Abstract
According to the existing mainstream automatic parking system (APS), a parking path is first planned based on the parking slot detected by the sensors. Subsequently, the path tracking module guides the vehicle to track the planned parking path. However, since the vehicle is non-linear dynamic, path tracking error inevitably occurs, leading to inclination and deviation of the parking. Accordingly, in this paper, a reinforcement learning-based end-to-end parking algorithm is proposed to achieve automatic parking. The vehicle can continuously learn and accumulate experience from numerous parking attempts and then learn the command of the optimal steering wheel angle at different parking slots. Based on this end-to-end parking, errors caused by path tracking can be avoided. Moreover, to ensure that the parking slot can be obtained continuously in the process of learning, a parking slot tracking algorithm is proposed based on the combination of vision and vehicle chassis information. Furthermore, given that the learning network output is hard to converge, and it is easy to fall into local optimum during the parking process, several reinforcement learning training methods in terms of parking conditions are developed. Lastly, by the real vehicle test, it is proved that using the proposed method can achieve a better parking attitude than using the path planning and path tracking-based method.
Highlights
The average proportion of cars and parking slots in big cities is about 1:0.8, and that in small and medium-sized cities is nearly 1:0.5, according to the data released by the National Development and Reform Commission of China
Because it realizes the end-to-end control from the parking slot to the steering wheel angle command, the control errors caused by path tracking are fundamentally avoided; Since the parking slot needs to be continuously obtained in the course of learning, we propose a parking slot tracking algorithm, which uses extended Kalman filter (EKF) to fuse the parking slot information with vehicle chassis information to achieve continuous tracking of parking slot; Given that the learning network output is hard to converge and it is easy to fall into local optimum in the parking process, several reinforcement learning training methods in terms of parking conditions, e.g., manual guided exploration for accumulating initial experience sequence, control cycle phased setting, and training condition phased setting, are designed
Given that the learning network output is hard to converge and it is easy to fall into local optimum in the parking process, several reinforcement learning training methods in terms of parking conditions are designed
Summary
The average proportion of cars and parking slots in big cities is about 1:0.8, and that in small and medium-sized cities is nearly 1:0.5, according to the data released by the National Development and Reform Commission of China. The lack of parking space makes the designed parking slot increasingly narrower. Parking environment is becoming complex progressively, and the increasingly higher requirement of the parking operation accuracy is raised, bringing great troubles to many drivers. Automatic parking system (APS) can increase parking safety and utilization rate of parking slot, so it has wide market application prospects. The smaller size of the parking slot requires very high parking accuracy for APS. Take the perpendicular parking slot as an example; it raises a higher demand of parking attitude for its narrow width. The BS ISO 16787-2016 [1] stipulates that the perpendicular parking inclination angle of APS should be confined within ±3◦ , imposing huge challenges to the performance of APS
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
