Predetermined Performance Fault-Tolerant Control for Vehicle Platoon Against Measurement Drifts and Multiple Input Constraints

Abstract

This article focuses on the problem of predetermined performance fault-tolerant control for vehicle platoon with measurement drifts and multiple input constraints. To overcome the difficulties of unmeasurable system states caused by measurement drifts and sensor failures, this article introduces auxiliary variables to build up new tracking errors. The dead zone and saturation constraints are transformed into linear functions and bounded perturbations, and the neural network approximation method is utilized to handle the system's unknown terms online, which further reduces the computational complexity and design difficulty of the algorithm. By constructing a predetermined performance function, a novel adaptive fault-tolerant controller is designed such that all control errors enter a prespecified range within a preset time. Finally, the validity of the developed fault-tolerant control algorithm is verified by MATLAB simulation of a platoon consisting of five programmable vehicles.