<p>The occurrence of actuator faults in heterogeneous vehicle platoon systems may generate large control signals to avoid performance loss, which can potentially lead to system instability and even vehicle collisions. Therefore, it is important to study the distributed collision-free vehicle formation control problem with actuator faults. This paper investigates the collision avoidance adaptive neural networks (NN) control problem for vehicle platoon with actuator faults. Firstly, based on given reference signals, distributed command governor via barrier functions generates a safe command signal for each vehicle. Then, based on the distributed command governor, a collision-free adaptive NN formation tracking fault-tolerant controller is designed by introducing the barrier functions and adaptive sliding surface. The main advantages of the developed control scheme are that it can ensure the formation tracking errors converge to a small neighborhood of zero and integrate actuator fault compensation with collision/obstacle avoidance. Finally, simulation results demonstrate the effectiveness of the proposed method in ensuring platoon safety and operational reliability under actuator fault conditions.</p>