To address the issues of curvature discontinuity and terminal tire non-return in the parallel parking of autonomous vehicles, a novel parallel parking path planning method based on the combination of the quintic polynomial curve and an improved sigmoid function was proposed. First, a vehicle kinematic model was established. Second, considering the position, front wheel angle, and yaw angle constraints during the parking process, a hybrid superimposed curve was designed. The parking path planning problem was converted into an optimal control problem, with the maximum curvature and curvature at both ends as objective functions, and the parameters were optimized using the simulated annealing algorithm. Subsequently, for path tracking control, the simulated annealing algorithm was used to optimize the prediction time horizon of the model predictive control algorithm. Finally, a series of actual vehicle experiments were conducted based on the Apollo Autonomous Driving Developer Suite, and the effectiveness of the proposed path planning method was validated. Therefore, this method can provide a certain reference for automatic parking path planning technology.
Parallel Parking Path Planning and Tracking Control Based on Simulated Annealing Algorithm
Int.J Automot. Technol.
International Journal of Automotive Technology ; 25 , 4 ; 867-880
2024-08-01
14 pages
Article (Journal)
Electronic Resource
English
Parallel Parking Path Planning and Tracking Control Based on Simulated Annealing Algorithm
Springer Verlag | 2024
|Parallel Parking Path Planning and Tracking Control Based on Adaptive Algorithms
Springer Verlag | 2021
|Two-section parallel parking path planning algorithm considering tracking error accumulation
European Patent Office | 2025
|Path Tracking and Control for Parallel Parking
IEEE | 2020
|