Incipient fault diagnosis and prognosis for hybrid systems with unknown mode changes

This work considers incipient fault diagnosis and prognosis for hybrid systems with unknown mode changes. A set of augmented analytical redundancy relations (ARRs), termed AARRs, is proposed to extend the capability of ARRs to identify degradation of components which are not represented by physical parameters. The methods utilize the unified constraint relations, named global AARRs (GAARRs), for health monitoring of hybrid systems. The mode information is provided by a mode tracker, which is based on mode-change signature matrix (MChSM) and is only triggered when inconsistency between monitored system and its nominal model is detected. Once a fault is detected, a hypothesis set is generated and a multiple hybrid differential evolution algorithm is adopted to identify the degradation dynamics and the unknown mode changes. The optimization problem is efficiently solved by a hybrid differential evolution algorithm, which is able to simultaneously handle real and binary unknown parameters. Simulation results are reported to validate the proposed method.