摘要
通过对铸件缺陷与影响因素间因果关系的分析 ,利用MATLAB中的NeuralNetworkToolbox仿真环境和BP模型算法建立了用于铸件缺陷分析与控制的神经网络模型 ,详细论述了模型结构的设计、数据处理、网络初始化、训练与仿真的过程。实践表明 ,基于MATLAB的铸件缺陷分析与控制模型能有效地提高效率及直观地结果显示 ,对提高铸件质量及进一步研究具有积极作用。
Analysing the relationship between casting defects and influence factors, the neural network model for the analysis and control about casting defects is set up based on MATLAB neural network toolbox and BP model. The processing, namely the structure design, the data processing, the network initilization, the network training and simulation about the model, is introduced. The result of simulation and practical operation shows that the method can improve the efficiency and display the results with graphs and text of the analysis and control about casting defects. It play a positive role for improving casting quality and making thorough research.
出处
《铸造》
CAS
CSCD
北大核心
2002年第6期362-365,共4页
Foundry
基金
甘肃省自然科学基金资助项目 (ZS981 A2 2 0 2 4 G)