Workflow-Based Fast Model Predictive Cloud Control Method for Vehicle Kinematics Trajectory Tracking Problem

Model predictive control (MPC) is one of the most popular approaches for vehicle trajectory tracking problem, since it provides optimal strategy by predicting its future behaviors, and at the same time ensures robustness. However, MPC requires a large amount of computing resources for optimization at each step. This results in poor performance of the algorithm. In this paper, a novel workflow-based MPC approach is proposed to accelerate the traditional MPC algorithm. First, a trajectory tracking method using MPC based on alternating direction method of multipliers (ADMM) algorithm is developed for online optimization. Then, we seperate the algorithm into multiple smaller computational tasks and provide an approach on establishing the workflow of MPC. Finally, it is shown that the workflow-based method improves the accuracy of trajectory tracking significantly and achieves the finer-grained discretization of continuous systems. The computation time is reduced by at most 62.89
$ \%$
.