English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Exploring the relationship between composition, process and mechanical properties of cold rolling steel is very important for steel production. The most direct ways is to directly put into actual production. According to the results of actual production, we can tell the composition and process and to obtain the property. But this method is expensive for steel companies. In this paper, we proposed a prediction model for predicting the mechanical properties of cold rolling steel. And based on the mechanical properties prediction model and particle swarm optimization, a process design method is proposed. This model can predict the material composition and process of the steel given the steel mechanical properties. The two models both achieved good results finally. Introduction We will deduce the final mechanical properties according to the composition and process of the steel materials, which is called the prediction of the steel properties [1]. And we will calculate the proper composition and process according to the required mechanical properties, which is called process design. How to predict the mechanical properties of steel is always popular research work for steel enterprises. A good steel mechanical properties prediction model can reduce the cost of time and funds. However, no one has yet proposed a mature method to solve the problem of steel properties prediction until now. And the process design method we need is much few and more difficult. In the past few decades, a large number of scholars have devoted to the study of steel mechanical properties prediction models. At present, the main ideas are divided into two types, one is based on the metallurgical mechanism model [3-4], the other is the mechanical properties prediction model[5-10] through statistics. In the metallurgical mechanism based model, people are mainly concerned to find out the relationship between the constituent elements of steel and the ultimate mechanical properties. [3] Such models often start from the process to explore the effects of metallurgical processes and materials on the composition of the steel material produced, and then the mechanical properties are calculated by the composition of the steel material obtained[4] . But such methods often require deep physical and chemical knowledge and fully understanding of each step, so it is often difficult to build a model through this way. [2] The data accumulation during recent decades has provided a basis for the establishment of the prediction model of steel mechanical properties through mathematical statistics. In the last decade, a large number of scholars have invested in this research[5]. The main idea is to fit a nonlinear model with high accuracy based on a large number of data. [2, 4-10] In 2001, Perzyk proposed a mechanical properties prediction model based on ANN network, and achieved good effect [4]. After that, some scholars used support vector machines and Bayesian networks respectively to establish the mechanical properties prediction model [6, 7]. D. M. Jones compares multiple linear regression, multiple nonlinear regression and neural networks, within which neural networks has achieved the best effect. [5]. This kind of method is different from the traditional metallurgical mechanism model, and it does not need to grasp deep metallurgy knowledge to establish an excellent model. The previous research mainly focused on how to put forward a high accuracy prediction model, but the enterprise is more likely to know what the composition and process of these steel types may need, given the mechanical properties of the required steel. L. A. Dobrzanski proposed a prediction model based on genetic algorithm, which can predict the corresponding material composition and process

Application of BP Neural Network and PSO Methods for Process Design of Cold Rolling Steel