基于复杂网络的多电机同步控制系统故障诊断与切换容错研究

Fault diagnosis and switching tolerance for multi-motor synchronous control based on complex dynamical network

针对多电机同步控制系统的传感器故障问题,受复杂网络同步输出耦合矩阵概念的启发,基于其中的信息和结构冗余,设计表征不同传感器输出数据一致性的相似度矩阵,通过在线对矩阵中元素大小及特征值的判断实现对故障传感器的检测、定位与估计;依据故障诊断结果,通过引入可靠系数对均值反馈的加权修正,实现对故障传感器的隔离与系统的切换容错.所提方法的创新之处在于充分利用了数据驱动方法中便利的数据获取与基于模型方法中能反映机理特征的各自优势,实现了基于模型与基于数据方法的有机融合,避免了繁杂的理论推导,具有算法简单、不需要改变系统原有的拓扑结构、易于工程实现等特点,同时适用于多故障情形.仿真实验验证了所提方法的有效性.

In view of the problem of sensor fault in the multi-motor synchronous control system,the similarity matrix which characterizing the consistency of different sensors output is designed,which is enlightened by the concept of the coupling matrix of complex network synchronous output based on the information and structural redundancy in synchronous control.The fault sensor is detected,located and estimated by judging the related elements and eigenvalues in the similarity matrix.And the fault sensors can also be isolated and the system is achieved fault-tolerance by improving the weighted mean feedback mechanism.The innovation of the method is to make full use of the advantages of the convenience online data acquisition in the data driven method and the characteristics of the mechanism based on the model method,realizes the integration of the model and the data method,and avoids complex theoretical derivation.It is also applicable to multiple faults situations and has the characteristics of simple algorithm,high execution efficiency and easy engineering implementation.There is no need to change the original topology of the system.Simulation experiments verifies the effectiveness of this method.