论文部分内容阅读
近年来,非线性非高斯系统的故障诊断问题一直是热点问题也是研究难点。针对非线性非高斯系统的故障检测与定位问题,运用基于粒子滤波(Particle Filter)的似然概率密度函数值的方法来检测系统故障,在检测到故障发生时,运用多模型方法与粒子滤波相结合,进行故障定位,从而进一步改善故障诊断的效果。仿真结果表明,该方法可以对非线性系统进行快速、准确的故障诊断。
In recent years, the fault diagnosis of non-linear non-Gaussian systems has always been a hot issue as well as a research challenge. Aiming at the problem of fault detection and localization in non-linear non-Gaussian systems, the system fault is detected by the method of likelihood density function based on Particle Filter. When a fault is detected, the multi-model method and particle filter phase Combined with fault location, thereby further improving the effectiveness of fault diagnosis. The simulation results show that this method can diagnose the nonlinear system quickly and accurately.