Blast furnace fault diagnosis rule exporting method based on deep neural network

Abstract

The invention discloses a blast furnace fault diagnosis rule exporting method based on a deep neural network, and belongs to the field of industrial process monitoring, modeling and simulation. The method comprises: firstly, modeling historical blast furnace fault data by adopting a deep neural network; then, for each fault, starting from an output layer of the network, establishing sub-models ofnodes of adjacent layers of the deep neural network by utilizing the decision tree in sequence, and exporting if-then rules; and finally, merging if-then rules layer by layer to finally obtain a blastfurnace fault diagnosis rule taking the blast furnace process variable as a rule antecedent and taking the fault category as a rule consequent. According to the method of the invention, the advantageof high diagnosis precision of the deep neural network is utilized, fault diagnosis knowledge is obtained from blast furnace historical data, the knowledge is converted into rules which are easily understood by blast furnace operators, man-machine collaborative knowledge and decision fusion is achieved, and the method can be widely applied to industrial systems with high reliability and accuracyrequirements for fault diagnosis.

Description

technical field [0001] The invention belongs to the field of industrial process monitoring, modeling and simulation, in particular to a method for deriving fault diagnosis rules for blast furnaces based on a deep neural network. Background technique [0002] In the iron and steel manufacturing process, the large-scale iron-making system dominated by large-scale blast furnaces is the key process of ferrite flow conversion. Large blast furnaces are the core equipment of the steel manufacturing process and are the largest chemical reaction vessels in the world. During the operation of large blast furnaces, if abnormal furnace conditions cannot be monitored, diagnosed and controlled in time, it will not only cause heavy losses of resources and equipment and reduce the age of blast furnaces, but also lead to accidents resulting in casualties and property losses. Therefore, ensuring the safe operation of blast furnaces is the top priority in the steel manufacturing process. [0...

AI Technical Summary

Problems solved by technology

This method is expensive to establish a rule base, and as the age of the blast furnace increases or the furnace conditions change drastically, the expert system will fail and lack the ability to enhance evolution

Based on machine learning methods, traditional white-box models such as decision trees and SVMs require a lot of training samples, but the number of fault samples in practice is often very small, so they cannot achieve good results; the current rapidly developing deep neural network Although the method has made great breakthroughs in accuracy, it is not trusted by blast furnace operators because it is a black-box model, and its credibility and diagnostic stability have been questioned, making it difficult to promote and use in practice; based on multivariate statistics The method has a very high misjudgment rate for large-scale blast furnaces with fluctuating raw material quality and complex and changeable operating environments

Therefore, there is still a big gap between the existing abnormal furnace condition diagnosis methods and practical application, and new paths and new methods need to be explored

Images