Equipment intelligent early warning method based on multiple-input-multiple-output ResNet

Abstract

The invention discloses an intelligent equipment early warning method based on a multi-input multi-output ResNet network. A deep residual network ResNet is used to process the data prediction regression problem. Through the reasonable construction of the training data set and the ResNet network, one network can perform prediction regression on multiple features at the same time. The number of training parameters is reduced, and the training speed and efficiency are improved. The method comprises the following steps: selecting equipment-related historical data, and preprocessing the equipment-related historical data; constructing a multi-input and multi-output ResNet network according to the multi-input and multi-output ResNet; training the multi-input multi-output ResNet network by using the preprocessed equipment related historical data to obtain an intelligent early warning model, and analyzing a training result to obtain a residual threshold; collecting relevant real-time data of the equipment, calculating a predicted value by utilizing the intelligent early warning model, and judging the running state of the equipment according to a residual error threshold value. The method isused for intelligent early warning of equipment.

Description

technical field [0001] The invention relates to the field of early warning of equipment failures, in particular to a method for predicting equipment states based on a multi-feature prediction deep residual network ResNet. Background technique [0002] In traditional industrial fields such as power generation and chemical industry, operators usually judge the operating status of equipment by monitoring the signals of various sensors connected to the equipment in the control system, but the monitoring of a single sensor signal cannot detect the abnormal state of the equipment in time. When the equipment is in a degraded state, the relevant sensor signal does not reach the alarm threshold set by the control system. If the operator does not find out, the equipment will run in a degraded state for a long time until the alarm set by the control system for the sensor signal occurs. At this time, the equipment is usually already A major failure occurred. [0003] The operating stat...

AI Technical Summary

Problems solved by technology

The application of deep neural network is mainly led by Internet companies. These companies have designed a large number of application scenarios for the application needs of users' daily life, covering all aspects of users' lives, but the application in the industrial field is still at a relatively low level.

On the other hand, limited by application scenarios, deep neural networks are currently mainly used to deal with single-feature classification problems, and there are few solutions for multi-feature regression prediction problems.

[0006] In the traditional convolutional neural network application, in theory, as long as the number of layers of the network is continuously increased, better results can be obtained, but in experiments, it is found that with the increase of the number of network layers, the phenomenon of network degradation occurs: with As the number of network layers increases, the training set loss gradually decreases, and then tends to be saturated. When the network depth is increased, the training set loss will increase instead. This is not an over-fitting phenomenon, because the training loss in over-fitting is keeps decreasing

Images