Automatic steel plate surface defect recognition method based on case-based reasoning

Abstract

The invention provides an automatic steel plate surface defect recognition method based on case-based reasoning in order to solve the problems that a mechanism model used for recognizing steel plate surface defects is difficult to establish, which does not facilitate detection and control of product quality. The automatic steel plate surface defect recognition method automatically recognizes seven defects of cracks, scaling, scarring, humping and others. A case-based reasoning recognition model is established through the steps of case library establishing, weight allocation, target case obtaining, numerical value normalization, case retrieval, case reusing, case storing and others. A recognition algorithm is implemented, so that recognition accuracy is improved. The problem that due to subjectivity and uncertainty of allocating case-based reasoning characteristic variable weights, the recognition accuracy rate is low is solved. The experiment result shows that the method can accurately recognize the defects on the surface of a steel plate and is good in sensitivity and distinctness, the processing efficiency of the defects on the surface of a steel plate is improved, the rejection rate is decreased, and the method is of great significance to case analysis on the defects on the surface in the production process of the steel plate, production operation method improving and product quality improving.

Description

technical field [0001] The invention utilizes the case reasoning technology in the field of artificial intelligence to realize the type identification of steel plate surface defects. Different types of surface defects determine the grade division of steel plates, and the identification and classification of defects can provide positive guidance for cause analysis and improvement of operation methods. Steel plate is one of the important products produced in the metallurgical industry. The identification and classification of its surface defects is an important link in the production process, which belongs to both the field of artificial intelligence and the field of metallurgical production. Background technique [0002] With the growth of the national economy and the enhancement of product quality awareness, the automation technology of the metallurgical industry has ushered in unprecedented opportunities for development. The national medium- and long-term science and techno...

AI Technical Summary

Problems solved by technology

[0006] The present invention obtains an automatic identification method for surface defects of steel plates. The method adopts case reasoning technology, and establishes a case reasoning recognition model through establishing a case library, weight distribution, case retrieval, case reuse and case storage, and realizes a recognition algorithm; To solve the problem of low recognition accuracy caused by the subjectivity and uncertainty of the weight of characteristic variables in the current distribution case reasoning, the defect types are identified through the numerical changes of the characteristic variables; the processing efficiency of steel plate surface defects is improved, and the cause of defects is analyzed. , Improving production operation methods and improving product quality have positive guiding significance

Images