Adaptive Fault-Tolerant Control of Platoons With Guaranteed Traffic Flow Stability

This paper investigates the fault-tolerant control problem for heterogeneous vehicular platoons where input quantization and dead-zone nonlinearity are also involved. Due to the occurrence of actuator faults, the maximum acceleration changes, which may invalid the traditional quadratic spacing policy. To tackle the dilemma, an improved quadratic spacing policy with the lower bound of fault factor is proposed. Furthermore, the improved quadratic spacing policy removes the assumption of zero initial spacing errors. Based on the new spacing policy, an adaptive fault-tolerant platoon control scheme is developed, by employing radial basis function neural networks (RBFNN) and PID-type sliding mode control method. The given scheme is proved to be capable of guaranteeing individual vehicle stability, string stability and traffic flow stability. The effectiveness of the scheme is verified through comparison simulation studies.