PARALLEL PARKING PATH PLANNING BASED ON IMPROVED ARCTANGENT FUNCTION OPTIMIZATION |
Qiping Chen 1, Lu Gan 1, Bo Chen 1, Qin Liu 2, Xiaobo Zhang 2 |
1Key Laboratory of Conveyance and Equipment Ministry of Education, East China Jiaotong University 2Jiangling Motor Co., Ltd. |
|
|
|
|
ABSTRACT |
Aiming at the problems of small parking space and poor parking effect for existing parallel parking, this paper
proposes a parallel parking path planning based on improved arctangent function optimization. Firstly, the vehicle parking
kinematics model is established, and the vehicle parameters are determined according to the three classical models. Secondly,
the parking space model is established, and the most reasonable parking space parameters are selected according to the
minimum parking space. Then, aiming at the problems of abrupt curvature of the designed arc-line-arc initial path, unsmooth
path, large yaw angle at the end and large parking space, an improved arctangent function model is proposed, the parking
constraints are established, the absolute value of vehicle yaw angle is taken as the objective function, and the parameters are
optimized by genetic algorithm. Finally, it is verified by simulation experiments. The results show that the method can
achieve smoother path, smaller parking space and more ideal parking posture, meet the requirements of stability, safety and
comfort in the process of parallel parking, and improve the ability of parallel parking. Therefore, this method can provide a
theoretical reference for path planning in automatic parking technology. |
Key Words:
Parallel parking, Path planning, Improved arctangent function, Parking constraints, Genetic algorithm |
|
|
|