基於GEP之視覺型循線自走搬運工

路徑跟隨對智慧型機器人是一項的重要任務。以往路徑跟隨大多採用磁軌或紅外線裝置。利用影像感測器像是攝影機作為輪型機器人的主要傳感器是一個不可避免的趨勢。本文將影像式路徑跟隨視為迴歸分析的問題，並採用基因表達程式規劃的技術來分析迴歸模型。該方法結合機器人的影像系統且在真實的環境中被驗證。實驗結果表明，該方法能夠產生精簡且準確的控制規則於路徑跟隨，並提升機器人的性能。 Path-tracking is an important task for mobile robotics. Traditional path-tracking methods are implemented with special sensors that detects tracks using a static model.Recently, visual sensors like CCD cameras have become essential for mobile robots. This article defines the control of visual-based path-tracking as a regression problem and uses the techniques of gene expression programming to discover such regression models. The techniques are integrated with a two-wheeled vision-based mobile robot. The experimental results show that the proposed method is able to produce simple and precise control rules for path-tracking and improves the robot’s performance.