Explicit model predictive control of multi-objective adaptive cruise of vehicle

In order to coordinate both the tracking control effect and real-time performance of adaptive cruise control(ACC) system, a multi-objective adaptive cruise control method of vehicle was proposed via the explicit model predictive control(EMPC) theory. Based on the kinematic relationship between vehicles, an adaptive cruise control kinematics model was established. The tracking error prediction model was derived in the forecast time domain by the predictive control theory. The multi-performance objective functions and constraints of vehicle safety, tracking, economy and comfort were determined. The closed-loop model predictive control system based on the repeated online optimization calculation, was transformed into an equivalent explicit polyhedral piece-wise affine(PPWA) system by the multi-parameter programming theory of explicit model predictive control. The optimal control laws from the distance error, velocity error, self-vehicle acceleration and rear vehicle acceleration to the desired acceleration were obtained by the off-line calculation. The search process of the online control was designed. The adaptive cruise control was realized by the explicit control laws in the partition of the current state vector. The longitudinal tracking conditions were simulated and verified, and the EMPC-ACC was compared with the traditional MPC-ACC. Compared result shows that in the sinusoidal acceleration and deceleration condition of lead vehicle, the single-step operation speed of EMPC-ACC controller improves by 53.51% on average compared with the MPC-ACC controller. Under the EMPC-ACC, the average distance tracking error is 0.220 3 m, and the average speed error is 0.340 1 m·s^-1. In the step acceleration and deceleration condition of lead vehicle, the single-step operation speed of EMPC-ACC controller improves by 72.96% on average compared with the MPC-ACC controller. The average distance tracking error is 0.331 9 m, and the average speed error is 0.399 1 m·s^-1 under the EMPC-ACC. It can be seen that on the premise of guaranteeing the longitudinal tracking performance, the proposed EMPC-ACC controller can effectively improve the real-time performance of the ACC.