摘要
首次利用人工神经网络技术对影响拉深过程中法兰下摩擦系数的工艺参数及润滑油参数进行了分析 ,提出了润滑油选择方案并描述了神经网络建模过程。神经网络预测计算结果与实际符合较好。对自行设计的试验装置进行了简要描述并提出了试验数据误差修正公式 ,实践证明 ,该公式有效的减少了试验误差。
Taking into consideration both the principal technological parameters and several parameters of lubricant,a model of artificial neural network is constructed to predict the average coefficient of friction under flange during drawing.The selections are made to choose proper lubricant under certain operating condition according to the predicted results.The predicted results show good agreement with experimental data.Then the experimental setup are described concisely which are designed by ourself and a formulation to correct experimental errors is introduced which works to give more accurate data.
出处
《锻压技术》
CAS
CSCD
北大核心
2000年第2期47-50,共4页
Forging & Stamping Technology
关键词
人工神经网络
摩擦系统
拉深
润滑油
误差修正
Artificial neural network Coefficient of friction Drawing Lubricant Error correction
