Once upon a time, there is a mobile robot named KePIX. A mobile robot with a fuzzy inference system based navigation control system which is given control behavior for obstacle avoidance and goal seeking. Fuzzy inference system was written in C/C++ using software Borland C++ Builder v.4.0 Professional during january to july 2005.
The fuzzy inference system for obstacle avoidance using 3 input, that are object distance value detected at askew-left (-45 degree), front side (0 degree), and askew-right (45 degree) position; 22 fuzzy rules, and 2 output, that are steering angle value and step velocity of the wheel.
The fuzzy inference system for goal seeking using 2 input, that are distance between the center point of the robot and the target point, and orientation angle of the robot toward the target point; 14 fuzzy rules, 1 output steering angle.
The navigation control system was tested successfully in simulation for the combination task of obstacle avoidance and goal seeking navigation.
In the real platform control, there was an error in calculation of the actual position of the robot caused by hardware limitation to perform control action that is each robot's wheel didn't move at the same velocity. Motion direction of the robot have tendency to turn to the left. Slip factor 3 degree on each wheel of the robot, and fluctuation power on the batteries condition caused limitation performance of the robot.
The fuzzy inference system for obstacle avoidance using 3 input, that are object distance value detected at askew-left (-45 degree), front side (0 degree), and askew-right (45 degree) position; 22 fuzzy rules, and 2 output, that are steering angle value and step velocity of the wheel.
The fuzzy inference system for goal seeking using 2 input, that are distance between the center point of the robot and the target point, and orientation angle of the robot toward the target point; 14 fuzzy rules, 1 output steering angle.
The navigation control system was tested successfully in simulation for the combination task of obstacle avoidance and goal seeking navigation.
In the real platform control, there was an error in calculation of the actual position of the robot caused by hardware limitation to perform control action that is each robot's wheel didn't move at the same velocity. Motion direction of the robot have tendency to turn to the left. Slip factor 3 degree on each wheel of the robot, and fluctuation power on the batteries condition caused limitation performance of the robot.
No comments:
Post a Comment