Model Free Adaptive Fault-tolerant Control for Signalized Intersections with Detector Fault

In this paper, a data-driven model free adaptive fault tolerant control method designed for signalized intersection with detector fault. First, the radial basis function neural network (RBFNN) is used to estimate the queue length in the case of detector fault. Then, the dynamic linearization is applied to construct the data relationship of intersection input and output. Then, based on the idea of queue length equalization, the estimated queue length is introduced into the algorithm to obtain a fault-tolerant signal control algorithm. Finally, the numerical simulation is carried out, in which superiority of the proposed method is validated.