Failure Data Acquisition System Based on WSN for Vehicle PHM Technology

Abstract

Prognostic and health management (PHM) technology needs plenty of historical failure data to build accurate model for failure analysis, diagnostic and prognostic. However the existing data collection equipment is not flexible enough for special vehicle because of the wired connection or lack of real-time information feedback in real vehicle experiment. A new failure data acquisition system based on wireless sensor network is proposed to acquire the sufficient historical failure data for the development of vehicle PHM technology. The proposed system architecture consists of several wireless failure data acquisition (WFDA) nodes, a gateway node, monitoring software and probability density ratio (PDR) algorithm working on a base station. Compared with other related acquisition systems, the WFDA node is small enough and suitable for working in a narrow space inside the vehicle. A double-buffer resampling strategy is specifically developed in this node to solve the contradiction of high sampling rate and low wireless bandwidth. The PDR algorithm embedded in monitoring software is used to detect abnormal data and show researchers the analysis results which can be relied to change the test item in time. Experiments results in the laboratory preliminary verified the effectiveness of system.