Artificial Neural Network in the Control Loop of a Wheeled Robot

Abstract

This paper describes the creation of a tracking system represented by a wheeled robot following a target (for example, a human). An artificial neural network (ANN) is used in the control loop of the robot to determine the range and bearing of the target, which are the tracking parameters. The main focus of the work is to develop a computer model of ANN capable of calculating near-optimal estimates of tracking parameters. In this case, the root mean square error (RMSE) of tracking parameter estimates is used as the criterion of efficiency of ANN operation. As a result, for the best ANN model the RMSE for the range was 0.009 m, and for the bearing was 1.006 degrees.