The paper describes a diagnostic system for electromechanical equipment with a decision-making block based on a neural network. An asynchronous gear drive was used as a control object. Decision making was carried out on the basis of a comprehensive analysis of vibration data (from the gear drive) and the current consumption of the induction motor. Vibration velocity, vibration acceleration and current in the phases of the stator winding of the drive electric motor are distinguished as diagnostic signs. The work shows the possibility of increasing the efficiency of diagnostics of electromechanical equipment by using complex analysis with the use of an intelligent decision-making unit. Analysis of the results of the neural network operation with the received vibration and current data showed that with a smaller number of iterations (training time) (by 40 %), 97.9 % of correct answers and a lower error value (by 12 %) were obtained.
Science Journal 