Simultaneous fault and mode switching identification for hybrid systems based on particle swarm optimization

This paper describes a methodology for simultaneous identification of fault parameters and mode switching events for hybrid systems. The method is developed based on the notion of Global Analytical Redundancy Relations (GARRs) from the bond graph model of the hybrid system. A unified formula with mode change time sequence and initial mode coefficients (IMC) is derived to represent the mode switching. Due to the discontinuous characteristic of the mode switching, an adaptive hybrid particle swarm optimization (AHPSO) employing the combination of real valued PSO (RPSO) and binary valued PSO (BPSO) is proposed to optimize different parts of solution simultaneously, a novel individual level adaptive method using fuzzy system is developed to dynamically adjust the algorithm parameters. GARRs are used as a fitness index of the AHPSO. Case studies of different energy domains are carried out to illustrate the efficiency of the proposed algorithm.