Unknown input observer-based fault diagnosis of speed sensors in dual clutch transmission

Abstract

Speed sensors in the dual clutch transmission (DCT) play an essential role in designing the vehicle launching and gear shifting strategy. The speed information can also be used to monitor whether the DCT system operates normally. The vehicle performance will degrade rapidly once the speed sensor occurs fault, which may lead to poor driving experience. With increasing working hours and poor working environment, the speed sensors are prone to failure. In order to monitor the sensor failure, this article proposes a robust speed sensor fault diagnosis algorithm based on an unknown input observer (UIO). First, the dynamic model of a DCT powertrain is constructed which captures the internal and external disturbance in the system. Based on the dynamic model, a sensor fault detection algorithm is presented by designing an UIO. Second, a set of UIOs is developed to identify which sensor occurs fault. Finally, a sensor fault estimation method using the UIO is proposed. Simulation results reveal that the speed sensor faults can be detected, isolated and estimated, which may further be used for fault-tolerant control of a DCT system.