Vision-based Automated Guided Vehicle Control by Using Fuzzy Kohonen Clustering Network

10.6567/IFToMM.14TH.WC.OS13.117

Meng-Ju Han；Jie-Ming Wang

Vision-based guiding ； Fuzzy Kohonen clustering network ； Automated guided vehicle

Proceedings of the 14th IFToMM World Congress

14th-4（2015 / 11 / 04）

337 - 341

英文

Automated guided vehicle (AGV) is widely used in the industrial field. Most existing methods treat the AGV controller by simple and intuitive representations such like PID controller. However, the PID controller lacks a standard procedure to generate proper parameters and let suitable parameters adjusting become a heuristic work. This paper presents a vision-based control method for mobile platform. Three modules are integrated to construct the control system, namely, color path detection module, path feature extraction module and control behavior decision maker. Based on Fuzzy Kohonen clustering network, a fuzzy concept control system is proposed to build a standard procedure. All parameters of the controller are self-generated without relying on human's experience. Experimental results show that the proposed visionbased guiding scheme appropriately responds to varies color path in a continuous manner.