자율주행차량의 주차를 위한 딥러닝 기반 주차경로계획 수립연구
Parking Path Planning For Autonomous Vehicle Based on Deep Learning Model
2
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
자율주차의 요소 중 하나인 경로계획(Path-planning)을 제안한다. 실제 주차장을 참고하여 수직주차와 수평주차로 주차장의 차로 너비, 주차 공간의 너비, 길이 등 주차장 구조와 주차 환경을 다양하게 설정한다. 출발점와 도착지점 등 각도와 환경을 다양하게 설정하여 경로데이터를수집하고 수집한 데이터를 Deep Learning model에 넣어 학습시켜 자동주차경로계획 모델을 제안한다. 분석결과, 기 알고리즘(Hybrid A-star, Reeds-Shepp Curve)과 딥러닝 모델 모두 장애물에충돌하지 않고 비슷한 경로를 생성하지만, 거리와 소모시간이 각각 0.59%, 0.61% 감소하여 효율적인 경로가 생성되었다. 또한, Switching point도 1.3개에서 1.2개로 감소하여 직진과 후진을최대한으로 줄여 운전자의 피로를 줄일 수 있을거라 생각된다. 마지막으로 경로생성시간은42.76% 감소하여 효율적이고 신속한 경로생성이 가능하여 향후 자율주행 중 자율주차의 경로계획생성에 활용될 수 있으며, 차량작도에 따라 이동하는 주차로봇의 경로생성에도 활용될 수있을 것으로 보인다.
Abstract
Several studies have focused on developing the safest and most efficient path from the current location to the available parking area for vehicles entering a parking lot. In the present study, the parking lot structure and parking environment such as the lane width, width, and length of the parking space, were vaired by referring to the actual parking lot with vertical and horizontal parking. An automatic parking path planning model was proposed by collecting path data by various setting angles and environments such as a starting point and an arrival point, by putting the collected data into a deep learning model. The existing algorithm(Hybrid A-star, Reeds-Shepp Curve) and the deep learning model generate similar paths without colliding with obstacles. The distance and the consumption time were reduced by 0.59% and 0.61%, respectively, resulting in more efficient paths. The switching point could be decreased from 1.3 to 1.2 to reduce driver fatigue by maximizing straight and backward movement. Finally, the path generation time is reduced by 42.76%, enabling efficient and rapid path generation, which can be used to create a path plan for autonomous parking during autonomous driving in the future, and it is expected to be used to create a path for parking robots that move according to vehicle construction.
참고문헌 (21)
- Kang, D. H., "Parking path planning using path folding method(2018)", Domestic Studies of the Society of Control Robot Systems Published by the Society of Liquor, 2018(6), pp.92~293 (2018)
- Kim, D. H., "Creating parking routes for automatic parking using M-space", Journal of Korea University, Robotics Association, 3(1), pp.1~8 (2008)
- Kim, J. H., "Parking path planning in constrained parking spaces using deep learning", Master, pp.2022
- Kim, J. H., "Automatic parking system implemented with MATLAB and Simulink | ADAS and autonomous stocks. New features of MATLAB and the utilization of deep learning for row systems", MathWorks Korea, Youtube, 1941~2020 (2022)
- Kim, J. M., "Development of an autonomous parking system based on real-time route planning of autonomous vehicles", Master, pp.2016
- Kim, M. S., "A comparative analysis of path planning and tracking performance according to the consideration of vehicle’s constraints in automated parking situations", Journal of Korea Robotics Society, 16(3), pp.250~259 (2021)
- LaValle, S. M., "Planning algorithms" (2006), Cambridge University Press
- Lee, J. W., "Research on domestic parking accidents and reverse automatic braking devices", 2016 Han The National Automobile Engineering Association's estimate (2016), Korea Insurance Development Institute Automotive Technology Research Institute
- Lee, S. I., "Improving RRT* family algorithm performance in circle and sphere obstacle environments using sampling range limitations", Journal of the Korean Aerospace Society (2022), Ajou University
- Lee, S. J., "Study on deep learning-based parking space optimal path search algorithm for autonomous parking systems", 2023 Summer of the Korean Society of Information Technology (2023), Gyeonggi University Kyodo
- Lei, C., "Multi-maneuver vertical parking path planning and control in a narrow space", Robotics and Autonomous Systems, 149, pp.103964~ (2022)
- MathWorks, "Use Hybrid A Star to generate code for path planning", MathWorks Help Center, ~ (2023)
- MathWorks, "Planner Hybrid A Star", MathWorks Help Center, ~ (2023)
- MathWorks, "Train deep learning-based sampler of motion planning", MathWorks Help Center, ~ (2023)
- Ministry of Government Legislation, "Enforcement rules of the automobile management act (Attachment 1)", ~ (2021)
- Reeds, J. A., "Optimal paths for a car that goes both forwards and backwards", Pacific Journal of Mathematics, 145(2), pp.367~393 (1990)
- SAE(Society of Automotive Engineers), "Levels of driving automation", ~ (2021)
- Seoul Transportation Corporation, "Customer convenience facilities - Transfer parking lot, Seoul Transportation Corporation usage information", ~ (2023)
- Thoresen, M., "Path Planning for UGVs based on Traversability Hybrid A", IEEE Robotics and Automation Letters, 6(2), pp.1216~1223 (2021)
- Vorobieva, H., "Automatic parallel parking in tiny spots : path planning and control", IEEE Transactions on Intelligent Transportation Systems, 16(1), pp.396~410 (2015)
- Yoo, Y. G., "A study on the reverse application method for improving the accuracy of the shortest path search of the A* algorithm", Han Journal of the Korean ITS Society, 12(6), pp.1~9 (2013)