本项目研究异构多无人机系统在个体执行器故障情况下的容错控制问题。异构多无人机系统是一类由多架不同种类且动力学特性差异显著的无人机组成的异构多智能体系统。相比于同构多无人机系统，它能提升任务配置能力和执行效率。然而，由于异构多无人机系统中不同种类个体发生执行器故障后对系统性能的影响也各不相同，因此其容错控制问题显著难于同构多无人机系统的容错控制问题。. 本项目针对异构多无人机系统中各类无人机的常见执行器故障，分析其产生机理并建立相应的故障模型及故障后的系统模型。在此基础上，充分考虑个体间动力学特性的差异，对可恢复和不可恢复执行器故障，均提出较为系统的基于被动及主动容错思想的控制理论和技术，确保异构多无人机系统故障情况下的一致跟踪及队形保持性能。随后利用现有的半物理实验平台，实现所提理论方法的实验验证，为异构多无人机系统的安全运行提供理论基础和技术支持。

The fault-tolerant control problem for heterogeneous multiple unmanned aerial vehicle (multi-UAV) systems with actuator faults is investigated in this project. The heterogeneous multi-UAV system is one typical case of heterogeneous multiple agent systems, which consists of different types of UAVs with substantially distinct dynamics. Compared with the homogeneous multi-UAV system, the heterogeneous one can improve the capability of task allocation and the efficiency. However, the effects on system performance are totally different when faults occur on different types of UAVs. Therefore, the fault-tolerant control problem for the heterogeneous multi-UAV system is much more difficult than that for the homogeneous one.. By analyzing the fault mechanism, the mathematical models are built for common actuator faults of each type of UAVs. Then, the model of the faulty heterogeneous multi-UAV system is built accordingly. Based on the modelling results, by fully considering the diversity of each individual UAV, both passive and active fault-tolerant control schemes and techniques are systematically proposed for the heterogeneous multi-UAV system with repairable or unrepairable actuator faults. With these methodologies, the consensus and formation keeping performances are guaranteed in the presence of faults. To verify the proposed methodologies, the experimental study on a hardware-in-loop simulation platform is conducted. As a result, this project can provide both theoretical foundations and technique supports for improving the safety of heterogeneous multi-UAV systems.