人工神经网络技术在成形模具表面激光强化中的应用

Application of Artificial Neural Network in Laser Die-surface Hardening

基于人工神经网络基本理论 ,建立在成形模具表面激光强化中预测材料表面最高温度、强化效果的 BP网络模型。借助于 MATL AB仿真软件中的神经网络工具箱作为开发平台 ,将试验样本数据和经过试验验证的数值计算结果作为补充的样本数据用于 BP网络的训练 ,利用训练好的 BP网络对非线性的样本数据规律进行拟合 ,实现激光加工工艺参数的优化 ,为实际生产和加工提供有效的依据。

Based on the basic theory of artificial neural network, BP network model is founded to predict the highest surface temperature and hardening effect in the process of laser die-surface hardening. Experimental result and numerical simulation result verified by experiment are used to train BP network by means of MATLAB software. Sample data regularity of nonlinear is curve fitting through BP network so as to optimize technological parameters and provide for effective foundation in actual production.