基于BP神经网络的铝合金表面处理工艺参数优化  

Optimization of Surface Treatment Process Parameters of Aluminum Alloy Based on BP Neural Network

作　　者：雷艳惠[1] Lei Yanhui(Xianyang Vocational Technical College,Xianyang Shaanxi 712000)

机构地区：[1]咸阳职业技术学院

出　　处：《现代工业经济和信息化》2019年第12期21-23,共3页Modern Industrial Economy and Informationization

摘　　要：近些年来,铝合金表面处理过程中的质量控制和工艺优化方面都在讨论如何对生产过程中的几十项参数进行有效的优化,在本研究中主要利用BP神经网络来优化合金表面处理工艺参数,整个过程可以分为两个部分,首先是神经网络的建立和训练,主压包括神经网络模型、BP神经网络的训练和测试,其中神经网络模型还包含三点:网络输出和输出变量、网络结构的设计、激活函数的选择。其次是正交设计优化工艺参数,主要包括三个方面:正交设计、优化结果和分析、结果验证。通过以上几个方面的分析,希望能够为以后相关方面的研究工作提供一些参考。In recent years, quality control and process optimization in the surface treatment of aluminum alloys have been used to effectively optimize dozens of parameters in the production process. In this study, BP neural network is mainly used to optimize the process parameters of alloy surface treatment. The whole process can be divided into two parts, the first is the establishment and training of neural networks. The main pressure includes two aspects: neural network model, BP neural network training and testing, and the neural network model also contains three points:network Output and output variables, network structure design, activation function selection;followed by orthogonal design optimization process parameters, including three aspects: orthogonal design, optimization results and analysis,results verification. The analysis of the above aspects provides some reference for the research work in related aspects in the future.

关 键 词：BP神经网络 铝合金 表面处理 

分 类 号：TG177[金属学及工艺—金属表面处理]