In this paper, active suspension system is selected as the research object. Back-stepping method is adopted, and virtual fault tolerant controller, main fault tolerant controller and adaptive law are designed by constructing Time-varying Barrier Lyapunov Function (TVBLF), to ensure that the displacement and speed of the vehicle do not violate the constraint boundary and achieve the goal of fast fault tolerance. For the unknown continuous function caused by the uncertain body mass and other factors in the system, Radial Basis Function Neural Network (RBFNN) is used for approximation. At the same time, the suspension space limitation and tire dynamic load are analyzed. Finally, the effectiveness of the proposed fault-tolerant control method is verified by simulation of deviation fault and partial failure fault. INDEX TERMS Fault-tolerant control, semi-vehicle active suspension system, time-varying displacement, velocity constraints.