Fuzzy Embedded Mobile Robot Systems Design through the Evolutionary PSO Learning Algorithm

The evolutionary learning algorithm called particle swarm optimization (PSO) is developed in this paper. The image model of the embedded mobile robot is automatically generated with the omni-directional image concept to approach toward the behavior of the embedded mobile robot. The circumvolutory environment is dynamically captured from the head of the mobile robot, which will directly be transformed into the Cartesian coordinate system. The required parameters of fuzzy rules are automatically extracted with the guide of the flexible fitness function, which is efficiently approach toward the multiple objectives of avoiding obstacles, selecting favorable fuzzy rules to drive the desired targets at the same time. Three illustrated examples with various initial positions for the discussed environment map containing different blocks size and locations are illustrated the efficiency of the PSO leaning algorithm. Simulations demonstrate that the proposed mobile robot with the selected fuzzy rules can avoid the obstacles and achieve the targets as soon as possible. Key-Words: Particle swarm optimization; Fuzzy systems; Mobile robots, Evolutionary learning, Omnidirectional image.