This paper proposes a novel adaptive dynamic surface control scheme for car-following tasks based on the probabilistic prediction of the preceding vehicle's future states. Firstly, the vehicle driving process in a short time is regarded as a discrete Markov chain, and the preceding vehicle's potential speed and acceleration for the car-following control system are predicted for a preset prediction time horizon via the Markov Chain Monte Carlo (MCMC) technique. Then, an adaptive dynamic surface control system for car-following tasks is constructed to handle the characteristics of parametric uncertainties, external disturbances and nonlinearities of vehicles, in which the certain term is estimated by the fuzzy logic technique in real-time, and the stability of the proposed adaptive car-following control system is proven by the Lyapunov theory. Finally, the performance of the proposed adaptive dynamic surface control scheme for car-following tasks is evaluated by experimental tests, and the results show that the proposed control scheme can achieve preferable performance and high control precision.
