Adaptive Sensor Fault Detection for Rail Vehicle Suspension Systems

Abstract

This paper develops an adaptive sensor fault detection scheme for rail vehicle suspension systems with uncertain parameters and sensor faults. A parameterized model of the input-to-output description of the rail vehicle suspension systems is developed, based on which different parameterized output estimators are constructed to detect different unknown sensor fault patterns. As a representative study, three fault detection estimators are presented using estimation errors between the sensor output and the output estimates. Robust adaptive parameter update laws are used to ensure desired system performance of the output estimators, for the construction of the fault detection scheme. The proposed adaptive detection scheme not only can handle large parameter uncertainties in rail vehicle suspension system models, but also can check whether sensor fault occurs in rail vehicle suspension system models or not, and identify the patterns of the sensor faults. Simulation study verifies the effectiveness of the developed adaptive sensor fault detection scheme.