Analysis and control of multi-agent flocking system containing mutations

Focused on common phenomena of individual variation and diversity in natural groups, in this paper, the concept of mutation strategy is proposed based on the research of multi-agent flocking movement system. Firstly, on the basis of the basic flocking model, three performance indicators such as individual pre-stable time, individual fluctuation, and system minimum steady-time are introduced. Then the dynamic performances of the original and mutation system are analyzed and compared to explore the influences of mutation exerted to the system. Besides, via the construction of learning control mechanism, the strategies of individual could be self-adjusted and optimized constantly in consistent with their update equation respectively, which led to the rapid achievement of new stable state of the mutation flocking system. Simulation results demonstrated the effectiveness of this learning control mechanism to enhance the rapidity and stability of such a mutation flocking system.