Alloy mechanical property prediction method based on BP neural network for rollers

Abstract

The invention discloses an alloy-cast-steel mechanical property prediction method based on an improved BP neural network for rollers, and belongs to the technical field of alloy-cast-steel rollers. The method includes the steps of firstly, using allay cast steel which is different in alloy composition and hot-processing technological parameter for conducting a series of mechanical property tests on the rollers, and collecting and screening required training sample data of an artificial neural network model; constructing the BP artificial neural network model containing an input layer, a hiddenlayer and an output layer, and then forming mapping relationships among the alloy compositions, hot-processing technological parameters and mechanical property of the alloy-cast-steel rollers; adopting the artificial neural network after training to predict the mechanical properties of the alloy-cast-steel rollers. The constructed BP neural network model is high in prediction accuracy and stability, excellent in promotion capability, and capable of providing a new approach and method for further researching and development of novel alloy-cast-steel roller materials, so that the production cost is reduced, and the development time is shortened.

Description

technical field [0001] The invention relates to the technical field of alloy cast steel rolls, in particular to a method for predicting mechanical properties of roll alloys based on BP neural network. Background technique [0002] The roll is the main deformation tool of the rolling mill. During use, they will fail due to defects such as peeling, cracks, and fractures, thereby shortening the service life of the roll. The service life of the roll is mainly determined by the intrinsic properties, namely strength and hardness. , by adjusting its chemical composition and heat treatment process parameters, to a certain extent, it can meet the performance requirements of the roll. Therefore, it has become a new challenge for the roll industry to study the relationship between the composition, process and performance of the material, so as to find out the roll material with excellent performance that meets the needs of the steel rolling industry. [0003] Through material design, ...

AI Technical Summary

Problems solved by technology

At present, when studying the relationship between the composition, process and performance of roll materials, experimental analysis methods and formula theory methods are usually used. Repeated experiments will consume a lot of time and property. The formula method must establish a mathematical regression model in advance, but Since the material properties are affected by many factors, and there is a nonlinear mapping relationship between these factors, the regression empirical formula is complicated and numerous.

Images