Reasearch on the Multi-Agent Modeling and Simulating Method of CAS and the Agent Rule Learning

The multi-Agent modeling and simulating method of complex adaptive system (CAS) is one of the representative subjects for research in the field of complexity science. A framework for the multi-Agent modeling and simulating method of CAS is set up, which is based on the analysis of this method’s requirements. Aim at the deficiency exists in the agent rule learning and evolving based on genetic algorithm, an Agent rule learning method based on improved immune genetic algorithm is proposed. Finally, two experiments of the application of this method in the interbasin water transfer management system are given out, based on SWARM platform, to verify the validity of this method.