http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
딥러닝 기반의 오토 튜닝 PID 제어를 이용한 차량 경로 제어 모델 개발
장승붕(Shengpeng Zhang),허효영(Hyoyeong Heo),백정열(Jeongyeol Baek),안상우(Sangwoo Ahn),탁태오(Taeoh Tak) 한국자동차공학회 2021 한국 자동차공학회논문집 Vol.29 No.1
The path tracking of the intelligent vehicle is a complex and key technology research. In the present study, a vehicle path tracking controller was designed by a deep learning auto-tuning PID(proportion integral derivative) controller. The Kp, Ki, and Kd parameters were optimized via deep neural network model in real-time. The yaw angle of the vehicle centroid and the lateral distance between the vehicle centroid and the road lane were used as inputs to the controller. In addition, the performance of the controller was compared and verified with 2-Dof, 7-Dof, and full vehicle models. The proposed controller was operated on a double lane change road and a curved road. Results showed that the proposed path tracking controller has high real-time and robustness.
[응용논문] 도로이탈방지 시스템의 운전자 간섭에 대한 연구
장승붕(Shengpeng Zhang),허효영(Hyoyeong Heo),안상우(Sangwoo Ahn),탁태오(Taeoh Tak) 한국자동차공학회 2021 한국 자동차공학회논문집 Vol.29 No.4
The article deals with the analysis of driver interaction based on the road departure prevention system(RDPS). Firstly, the controller of the RDPS, which uses the linear quadratic regulator(LQR) method, is modeled. The controller is identified by the two-degree-of-freedom vehicle model and the steering model. The various weight coefficients of LQR are investigated through the vehicle driving trajectory. The performances of the RDPS model are validated with various lateral velocities in the straight road and the curved road, respectively. Then, the driver interaction based on RDPS is carried out in the driving simulator. The motion of the driver is analyzed by using motion capture devices. The results show that the vehicle is unstable when the driver intervenes with the RDPS. Also, when the RDPS was intervened by the driver, the right curved road was more unstable than the left curved road.