当前位置：首页 > 过刊浏览->2020年41卷第10期
基于人工神经网络模型的铸造涂料性能预测
Performance Prediction of Foundry Coatings Based on Artificial Neural Network Model
浏览(1799) 下载(6)
- DOI:
- 作者:
- 王佳宇 赵溶 李琪 李日
WANG Jiayu;ZHAO Rong;LI Qi;LI Ri
- 作者单位:
- 河北工业大学材料科学与工程学院 宁波特种涂料厂
School of Materials Science and Engineering,Hebei University of Technology,Tianjin 300401,China;Ningbo Special Coatings Factory,Ningbo 315000,China
- 关键词:
- 铸造涂料 BP神经网络 ELM神经网络
casting coating BP neural network ELM neural network
- 摘要:
- 针对铸造涂料的成分与涂料基本性能之间的复杂关系,提出利用人工神经网络对实验数据进行处理,建立实验因素与结果之间的神经网络模型。结果表明,在误差允许的范围内,构建的B P神经网络和ELM神经网络都可以实现对铸造涂料性能的预测。程序运行时间主要受神经网络模型的影响,对于悬浮率(2h)和悬浮率(24 h),ELM神经网络模型比BP神经网络模型具有较高的预测精度和运行速度。In view of the complex relationship between the composition and the basic properties of casting coatings,an artificial neural network was proposed to process the experimental data and establish a neural network model between experimental factors and results.The results show that both BP neural network and ELM neural network can predict the performance of casting coatings within the allowable range of error.Program running time is mainly affected by the neural network model.For suspension rate(2 h)and suspension rate(24 h),the ELM neural network model has higher prediction accuracy and running speed than the BP neural network model.