Fault Detection Prediction Analysis of Multi-sensor Data Fusion Architecture and Isolation Using Pseudo Sensor Enhancement Method (PSEM)

摘要

The recent technology of intelligent systems is using multiple sensors to assess real-time information of the internal system states and the environmental operating conditions. In the case of sensor (s) transient faults, the disconnection of the readings causes inappropriate actions. So the need for sensor fault detecting and estimating multiple sensor faults becomes an important issue to optimize and enhance the system performance. In this paper, fault detection prediction analysis of multi-sensor architecture is introduced to study different number of sensors with different arrangement architectures and to detect transient faults in each sensor. Then, a new proposed pseudo sensor enhancement method (PSEM) is presented and investigated to isolate and compensate for the transient faults. The simulation results of different sensor architecture configurations without and with applying the PSEM have been obtained and analyzed to demonstrate the enhancement in operational performance of the proposed method to meet the requirements for multiple sensor intelligent systems.