基于区间解析冗余关系的参数不确定性研究

本文主要目的是拓展基于区间解析冗余关系法I-ARRs(Interval valued Analytical Redundancy Relations)对不确定性混杂系统的故障检测。该方法产生一组对故障估计的阈值,一旦标称残差超过阈值,则故障能被检测。现有的故障检测方法主要是针对参数性故障,因此对测量信号导数的估计呈现出各种限制。而且在I-ARRs中对测量信号导数十分敏感,这不可避免地涉及到对残差的估计。为解决这一问题,利用鲁棒滑模微分器方法得到对含噪声测量信号导数最优估计,降低残差噪声,以减小虚警、漏报,从而提高鲁棒性。通过典型电力电子电路BUCK电路仿真,验证了此方法的有效性。

Research on Parameter Uncertainty Based on Intervalvalued Analytical Redundancy Relations

The main purpose of this paper is to extend the fault detection of uncertain hybrid systems based on the method of interval valued analytical redundancy relations (I-ARRs). This method generates a set of thresholds for fault estimation. Once the nominal residual exceeds the threshold, the fault can be detected. The existing fault detection method is mainly for parametric faults, so it presents a variety of limits on the estimation of the derivative of the measurement signal. And the I-ARRs is very sensitive to the derivative of measurement signal, which inevitably involves the estimation of residuals. In order to solve this problem, the robust estimation of the derivative of the noise measurement signal is obtained by using the robust sliding mode differential method, which reduces the false alarm and omission, improves the robustness. The proposed method is verified by the simulation of BUCK circuit of typical power electronic circuit.