Neural network application for assessment of the asynchronous motors energy consumption

The article deals with the estimation of the asynchronous motors power consumption using the neural networks. As asynchronous motors with squirrel-cage rotor are the largest part of the industrial enterprises load, the change of their power consumption at the voltage deviations can affect the operating modes of the enterprise power supply system and other electrical receivers. Application of neural networks allows to determine the non-linear regularities in noisy heterogeneous data, provides valid calculations results in a large number of input parameters. At presence of input data for network training the created neural network helps simply enough to find the power consumption of an asynchronous motor at electric network voltage deviations from the nominal value in a wide range of voltages, load factor changes, and also if it is necessary of other data initial.