Information system of agricultural robotic KAMAZ cars

Abstract

The article is devoted to the development of an information system for the complex of agricultural vehicles. It is shown that with the development of intelligent technologies, wireless communications, and the Internet, the task arises of creating a single information space for agricultural vehicles. The structure of sensors and measuring instruments installed on an unmanned KAMAZ car is described. The block diagram of the control system for the set of robotic cars is presented. The general structure of the information system, which includes a navigation system, technical vision, control units, a remote diagnostics subsystem, and a data transmission system, is considered. The technical diagnostics of the units is carried out using embedded and onboard thermal imagers. Thermograms from thermal imagers are analyzed using a convolutional neural network. Deep neural network and machine learning provided high accuracy in the detection and classification of equipment failures. The decision support system generates control signals for the units and transfers information to the highest level of control of the robotic complex.