METHOD FOR ESTIMATING THE TIME OF USEFUL USE OF EQUIPMENT BASED ON NEURAL NETWORKS

A method for predicting the useful time of equipment based on the processing of diagnostic data using parallel recurrent and convolutional neural networks is proposed. Images for the convolutional network are formed on the basis of the wavelet transform of diagnostic data. Neural networks operate in a multivalued classification mode, which is used in the method to refine the prediction of the useful time of equipment based on the recursive least squares method. The results of a model experiment performed using a program developed in the MatLAB environment that implements the proposed method are presented.