Sensor Selection and Location Scheme for Prognostic and Health Management

The performance of electronic vehicle PHM relies not only on the diagnostic and prognostic algorithms used, but also on the types, location, and number of sensors selected. The paper firstly presents the architecture for PHM. Sensor localization and selection for fault diagnostic purposes is the importance part. It introduces the new sensor approach for PHM, such as smart sensor. Sensor localization and selection for fault diagnosis has been studied. A novel scheme for a diagnostic and prognostic system to integrate the functions of sensor localization and selection, feature extraction, mode identification, fault diagnosis and prognosis is introduced. The detailed process includes modeling, FMECA research, FOM, optimization algorithm and performance assessment .The algorithm combines particle swarm optimization method with a heuristic search algorithm to solve the NP question