An End-to-End Deep Reinforcement Learning Model Based on Proximal Policy Optimization Algorithm for Autonomous Driving of Off-Road Vehicle

Wang, Yiquan; Wang, Jingguo; Yang, Yuchao; Li, Zhaodong; Zhao, Xijun

doi:10.1007/978-981-99-0479-2_248

Cited by 3 publications

(1 citation statement)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This strategy allows autonomous navigation on a forest path in simulation. According to study [74], the RL agent was fed drivable area segmented images and the associated navigation strategy was developed to avoid motorcycles, pedestrians, and roadblocks independently. Simulations and real-world tests have validated its effectiveness.…”

Section: End-to-end Navigation Of Ugvsmentioning

confidence: 99%

Human-guided reinforcement learning: methodology and application to autonomous driving

View full text Add to dashboard Cite

Ph.D. journey would not have been possible without his unwavering and unconditional full support, both academically and personally. Even though I had some experience writing publications during my master's studies, it wasn't until I came under Chen's supervision that I realized the key to building a high-quality research career. In spite of his busy schedule, Chen has always had at least two individual meetings with me per month since the beginning of my Ph.D. career, giving me insights into my research. Chen provided me with not only research topic guidance, but more importantly, helped me to develop the analytical skills, sharpen the logical thinking, and enhance the academic writing abilities. His enthusiasm and creativity to solve critical problems in science and engineering inspired me to strive for a consistent and ever-improving goal in my doctoral work, a underresearched but valuable area, and eventually built a relatively complete research architecture. I am also grateful to Chen for the meaning he has to me personally. A family problem hindered my academic study for a considerable period of time. The support Chen provided for my personal issue was of the greatest importance to me. I also appreciate two outstanding members of my thesis advisory committee, Prof.Chen I-ming, and Prof. Guan Cuntai, for their critical advice and guidance to the completion of my research works and this thesis.My research was greatly helped by my extraordinary collaborators and close friends.Mr. Huang Zhiyu, who began his doctoral degree at the same time as me, collaborated with me in most of my research projects and gave me insightful and valuable advice throughout the studies. My scientific growth have also been helped by Yang Haohan, Zhou Yanxin, and Huang Wenhui. I am also thankful to my team mem-

show abstract

Section: End-to-end Navigation Of Ugvsmentioning

confidence: 99%

Human-guided reinforcement learning: methodology and application to autonomous driving

View full text Add to dashboard Cite

show abstract

Learning a Safe Driving Policy for Urban Area from Real-World Data

Mahmud,

Huang

2024

2024 6th International Conference on Reconfigurable Mechanisms and Robots (ReMAR)

View full text Add to dashboard Cite

A Review of Research on Longitudinal Control of Intelligent Vehicles Based on Drive/Brake by Wire

Shi,

Qian,

Chadia

et al. 2024

WEVJ

View full text Add to dashboard Cite

In recent years, with the rapid innovation of science and technology, wire control technology, as a key technology, has achieved the transmission control of vehicles through the form of “electrical signals”, which has become an important foundation for realizing the high degree of intelligence of vehicles. This paper provides a comprehensive overview of the wire control technology, its application and longitudinal control strategy, and focuses on the longitudinal control technology of intelligent vehicles based on drive/brake by wire. The specific content includes five parts: first, the principles and characteristics of wire control technology and its application in intelligent vehicles are introduced; then, two commonly used longitudinal control strategies are described; then, the application of classical control technologies (such as PID, MPC, and sliding-mode control) in the longitudinal control of intelligent vehicles is discussed, including their working principles, characteristics and related research; subsequently, the AI control technology (deep reinforcement learning) is presented in the longitudinal control of intelligent vehicles, discussing its theoretical basis, the current status of algorithm research, control methods, and practical applications, etc.; finally, the paper summarizes the advantages and disadvantages of the classical control technology and AI control technology, and looks forward to the application and development prospects of these two control technologies in the control of intelligent vehicles.

show abstract

An End-to-End Deep Reinforcement Learning Model Based on Proximal Policy Optimization Algorithm for Autonomous Driving of Off-Road Vehicle

Cited by 3 publications

References 40 publications

Human-guided reinforcement learning: methodology and application to autonomous driving

Human-guided reinforcement learning: methodology and application to autonomous driving

Learning a Safe Driving Policy for Urban Area from Real-World Data

A Review of Research on Longitudinal Control of Intelligent Vehicles Based on Drive/Brake by Wire

Contact Info

Product

Resources

About