TDMPC-based trajectory tracking control method of intelligent vehicle considering model uncertainty from system time delay

To overcome the constraint of communication delay on tracking accuracy, a time-delay model predictive control method is proposed. Integrating trajectory coordinate dynamics and time-delay state equations to construct an enhanced composite model with time-varying delay characteristics. Designed the optimal weighting method to predict communication delay, transformed the uncertain delay equation into a time series model with external inputs, and constructed an augmented prediction equation based on the Diophantine equation. The controller solves the objective function online through rolling optimization and introduces feedback correction to compensate for time delay errors. The key content is to abandon the traditional ideal time series assumption and achieve effective modeling of time-delay uncertainty, as well as collaborative optimization of time-delay prediction and feedback correction. The results show that TDMPC significantly reduces the lateral tracking error compared to standard MPC under communication delay, and the amplitude of lateral angular velocity fluctuation is significantly reduced, significantly improving the tracking stability of complex time-delay scenarios.