Abstract. The article is analytical. It considers the construction principles of the automation system structure which realize the concept of «unmanned mine». All of these principles intend to deal with problems caused by a continuous complication of mining-and-geological conditions at coalmine such as the labor safety and health protection, the weak integration of different mining automation subsystems and the deficiency of optimal balance between a quantity of resource and energy consumed by mining machines and their throughput. The authors describe the main problems and neck stage of mining machines autonomation and automation subsystem. The article makes a general survey of the applied «unmanned technology» in the field of mining such as the remotely operated autonomous complexes, the underground positioning systems of mining machines using infrared radiation in mine workings etc. The concept of «unmanned mine» is considered with an example of the robotic road heading machine. In the final, the authors analyze the techniques and methods that could solve the task of underground mining without human labor.

It is proposed to conduct fault diagnostic test on electric drives of mining shovels based on the results of monitoring the current values of electromagnetic and mechanical parameters and variables of electric drives obtained in the course of their operation using the modern computer technology. The structure of the developed system of functional diagnostics, allowing to monitor the status of the drive and identify emerging fault is shown in the paper. To determine in real time the current parameters and variables of DC motor which can't be measured during their operation, the dynamic identification was used based on the measured current and voltage of the motor windings, and mathematical estimation methods. Parameters of the mechanical subsystem of electric drive are identified by a mobile measuring system. The authors also give the structure and characteristics of the one-step neural network predictor of current, used to predict the current values in the armature and field windings of motor. The analysis of the technical state of the electric drive by a set of attributes is performed in a special analyzer, built on the basis of pre-trained artificial neural network. The results of these studies support the possibility of creating a diagnostic system for the main electric drives of mining shovels using the estimation methods and apparatus of artificial neural networks.