多智能体系统主动容错控制及其容错能力分析

Based on the distributed formation control of multiple unmanned aerial/air vehicles (UAVs), by using graph theory, matrix operation technique, Lyapunov stabilty theorey, adaptive control and the other theoreies and techlogies, combining with model-based fault diagnosis (FD) and fault tolerant control (FTC) technique, the projection researches the problem of adaptive FD and FTC for nonlinear multi-agnent systems (MASs) with global and local faults. . First of all, the faults are divided into global and local faults, and a three-steps-based FD method is proposed for MASs. The first step in the method is fault detection. The second step is to design a differentiation algorithm such that the global faults can be distinguished from the local faults. The final step is fault isolation, which is divided into two cases: (1) for global faults, fault isolation algorithm is deigned to iaolate the global faults; (2) for local faults, local fault isolation strategy is designed to isolate the faulty agent(s), further to isolate the faul(s) occurred in the faulty agent(s). . Then, based on fault isoalation, the corresponging fault models are respectively established for various fault types, and many distributed fault tolerant control schemes are respectively deigned for various failure cases. . In addition, the fault tolerance of multi-agnent systems is quantitatively analyzed, and the corresponding measures are explored to improve the fault tolerance of the multi-agnent systems. . Finally, the distributed formation control problem of multiple UAVs is deeply investigated. Based on the proposed approaches in the projection, a new distributed fault tolerant formation control framework is designed for the multiple UAVs. . The projection enlarges the research fields of adaptive fault diagnosis and fault tolerant control, which makes the project has the important academic value and practical meaning.

本项目以多无人机编队控制为背景，利用图论、矩阵运算技术、Lyapunov稳定理论及自适应控制等理论与技术，融合基于模型的故障诊断与容错控制技术，研究具有全局故障与局部故障的非线性多智能体系统的故障诊断与容错控制问题。首先，在将故障分为全局与局部故障的基础上，提出基于三步法的故障诊断方法。第一步是故障检测。第二步是全局与局部故障区分。第三步为故障分离：(1)针对全局故障，建立全局故障的分离算法；(2)针对局部故障，设计分离算法——先确定故障智能体，再分离出局部故障。然后，建立各种故障模型，针对故障发生情况，设计相应的分布容错控制方案。接着，定量分析多智能体系统的容错能力，探索提高其容错能力的措施。最后，深入研究多无人机编队容错控制问题，并运用所提故障诊断与容错控制方法，建立一套无人机编队故障诊断与容错控制框架。本项目拓宽了自适应故障诊断和容错控制的研究领域，具有重要学术意义和实际价值。

本项目以多无人机编队控制为背景，利用图论、矩阵运算技术、Lyapunov稳定理论及自适应控制等理论与技术，融合基于模型的故障诊断与容错控制技术，研究具有全局故障与局部故障的非线性多智能体系统的故障诊断与容错控制问题。首先，在将故障分为全局与局部故障的基础上，提出基于三步法的故障诊断方法。第一步是故障检测。第二步是全局与局部故障区分。第三步为故障分离：(1)针对全局故障，建立全局故障的分离算法；(2)针对局部故障，设计分离算法——先确定故障智能体，再分离出局部故障。然后，建立各种故障模型，针对故障发生情况，设计相应的分布容错控制方案。接着，定量分析多智能体系统的容错能力，探索提高其容错能力的措施。最后，深入研究多无人机编队容错控制问题，并运用所提故障诊断与容错控制方法，建立一套无人机编队故障诊断与容错控制框架。本项目拓宽了自适应故障诊断和容错控制的研究领域，具有重要学术意义和实际价值。