Optimal Genetic Fuzzy Obstacle Avoidance Controller of Autonomous Mobile Robot Based on Ultrasonic Sensors
Qiao Liu,Yong-Gang Lu,Cun-Xi Xie
DOI: https://doi.org/10.1109/robio.2006.340327
2006-01-01
Abstract:In order to avoid obstacles efficiently and reach the goal quickly under multi-obstacle environment, we studied the path planning question of autonomous mobile robot (AMR) based on ultrasonic sensor information by combining genetic algorithm with fuzzy logic control. Firstly, the principles and configuration of ultrasonic sensors were introduced. Secondly, the dynamic model and kinetic equations of AMR were constructed. Then, according to the number of obstacles, the avoiding behavior and rules were presented, moreover, the obstacle-selecting and avoidance rules and flow chart of AMR under multi-obstacles environment were also proposed. Based on above, we designed a fuzzy controller to modify the moving direction of AMR by defining or establishing input variables, output variables, fuzzy membership functions, fuzzy rule base including 25 If-Then fuzzy inference rules and defuzzification method. At last, a genetic algorithm was added for optimal searching parameters which includes the 5 times 5 consequent variables of the control rule table, the searching parameters, the bottom parameters of triangular membership functions and scaling factors. By setting the total route length as the target function, we founded the optimal genetic fuzzy controller for various obstructive environments through Matlab 6.5 simulation. The simulation results show the optimal controller under obstructive environment has better adaptability and passes shorter route in complex environment.