Method for predicting quality of medium and heavy plates based on evolution fuzzy correlation rule

Abstract

The invention provides a method for predicting the quality of medium and heavy plates based on an evolution fuzzy correlation rule. A fitting function can be used for replacing a consequent of the fuzzy correlation rule, thereby enabling the fuzzy correlation rule to achieve the regression prediction function on the basis that a classification function is achieved. The method comprises the steps:obtaining the fuzzy correlation rule corresponding to steel rolling data, wherein the steel rolling data is the steel rolling data of the medium and heavy plates; carrying out the learning of the consequent of the fuzzy correlation rule, and constructing a fitting function on the basis of the consequent of the fuzzy correlation rule; obtaining the testing data, judging whether the testing data ismatched with the fuzzy correlation rule or not, carrying out the prediction through a fitting function generated according to the matching rule if the testing data is matched with the fuzzy correlation rule, or else carrying out the weighing through the fitting functions of all fuzzy correlation rules at a current moment, and obtaining a prediction result. The invention relates to the technical field of data mining.

Description

technical field [0001] The invention relates to the technical field of data mining, in particular to a method for predicting the quality of medium and thick plates based on evolutionary fuzzy association rules. Background technique [0002] Steel plates with a thickness of 4.5mm to 25mm are medium and heavy plates, and the quality prediction of medium and heavy plates is of great significance in actual production. However, its performance is affected by many factors such as production material composition and production process, and it is difficult to accurately describe it with mathematical formulas. Although the artificial intelligence method represented by artificial neural network has strong nonlinear approximation ability, the neural network learning algorithm is prone to over-fitting phenomenon in the learning process. [0003] With the development of computer and database technology, a large amount of data in the production process of the steel rolling industry has b...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a method for predicting the quality of medium and thick plates based on evolutionary fuzzy association rules, so as to solve the problem in the prior art that the fuzzy association rule mining algorithm can only be used for classification and cannot be used for prediction

Images