A complete system for fault diagnostics of the longitudinal controllers in a platoon of automated vehicles is developed. The diagnostic system is designed to provide automated monitoring and fault identification of all the sensors and actuators used in the longitudinal control system, including radar sensors and inter-vehicle communication. The system uses several reduced-order nonlinear observers constructed from a longitudinal dynamic model of the vehicle. Multiple estimates of signals are obtained by designing each observer to utilize a different sensor measurement. Different combinations of all the available sensor signals and the observer estimates are then processed to construct a bank of 10 different residues. We show analytically that a fault in any of the sensors or actuators creates a unique subset of these residues to grow so as to enable exact identification of the faulty component.
Simulation results are presented to demonstrate the working of the fault diagnostic system in the presence of various faults.