The Use of Deep Learning Methods in Diagnosing Rotating Machines Operating in Variable Conditions

This paper presents the use of artificial neural networks in diagnosing technical condition drive systems operating under variable conditions. The effects temperature and load variations on values diagnostic parameters were considered. An experiment was conducted a testing rig where introduced corresponding to main gearbox bucket wheel excavator. signals vibration acceleration body, rotational speed, current consumption motor for different oil measured. Synchronous analysis performed, order amplitudes current, determined. Such datasets learning vectors set deep networks. A new approach proposed this is train network using consisting only data from efficient system. responses trained undamaged system performed against response recorded three damage states: misalignment, unbalance, simultaneous misalignment unbalance. As result, parameter as normalized measure deviation results developed faulted result condition.