异质多智能体系统的网络建模及合作行为分析与调控

With the expansion of multi-agent system application and the increasing scale of the system, the multi-agent systems are heterogeneous in many aspects, such as network structure, agent dynamics and so on. Based on the heterogeneity of multi-agent system, this project researches on the network modeling of multi-agent system, and analysis as well as control of cooperative behavior of multi-agent system. Firstly, this project constructs the network model of multi-agent system, such as multilayer network and dynamic network based on time series data, analysing heterogeneity of multi-agent system. Then, this project explores cooperative behavior evolution under the network model of multi-agent system in addition to interaction model of agents and behavior decision-making model. Lastly, this project researches controllability and optimal control strategy of multi-agent system, incorporating heterogeneity of the system and cooperative behavior dynamics into the behavior evolution of the system based on controllability theory framework of complex networks. To analyse and control cooperative behavior of multi-agent system through constructing the network model of the system heterogeneity is significant for cooperation of multi-agent system and its application.

随着多智能体系统应用领域和系统规模的不断扩大，多智能体系统在很多方面存在异质性，包括系统网络结构、智能体动态等。本项目立足多智能体系统异质性的特点，研究多智能体系统的网络建模、多智能体系统合作行为分析以及控制问题。首先，基于多智能体系统的时间序列数据，分析多智能体系统的异质性特点，构建多智能体系统的多层网络和动态网络等网络模型；然后，基于多智能体系统的网络模型、智能体间交互作用关系模型以及智能体行为决策模型分析合作行为的演化；最后，基于复杂网络的可控性理论框架，将多智能体系统的异质性特点和合作行为动态融入到系统的行为动力学中，研究多智能体系统合作行为的可控性和最优控制策略。通过建立多智能体系统异质性的网络模型，分析和调控多智能体系统的合作行为，对于多智能体系统的合作问题及其应用具有重要的意义。

随着多智能体系统应用领域和系统规模的不断扩大，多智能体系统在系统网络结构、智能体动态等方面存在异质性。本项目立足多智能体系统异质性的特点，研究多智能体系统的网络建模、多智能体系统合作行为分析以及控制问题。首先，基于多智能体系统的时间序列数据，分析多智能体系统的异质性特点，构建多智能体系统的多层网络和动态网络等网络模型；然后，基于多智能体系统的网络模型、智能体间交互作用关系模型以及智能体行为决策模型分析合作行为的演化；最后，基于复杂网络的可控性理论框架，将多智能体系统的异质性特点和合作行为动态融入到系统的行为动力学中，研究多智能体系统合作行为的可控性和最优控制策略。通过建立多智能体系统异质性的网络模型，分析和调控多智能体系统的合作行为，对于多智能体系统的合作问题及其应用具有重要的意义。