engleski

Neural Network in the Autonomous Mobile Robot Path Tracking Control

This paper presents implementation of a non-linear PD-controller for autonomous mobile robot path tracking control by a neural network (NN) in a real-time working environment instead of using classical form of the same type of controller. Controller-guided torque (turning moment) on two independent robot wheels insure movement of robot in unorganized surroundings. Two types of static-feedforward one learning step neural networks (Time-Discrete Static-Feedforward Neural Network and Radial Basic Function Neural Network with modified learning algorithm) were used and test results were presented and compared.

Autonomous mobile robot; Path tracking control; Time-Discrete Static-Feedforward Neural Network; Radial Basis Function; Modified learning method

nije evidentirano