Model-based Distributed Diagnosis for Electro-hydraulic Suspension System

This paper presents a model-based distributed fault diagnosis approach for electro-hydraulic suspension system by using the diagnostic hybrid bond graph (DHBG) technique and bat algorithm (BA). Electro-hydraulic suspension system is a complex nonlinear system and if the centralized diagnosis approach is used, some problems, like poor scalability and high computationally cost, exist. To overcome these problems, a distributed approach is developed where the DHBG model of global system is decomposed into several local submodels by temporal causal graph technology. In each submodel, the distributed analytical redundancy relations are derived from the DHBG of submodel and the distributed fault signature matrix is constructed to isolate the possible faults set (PFS). For fault estimation, the BA is used to find out the true fault parameters from PFS. Simulation results validate the developed distributed fault diagnosis method in Electro-hydraulic suspension system is effective.