基于跳变参数估计的故障诊断方法

本文讨论研究了系统跳变参数估计的故障诊断方法。利用卡尔曼滤波法检测及估计跳变参数,从而减少了系统模型参数辨识的计算量。通过选择跳变模型参数集、非跳变模型参数集以及基于知识的方法,减少方程p=f~(-1)(θ)的待求变量数量,提高了故障定位速度。用广义Newton—Raphson方法求解故障定位方程,保证了较高的故障定位精度。

Fault Diagnosis Based on Jumping Parameter Estimation

A jumping parameter estimation method for fault diagnosis is presented in this paper. The calculation for estimating the model parameters is simplified since the jumping parameters are tested and estimated with the method of Kalman filter. The number of physical parameters being solved in the equation p=f-1(6) is limited via selecting the sets of jumping and non-jumping parameters, and using knowledge based methods. So the fault localization time is shortened. The fault localization accuracy is increased by using the Newton-Raphson method for solving the fault localization equation.