Prediction and Optimization of Key Performance Indicators in the Production of Stator Core Using a GA-NN Approach

4th International Conference on Mechanical, Materials and Manufacturing

材料科学;机械制造

Rajora, M.^1 ; Zou, P.^2 ; Xu, W.^3 ; Jin, L.^3 ; Chen, W.^3 ; Liang, S.Y.^1

George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta

GA, United States^1

Mechanical Engineering College, Donghua University, Songjiang District, Shanghai, China^2

Shanghai Electric Group Co., Ltd., Central Academe, Shanghai, China^3

关键词: Decision making process;    Engineering problems;    Intelligent techniques;    Key performance indicators;    Large amounts of data;    Levenberg-Marquardt;    Neural network (nn);    Optimization algorithms;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/272/1/012011/pdf
DOI  :  10.1088/1757-899X/272/1/012011

学科分类：材料科学（综合）

来源: IOP

【 摘 要 】

With the rapidly changing demands of the manufacturing market, intelligent techniques are being used to solve engineering problems due to their ability to handle nonlinear complex problems. For example, in the conventional production of stator cores, it is relied upon experienced engineers to make an initial plan on the number of compensation sheets to be added to achieve uniform pressure distribution throughout the laminations. Additionally, these engineers must use their experience to revise the initial plans based upon the measurements made during the production of stator core. However, this method yields inconsistent results as humans are incapable of storing and analysing large amounts of data. In this article, first, a Neural Network (NN), trained using a hybrid Levenberg-Marquardt (LM) - Genetic Algorithm (GA), is developed to assist the engineers with the decision-making process. Next, the trained NN is used as a fitness function in an optimization algorithm to find the optimal values of the initial compensation sheet plan with the aim of minimizing the required revisions during the production of the stator core.