Integrated design and YOLO based control framework for autonomous EV charging robot platforms

摘要

• The paper presents the development and engineering of a robotic arm mechanism specifically tailored for electric vehicle (EV) charging, ensuring compatibility with various EV models regardless of their manufacturer. • The robotic arm incorporates advanced autonomous capabilities, including precise vehicle detection, plug-in procedures, and real-time battery monitoring, enhancing the efficiency and safety of the EV charging process. • The research integrates a YOLOv8-based charging port detection system, enabling accurate and efficient identification and interaction with EV charging ports. • The study offers a thorough analysis of the robotic arm's performance, demonstrating its effectiveness through experimental results and comparative analysis with conventional charging solutions. The increase in demand for convenient and efficient charging solutions has experienced a significant upsurge due to the rapid adoption of electric vehicles (EVs). Consequently, the primary objective of this manuscript is to offer a comprehensive analysis of the development, execution, and assessment of an independent mechanical appendage specifically tailored for charging EVs. The demand for efficient and practical charging solutions is increasing as the demand for electric vehicle (EV) adoption increases. This investigation centers on the conception and engineering of a robotic arm mechanism that has been individually tailored to integrate with EVs. This robotic arm's design aimed to have autonomous capabilities, which include exact automobile identification, plug-in procedures, and real-time battery monitoring, ensure the charging process's security and efficacy. This study aims to give a complete examination of the modelling and simulation development of arm and evaluation of YOLOv8 based charging port detection exclusively for EV charging applications. The proposed model is capable of classifying charging ports precisely. The CCS1 metrics are: 100% recall, 0.962 mAP50; Type 1 – 100 % recall, 0.993 mAP50; CHAdeMO – 82.6% recall, mAP50; GB-T – 92.3% recall, mAP50; Tesla – 100% recall, mAP50; Type 2 ports – 94.1% recall, mAP50. The research comprises the design of and engineering of a robotic arm system that is optimized to interface with EVs intuitively.