Autonomous Mobile Robot Design with Behaviour-based Control Architecture using Adaptive Neuro-Fuzzy Inference System (ANFIS)

Autonomous Mobile Robot (AMR) is a robotics-based platform that requires a control system to function correctly. Several robotics architectures have been proposed. Unfortunately, designing a control architecture is difficult and time-consuming, resulting in inaccurate control actions. As a result, the behaviour-based approach has emerged as the primary alternative to traditional robot control. This paper's contribution is a design for an AMR based on a behaviour-based control architecture. This paper discusses three AMR behaviors: planned goal-seeking behavior, wall-following behavior, obstacle-avoidance behavior, and color-detecting behavior. The Adaptive Neuro-Fuzzy Inference System (ANFIS) approach will be used to control all behaviour, which combines Fuzzy Logic and Artificial Neural Network to produce a reliable control algorithm. Competitive coordination is used to design behaviour coordination, particularly for obstacle-avoidance behaviour. Several experiments have been carried out. Based on the AMR's movement, it can be stated that all behaviours work well, where AMR can avoid obstacles, detect colours, follow walls, and decide the direction of motion.