Title screen

[References: 17 tit.]

One of the key tasks of mobile robotics is navigation, which for Outdoor-type robots is exacerbated by the functioning in an environment with a priori of unknown characteristics of underlying surfaces. In this paper, for the first time, the learning navigation system for mobile robot based on the group method of data handling (GMDH) is presented. The paper presents the results of training of models both for evaluating the robot’s pose (coordinates and angular orientation) in heterogeneous environment and classification of the type of underlying surfaces. In addition to the direct readings of the on-board sensors, additional parameters (reflecting how the robot perceives the surface terramechanics) were introduced to train the models. The results of testing of the obtained models demonstrate their performance in an essentially heterogeneous environment, when areas of the underlying surfaces are comparable with the robot’s dimensions. This testifies the operability of developed GMDH-based learning system for mobile robot navigation.