Simulation in Operational Research form Fine Details to

  • Slides: 17
Download presentation
Simulation in Operational Research form Fine Details to System Analysis

Simulation in Operational Research form Fine Details to System Analysis

Operational Research as DST • A system (mainly systems of system) are a composition

Operational Research as DST • A system (mainly systems of system) are a composition of many small interacting fine details • Operations Analysis is a tool to support decision making and as such should provide realistic and accurate evaluation Both God and the Devil are in the fine details

Large System Characteristics • It is impossible to provide a generic representation of all

Large System Characteristics • It is impossible to provide a generic representation of all systems • Descriptive characteristics include: – Many individual components – Some components react asynchronously – There is no single “motivation” for all components – Interaction time vary • And it is not RANDOM

System of Systems Representation • Representing Large Systems as Systems of Systems is a

System of Systems Representation • Representing Large Systems as Systems of Systems is a “simplification approach” • Individual “encapsulated” components with a singled controlled motivation can be represented as an individual system • Large number interactions cannot be collected to a single representation unless we go to Thermodynamics which is much larger numbers

Large Scale & Complex Systems • Large scale long acting systems are “difficult” to

Large Scale & Complex Systems • Large scale long acting systems are “difficult” to formulate • Interactions of Systems over time with accumulating varying feedbacks constantly change and are difficult to formulate • The Mathematical approaches left are – Numerical analysis – Simulation

The Advantage of Simulation • Simulation as a form of numerical analysis • Simulation

The Advantage of Simulation • Simulation as a form of numerical analysis • Simulation has several advantages: – Easy to visualize – Can be interactive – Simple to understand the effects Therefore easy to control and modify

Warfare Management an Example of a complex System of Systems

Warfare Management an Example of a complex System of Systems

The Combat Theater • The Combat Theater is a large ever-changing system • Multi

The Combat Theater • The Combat Theater is a large ever-changing system • Multi dimensional interaction change the behavior over time • Advance command control systems, precision collection assets, and precision weapon constantly affect behavior Targets Behavior Target New Target Doctrine Collection Assets Targets for Attack New Own Force Doctrine Fire Assets BDA

High Resolution Simulation • The multiple interaction and the changing environment over time requires

High Resolution Simulation • The multiple interaction and the changing environment over time requires high resolution simulation of all elements in the “system” – High resolution yes – High fidelity sometimes • Only the high resolution simulation can account for all interactions and the changing interactions

Simulating Bottom Up • Simulation of all interactions requires the simulation of: – All

Simulating Bottom Up • Simulation of all interactions requires the simulation of: – All entities: Systems, Platforms, Commanders – Behaviors • The simulation represents all mapped interactions and therefore considers changes over time

Representation of the Entities • The Bottom up simulation requires for the representation of

Representation of the Entities • The Bottom up simulation requires for the representation of all entities: – Specific systems (Sensors, Weapons, communication) – Individual platforms (Size, maneuverability, visibility, carried systems) – Behavior (How to use the capabilities) • Also groups are represented and have a dedicated behavior

Hierarchical Representation of Systems • Interacting systems may be represented as systems of systems

Hierarchical Representation of Systems • Interacting systems may be represented as systems of systems • Knowledge representation can be easily represented as interacting State Machines • Each machine represents a single entity but affects directly and indirectly others

Using Our Knowledge • The difficulty in representing the individual “behavior” of each system

Using Our Knowledge • The difficulty in representing the individual “behavior” of each system is simple • Knowledge can be gathered to represent each individual system: – – Objectives Means of operations Reactions Affects on others

Simulating the Whole Arena • The Combat Theater is represented by: – Multiple parallel

Simulating the Whole Arena • The Combat Theater is represented by: – Multiple parallel asynchronous interactions – Interactions are repeated – The arena and elements in the arena change as part of the interactions – There are constant feedbacks between the interactions – Knowledge is learned so effects and behaviors are different • The only way to evaluate the arena is to fully simulate it

The Complete Simulation Cycle Directive Situation Perception Behavior Knowledgebase “Brain” Actions Motion Model Motion

The Complete Simulation Cycle Directive Situation Perception Behavior Knowledgebase “Brain” Actions Motion Model Motion Sensor Activation Sensors Target Selection Own Platform Knowledge orm f t la ge P n led w O ow Kn Ow Kn n P l o w at le for d ge m The Simulated Arena Controllers ECM Activation ECM Weapon Discharge Weapons Communication

Supporting Operations Research • Aspects of the simulated arena are exported to offline analysis

Supporting Operations Research • Aspects of the simulated arena are exported to offline analysis • Analysis can consider individual scenarios, multiple re-runs, specific aspects of the scenario

Summary • Simulation is a tool providing the data for Operational Analysis: – Computing

Summary • Simulation is a tool providing the data for Operational Analysis: – Computing individual effects – Enabling visualization – Enabling multiple interactions • The important thing is to use simulation to realistically generate the arena and not statistically represent the arena loosing the importance of simulation