K连通抗毁性拓扑条件下异构群体的协同一致

The fault-tolerant topology with K-connectivity is the foundation and important guarantee for heterogeneous group which completes the task through the distributed cooperative control in complicated environment. However, due to the topological connectivity is the global property of the network, it is a challenge for the cooperative control of heterogeneous group in the condition of connected topology which can only use local information. For improving the performance of fault tolerance of network topology and the ability of cooperative control of heterogeneous group in complicated and uncertain environment, the multiple dynamic assembly tasks is considered in this study. Firstly, the optimal index of the fault tolerance K is obtained using the combinatorial optimization method, and then a distributed K-connected topology with approximately minimum power is constructed. Furthermore, considering the K-connected topology of the heterogeneous group as control objectives and constraint conditions, the distributed consensus algorithm is designed combining the non-smooth control and leader-follower control, and then the consensus of heterogeneous group in the condition of fault-tolerant topology with K-connectivity is achieved, and a series of theoretical results is obtained. The study is fully related to the hot topic of the cooperative control in multi-agent systems, its theoretical research results can effectively solve the cooperative control problem of complex multi-agent systems in the condition of multiple conflict targets.

在复杂环境下，K连通抗毁性拓扑结构是异构群体通过分布式运动协同控制完成既定目标任务的基础和重要保障。但由于拓扑连通性是网络的全局结构属性，这给仅能利用局部信息进行拓扑连通性条件下的异构群体运动协同控制带来极大的挑战。为提高异构群体系统在复杂环境下的拓扑结构抗毁性和协同运动控制能力，本研究面向多点动态集结任务，首先通过组合优化思想得到系统完成目标任务的最优抗毁性指标K，进而设计出全局功耗最小的分布式K连通抗毁性拓扑结构构建方法。同时，将异构群体的K连通拓扑结构同时作为控制目标和约束条件，结合非光滑控制与领航跟随控制思想设计出分布式一致跟踪控制协议，实现有限时间K连通拓扑条件下异构群体的一致性控制，得到异构群体同时实现K连通的抗毁性拓扑结构和一致的充要或充分条件。本研究选题紧绕群体系统的协同一致这一控制科学领域的热点问题，其理论研究成果可以有效解决多冲突目标下的复杂群体系统的协同控制问题。

复杂环境下多智能体系统的分布式协同一致是当今控制领域的热点问题。本项目围绕多智能体系统拓扑结构抗毁性和异构群体的协同一致问题，分别针对：非线性多智能体系统的分布式一致性控制问题、抗毁性拓扑条件下异构群体的协同一致问题、复杂环境下异构群体的信息处理和异常诊断问题、基于物联网的多智能体系统路径规划与协同避障问题、复杂环境下异构群体的协同控制在物流领域中的应用问题等展开了具体研究，提出了新的分析与设计方法，设计了基于分布式信息和局部信息交互的控制协议，获得系统达到协同状态一致的充要及充分条件，得到一系列较成体系的研究成果和研究方案，并在集装箱港口无人堆场机械群的远程控制中得到广泛应用。