论文部分内容阅读
提出一种基于演化博弈理论的多智能体系统协调控制方法.在所建立的数学模型框架中智能体根据其自身的利益,通过局部交互,在博弈竞争中学习,自主调整其行为.根据系统整体性能的要求,通过选择合适的博弈类型、设计适宜的收益计算方法、更新进化规则等,实现对多智能体系统的控制.在演化过程中,无需指定某些特定个体的具体动力学行为,只需通过种群的自适应进化即可实现整体目标.以分工合作问题为例,详细解析所提出的控制方法,通过理论分析和仿真验证该方法可以实现多智能体系统的自适应协调控制.
This paper proposes a multi-agent system coordinated control method based on evolutionary game theory.In the framework of the mathematical model established, the agents learn and adjust their behaviors independently through game interaction based on their own interests through local interaction, Performance requirements, and control the multi-agent system by selecting the appropriate game type, designing the appropriate calculation method of return, updating the evolution rules, etc. In the evolution process, there is no need to specify the specific dynamic behavior of some specific individuals, only We need to achieve the overall goal through the adaptive evolution of the population.Taking the problem of division of labor and cooperation as an example, the proposed control method is analyzed in detail, and through the theoretical analysis and simulation, this method can realize the adaptive coordination control of multi-agent system.