after particular inputs
• A turnstile; Inserting a coin unlocks it, and pushing it relocks it. If it is unlocked, then inserting a coin, or pushing it will not affect the state
• State Diagram: Each node is a state, and each directed arc indicates a path from one state to another
• State Table: Tabular output/transition by row is state, and each column is an input symbol
• Communication Diagrams model scenarios along with objects (icon and links), paths to each message, and the messages are presented as sequence diagram

Sequence Diagram Elements

• Sequence Numbering helps understand/trace the message
• Simplest: 1, 2, 3
• Diagrams with 1, 1.1, 1.2
• Shows how objects operate together and in order
• Type of interaction diagram in UML
• Visualizes the behavior of a system using interactions between parts and objects over time