Spinning whole-process energy consumption monitoring method based on feature self-matching transfer learning

Abstract

The invention relates to a spinning whole-process energy consumption monitoring method based on feature self-matching transfer learning, and the method comprises: learning knowledge rules from an oldfactory historical data set through a deep convolutional neural network, and migrating the knowledge rules to the abnormal trend recognition of the spinning energy efficiency of a new factory, so as to solve a problem that an abnormal sample of the new factory is not sufficient. Meanwhile, for the problem of negative migration caused by mismatching of data features between a source domain and a target domain in the migration process, a feature self-matching layer network based on clustering is designed, the distance of similar features is minimized through a feature matching matrix, outlier features are eliminated, positive migration of effective knowledge is promoted, and negative migration of invalid interference knowledge is inhibited. Compared with an existing method, the method of theinvention is advantageous in that the model provided by the invention has higher spinning energy efficiency anomaly detection precision and lower missing report rate.

Description

technical field [0001] The invention relates to a method for monitoring energy consumption in the whole process of spinning, in particular to a method for monitoring energy consumption in the whole process of spinning based on feature self-matching transfer learning, and belongs to the technical field of production energy consumption monitoring. Background technique [0002] Spinning is a typical energy-intensive industry for people's livelihood. In 2018, China's spinning power consumption reached about 70 billion kWh, but the effective utilization rate was less than 75%, and a large amount of energy consumption was wasted in the abnormal discovery of lagging behind. In the yarn production process, due to production events or changes in the production environment, the power consumption per ton of yarn often deviates from the normal value, and the lagging discovery leads to a lot of energy waste. Energy efficiency monitoring in the production process is one of the most effect...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to realize the early monitoring of energy consumption in the whole process of spinning, transfer the historical energy consumption data of the old factory to the energy consumption monitoring model of the new factory through the transfer learning method, improve the detection accuracy of energy consumption symptoms, and solve the problem of new factories Energy consumption monitoring cold start problem

Images