线性系统参数故障检测与诊断的一种两级Kalman滤波算法

A TWO STAGE KALMAN FILTERING ALGORITHMFOR PARAMETER FAILURE DETECTIONAND DIAGNOSIS IN LINEAR SYSTEMS

本文研究了线性系统的参数故障检测与诊断问题,提出了一种适于系统内部及控制器故障的两级Kalman滤波算法。该算法中,初级滤波器建立在系统无故障的假设上,对其新息序列作简单的假设检验即可实现故障的检测。当系统发生故障时,次级滤波器即从初级滤波器的新息序列中估计出参数故障的大小。仿真结果表明,这一算法是可行且有效的。

The problem of parameter failure detection and diagnosis in linear systems is discussed in this paper. A two stage Kalman filtering algorithm, which is suitable for the detection of both in-ternal failures and controller failures of real systems, is proposed based on the error model. The first filter is designed on the basis of the assumption that the system is normal. By making a simple hypothesis testing about its innovation sequence, failures in system can be detected on time. After being detected, the failure parame-ters can be estimated in the second stage from the innovation sequence of the first one. By simulation it is shown that the algorithm proposed in this paper is feasible and efficient.