Wall-following and Navigation Control of Mobile Robot Using Reinforcement Learning Based on Dynamic Group Artificial Bee Colony

This study proposes an efficient wall-following and navigation control model that includes three control modes, namely w all-f ollowing (WF), t oward-g oal (TG), and b ehavior m anager (BM). To achieve an adaptive controller for WF mode, an efficientr ecurrent f uzzy c erebellar m odel a rticulation c ontroller (RFCMAC) based on d ynamic g roup a rtificial b ee c olony (DGABC) is proposed for implementing reinforcement learning process. The fitness function includes four assessment factors which are defined as follows: (1) maintaining safe distance between the mobile robot and the wall; (2) ensuring successfully running a cycle; (3) avoiding mobile robot collisions; (4) mobile robot running at a maximum speed. Moreover, the BM is used to switch WF mode and TG mode, and is employed as an escape mechanism based on the relationship between the robot and the environment. The experimental results show that the proposed DGABC is more effective than the traditional ABC in WF mode. The proposed control method also obtains a better navigation control than other methods in unknown environments.