The model predictive control (MPC) algorithm is a path tracking control algorithm based on model optimization, whose performance is influenced by the accuracy of the vehicle model and the weight parameters. An optimization method for the MPC algorithm based on time-delay compensation is proposed to address the issue of inaccurate vehicle model caused by actuator hysteresis in real vehicles. First, considering the impact of control loop delay and actuator hysteresis on traditional models, the time-delay vehicle error prediction model is updated to incorporate time-delay compensation into the MPC. Second, to enhance the accuracy of the time-delay vehicle error prediction model, an improved salp swarm optimization algorithm is used for adaptive parameter identification of the time-delay system. Finally, a simulation and real vehicle platform are constructed to analyze the proposed optimization method for the MPC algorithm based on time-delay compensation. The experimental results show that compared to traditional models, the error of the MPC algorithm strategy with time delay compensation is reduced by 47.26%.