基于广义熵优化准则的非线性非高斯系统故障分离方法

Generalized Entropy Optimization Principle Based Fault Isolation for Nonlinear Non-Gaussian Stochastic Systems

针对非高斯随机系统研究了一种基于滤波的故障分离方法。通过故障、扰动的统计信息获取状态估计误差和输出估计误差的(联合)概率密度函数,再根据所建立的广义熵优化准则设计滤波器。仿真结果显示,在故障和干扰均出现及主要故障不出现两种不同的情况下系统响应完全不同,说明该广义熵优化准则能有效分离故障。

This paper studies fault isolation problems for nonlinear non-Gaussian systems by using a filtering method. The PDF or JPDF is obtained based on the statistic information of the faults and disturbances, the generalized entropy optimization is established and the filter is designed. Finally, a simulation is given to demonstrate the effectiveness of the proposed method.