论文部分内容阅读
用人工神经网络模型分析了时效参数对铝青铜硬度的影响。用“舍一法”训练了模型。模型对训练样本的计算值与实测值在散点图中沿着45°角平分线分布,统计学指标为:均方误差(MSE)为2.1388,相对均方误差(MSRE)为6.59%,拟合分值(VOF)为1.8301。用训练后的网络模型进行预测,得到的散点大致分布于45°角平分线附近,统计学指标为:均方误差为1.9512;相对均方误差为5.62%;拟合分值为1.7783。对时效参数的影响分析表明:时效温度和时效时间对硬度的影响,都存在一个最佳值,在时效温度和时效时间分别为450℃和30 min时,铝青铜的硬度达到最大值。
The effect of aging parameters on the hardness of aluminum bronze was analyzed by artificial neural network model. Trained the model with The calculated and measured values of the model for the training samples are distributed along the scattergram at a 45 ° angle. The statistical indicators are as follows: the mean square error (MSE) is 2.1388, the relative mean square error (MSRE) is 6.59% The combined score (VOF) is 1.8301. After training, the network model is used to predict. The scatter points are distributed around the bisector of 45 °. The statistical indexes are as follows: the mean square error is 1.9512; the relative mean square error is 5.62%; the fitting score is 1.7783. The analysis of the aging parameters shows that there is an optimum value of the aging temperature and the aging time, and the hardness of the aluminum bronze reaches the maximum when the aging temperature and the aging time are 450 ℃ and 30 min respectively.