Motion Controlling of a Four Wheels Vehicle Basing on an Adaptive Neuro Fuzzy Inference System (Anfis)

Abstract

A robot motion control is a very complicated task that many strategies were presented to deal with , the (AI) Artificial Intelligence represents one of the most appropriate methods that solve the robot path planning , navigation  and colliding avoidance.

 In this paper , a hybrid soft computing technique has been proposed to control a four wheel vehicle system , this technique is known as (ANFIS) Adaptive Neuro Fuzzy Inference System , which combines between two of the AI methods which are the Artificial Neural Network (ANN) and the Fuzzy Inference System (FIS) to get the integrated system properties . ANFIS method is used due to its ability to learn the human experience in order to develop various robot motion strategies .In this proposed method , the distance between the vehicle and each of the nearest barrier and the target represent the inputs to the ANFIS processing structure , while the vehicle speed , approach angle , and the distance between the current and updated positions of the vehicle , represent the outputs that will show better robot performance in term of accuracy .Besides , this suggested low cost system accomplishes its mission of reaching target using simplest paths and avoiding colliding with barriers, in addition to overcoming the complex computations and providing very acceptable tolerances for the uncertainties that exists in the work space data .