应用神经网络预测橡胶配方的性能

Forecasting Rubber Mix Proportion Quality with Artificial Neural Network

建立了以配合前用量和胶料物理性能的人工神经网络模型， 训练后的神经网络能较好的拟合实验数据， 对９ 个橡胶配方的性能进行回想和预测结果与试验值的偏差小， 说明利用神经网络方法精度高，

An artificial neural network model about mix rubber physical quality is established. After being trained, the neural network, the model can fit experimental data. The results of recalling and forecasting are accurate. It is shown that the model is of high precision and reliability.