Design of visual feedback tracking algorithm for nonholonomic mobile robots based on neural network

With the rapid development of the computer and the electronic technique, the application of robots has been widen. The robot visual serving control system may mimic the human eyes. Then the vision information is used as a feedback to improve the ability of the robot adaption to the environment. However, traditional algorithms which need the calibration of visual parameters spend much time and become technical bottlenecks. This paper presents the development background of the robot and the concept of nonholonomic mobile robots with visual servoing feedback. Second, the deficiency exists in traditional algorithms and fuzzy controller. Third, BP neural network PID is proposed to design controller. Combining BP neutral network with PID controller is used to manipulate mobile robots firstly. The complex deduces of common tracking controllers is simplified and tracking control problem with non calibrated virtual parameters is solved. Finally, we program the simulation code. The simulation results show that the method is effective. © 2013 IFSA.