Robust Pattern Recognition Based Fault Detection and Isolation Method for ABS Speed Sensor

Anti-lock braking system (ABS) is considered an essential safety in electric vehicles that works to grant a reliable vehicle driving experience, and it very important ensure the security of such onboard system. This work presents detailed analysis associated with comparison includes several techniques based on pattern recognition for biasing fault detection wheel speed sensors. These are K-nearest neighbor (KNN), support vector machine (SVM) decision tree (DT), which were selected among other have been studied. The MATLAB Simulink model ABS was implemented, data extracted from healthy unhealthy operating conditions order be used train each technique individually. An offline test applied these trained FDI models using same implemented express performance one. Specifically speaking, accuracy sensitivity algorithm’s efficiency comparison, 99.9 % Fine KNN, 75 Coarse Gaussian SVM, 61.5 Tree. From result, considering issues mentioned above, can concluded KNN classifier superior both SVM TREE classifiers.