复杂动态环境下非完整多智能体系统的蜂拥控制

With the rapid development of information technology, artificial intelligence and other science and technology, cooperative control among multiple agents with nonholonomic constraints, such as unmanned aerial vehicles and unmanned vehicles, is playing an important role in social production and life. It has become a research hotspot in Control Science and Engineering. Inspired by the flocking behavior of fish and birds in nature, it is important to study and reveal the flocking evolution and control mechanism of nonholonomic multi-agent systems in complex dynamic environments, which has important theoretical value and wide application prospects. Based on considering the node uncertainties, the topological uncertainties of communications and the uncertainty of the motion space and their coupling effects, this project will be devoted to: establishing node and network models to reveal the uncertainty and nonlinear characteristics for nonholonomic multi-agent systems; studying the dynamic evolution rule and flocking control method of nonholonomic multi-agent systems; designing the optimization algorithm for flocking control of nonholonomic multi-agent systems under multi-factor constraints; building new theories and methods of flocking control of nonholonomic multi-agent systems to provide the theoretical basis and technical support for the cooperative operation of practical industrial systems, such as multi-robot formation, unmanned aerial vehicle networking.

随着信息技术、人工智能等科学技术的快速发展，多个具有非完整约束的智能体(如无人机、无人车)间的协同控制在社会生产、生活中正发挥着重要作用，已成为控制科学与工程界的研究热点。受自然界中鱼群、鸟群等生物群落蜂拥行为的启发，研究并揭示复杂动态环境下非完整多智能体系统的蜂拥演化规律与控制机理具有重要的理论价值与广阔的应用前景。本项目将综合考虑节点不确定性、通讯拓扑不确定性、运动空间不确定性及其耦合作用，拟开展如下研究：建立具有不确定性与非线性特性的非完整多智能体系统的节点与网络模型；研究复杂动态环境下非完整多智能体系统的动态演化规律与蜂拥控制方法；设计多因素约束下非完整多智能体系统的蜂拥控制与优化算法，发展并建立非完整多智能体系统蜂拥控制的新理论与新方法，为多机器人编队、无人机组网等工业系统应用提供理论依据与技术支撑。

在项目团队成员的共同努力与合作下，综合考虑系统的非完整特性、网络拓扑结构、不确定性、时延等因素约束，研究了复杂动态环境下非完整多智能体系统的组网与建模、动态演化规律、蜂拥控制与优化设计等问题。针对非完整系统的非线性特性与网络拓扑结构特征，建立了相应的非完整多智能体系统的网络模型与稳定性理论；针对不同的网络拓扑结构和任务需要，建立了相应的蜂拥、多蜂拥协同控制理论，实现了非完整多智能体系统的蜂拥、多蜂拥等协同、多协同控制；针对外部扰动、时延、拓扑切换等因素约束，分析了其对系统动态演化的影响，建立了非完整多智能体系统相应协同控制的优化算法，发展了多智能体系统的相关控制理论与方法，为无人机组网、多机器人编队等工业系统的协同运作提供理论依据与技术支撑。目前，在本项目资助下，在国内外刊物和会议上发表论文11篇，其中SCI源刊论文10篇，圆满完成了研究计划中预期的研究任务。