Fault detection and diagnosis of singular stochastic distribution systems

In this paper, a new filter-based method for fault detection and diagnosis (FDD) of linear singular stochastic distribution systems is presented. It is supposed that only the output probability density function (PDF) of the system is measured instead of the output signal itself. Linear B-spline expansions are adopted to decouple the input from the output PDF. This results in a dynamic model between the input and the weights of B-spline basis functions. Singularity means some weights are algebraically related to the input. For this complicated system, both detection and diagnostic filters are presented as well as the adaptive tuning rule for fault. Simulations are provided to demonstrate the efficiency of the proposed approach.