基於主僕式架構之自動充電系統設計與開發

Abstract

本研究開發了一種基於主僕式架構的自動充電系統，可實現智慧自動充電服務。自動充電系統包含一自主移動式充電機器人、電源拖車、伺服器、使用者介面。自主移動式充電機器人做為主僕式架構中的主單元，其結合了機械手臂和自主移動平台，透過搭載的自主移動平台抵達目標位置後再利用視覺系統輔助機械手臂進行電源拖車上方的充電槍拿取與對位插孔等功能。電源拖車做為自動充電系統的僕單元，由供電平台和充電功能平台所組成，可實現符合開放點充電協議(Open Charge Point Protocol, OCCP)充電協議之充電功能。自主移動式充電機器人做為主單元具有高度智慧及自動化的特點，本論文基於機器人作業系統 (Robot Operating System, ROS) 實現同步定位、構建地圖、導航、路徑規劃和避障等功能。最後設計一使用者介面獲取用戶充電資訊與伺服器進行串接，並基於一基本規則庫進行充電排程，提高自動充電系統的自動化與智慧化程度。由實驗結果可知，本論文所提出之自動充電系統確實可以準確根據用戶於使用者介面所輸入之充電需求，進行排程、充電以最大化滿足用戶需求。This study developed an automatic charging system based on a master-slave architecture, enabling intelligent automatic charging services. The system consists of an autonomous mobile charging robot, a power trailer, a server, and a user interface. The autonomous mobile charging robot acts as the master unit in the master-slave architecture, integrating a robotic arm with an autonomous mobile platform. After reaching the target location via the mobile platform, the robot uses a camera vision system to assist the robotic arm in tasks such as picking up the charging gun and aligning it with the charging port on the power trailer. The power trailer, acting as the slave unit of the automatic charging system, comprises a power supply platform and a charging function platform. These two platforms cooperate with the OCCP charging protocol to achieve reliable charging capability.As the master unit, the autonomous mobile charging robot features a high level of intelligence and automation. This thesis, based on the Robot Operating System (ROS), implements functions including simultaneous localization and mapping (SLAM), navigation, path planning, and obstacle avoidance. Additionally, a user interface is designed to acquire user charging information and interface with the server. The charging scheduling is managed based on a basic rule-based system, enhancing the automation and intelligence of the charging system. Experimental results demonstrate that the proposed automatic charging system can accurately schedule and charge based on user input from the front-end interface, maximizing user satisfaction.