论文部分内容阅读
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。
For the first time, artificial neural network technology was used to analyze the process parameters and lubricating oil parameters that affect the friction coefficient of the flange during drawing. The selection scheme of lubricating oil and the neural network modeling process were also described. Neural network prediction results in line with the actual good. The self-designed test device is briefly described and the error correction formula of the test data is proposed. Practice has proved that the formula effectively reduces the test error.