为了实现传感器故障隔离, 传统方法针对不同故障独立选取最优一致向量。指出为了正确实现故障隔离, 对不同故障选取的最优一致向量应是相互关联的,并由此提出一种改进的指标函数以获得最优一致向量。基于梯度法得到最优一致向量的数值解。仿真实验表明提出的方法可正确实现传感器冗余系统的故障隔离。 关键词: 故障隔离; 传感器; 一致向量; 直接冗余; 指标函数; 梯度法 Abstract : Traditional methods select optimal parity vectors for different faults independently to realize sensor fault isolation. It is pointed out that the optimal parity vectors selected for different faults have relations with each other in order to realize fault isolation. An improved index function is presented to obtain the optimal parity vec2 tors. The numerical solution of optimal parity vectors is obtained by gradient method. Through simulation experi2 ments ,it is shown that the method can realize fault isolation in sensor redundancy systems correctly. Key words : fault isolation ; sensor ; parity vector ; direct redundancy ; index function ; gradient method