Sensor Optimization Selection Based on Fault Detectability and Trackability

Abstract

Abstract. Correctly selecting and reasonably arranging sensors are critical to high fidelity health assessment and low testing costs. A novel approach of sensor optimization placement for health monitoring based on fault detectability and trackability is proposed in this paper. Firstly, the requirements of sensor selection for health monitoring, the definitions and calculations of fault detectability and trackability are presented. Thus, a Sensor Optimization Selection Model (SOSM), whose objectives are to maximize the fault detectability and trackability and minimize cost of sensors, is built. Afterwards, an Adaptive Simulated Annealing Genetic Algorithm (ASAGA) is implemented to solve the SOSM. Finally, the real gearboxes and experimental data are used to verify the effectiveness of the SOSM proposed in this paper and its solution. The results from this study have shown that the approach can provide a better strategy for health monitoring in order to reduce the test cost, improve the reliability and the capability.