基于神经网络固体渗硼的性能预报

Property Prediction of Solid Boriding Based on Artificial Neural Network

将人工神经网络理论和算法应用于固体渗硼的研究,在对人工神经网络训练的基础上,建立了固体渗硼工艺与渗层性能预报的数学模型。试验结果验证了性能预报神经网络模型的可靠性,为解决固体渗硼性能预报问题提供了一条先进、合理的途径。

The theory and the algorithm of the artificial neural network were applied in the research of the solid boriding, On the basis of the training of network, the mathematics model of the relationship between the solid boriding and prediction of properties was built. The experimental results verify the reliability of property prediction of artificial neural network model. It presents an advanced and effective approach to the property prediction of solid boriding.