| 제목 |
카메라 기반 다중 파티클 필터 및 가중 비용 함수를 이용하는 자율주행 모빌리티의
경로 추종 제어 알고리즘 개발 |
| 분야 |
차량동역학 및 제어 |
| 언어 |
Korean |
| 저자 |
김세환(한경국립대학교), 오광석(한경국립대학교) |
| Key Words |
Steering control(조향 제어), Multi-particle filter(다중 파티클 필터), Path tracking(경로 추종), Sliding mode approach(슬라이딩 모드 접근), Camera(카메라), Autonomous mobility(자율주행 모빌리티) |
| 초록 |
Model uncertainty and changes in sensor specifications in autonomous mobility can have a negative impact on path tracking. Therefore, in this paper, we are proposing a path tracking control algorithm based on multi-particle filtering and a weighted cost function. This algorithm is designed to ensure path tracking, even when there are changes in camera specifications. A pixel-based control error was derived by using multiple particle filters for path tracking, instead of image processing techniques. A simplified error dynamic was designed, based on the derived pixel-based control error. Then, the desired steering angle was derived after applying weights to the cost function, which was designed by using the sliding approach and Lyapunov to reduce control errors. Next, the path tracking performance was evaluated by using two different cameras on an S-curved path based on autonomous mobility. The evaluation led us to confirm reasonable desired path tracking results, even when using cameras with different specifications. |
| 원문(PDF) |
다운로드
Journal Site
|
