Evolutionary Computing Based Methodologies for Modelling, Simulation and Analysis

Home

Project Objectives

In today’s highly complex, dynamic and interlinked world, defence forces all around the world are facing increasing demand to provide robust tactical and operational plans for both wartime as well as peacetime missions. Simulation is an important technique in meeting this challenge to evaluate plans with different configurations against a wide range of scenarios. Coupled with the method of data farming, these plans can be analysed through the process of simulation experimentation.

However, conducting simulation-based experimentation is a time consuming process. This is especially true when a wide range of scenarios and configuration parameters have to be considered in order to determine the best plan to use. Very often, these plans have to be regenerated when the tactic of the enemy changes. Increasingly, evolutionary techniques such as genetic algorithms and ant colony optimization are used in the process of modelling, simulation and analysis in order to reduce the time taken for determining the best plan.

The objectives of this project are to determine answers to the following questions: i) How can a robust plan be generated; ii) How can such robust plan be evolved against an adaptable enemy; iii) How can a simulation model used to generate tactical and operational plans be adapted and evolved to allow for more diverse solutions to be generated. These set of objectives will be achieved by researching and developing a set of algorithms and techniques that will enable evolutionary computing-based methodologies to be used in the process of modelling, simulation and analysis, and also to apply these techniques to case studies from the defence domain. Specifically, the project will look into the research topics specified in the following research development tasks (RDTs):

  • RDT1: Evolutionary Computing Techniques for Automated Red Teaming (ART)
  • RDT2: Evolutionary Computing Techniques for Automated Co-Evolution (ACE)
  • RDT3: Evolutionary Computing Techniques for Evolvable Simulation (EvoSim)
  • RDT4: Application of ART, ACE and EvoSim to Military Scenarios


Project Members

  • DSO Co-Investigators: Chwee Seng Choo
  • DSTA Co-Investigators: Darren Ong Wee Sze


Project Details

  • Funding: Defence Science and Technology Agency (DSTA)
  • Starting Date: April 2009
  • Duration: 3 years

logos.jpg