基于神经网络包装缓冲特性预测模型的研究

A Prediction Model for Package Impact Cushioning Characteristic Based on Neural Network

文中描述了冲击缓冲特性的预测模型,研究了不同的试验方案,其预测模型结果与实际值最大相对误差的变化规律,找出了能满足工程设计要求的最少试验方案。通过实例验证了该预测模型,其预测模型的结果对工程设计是十分便利。

A prediction model for impact cushioning characteristic of package was introduced. The changing law of the maximum relative error of the predicted results to actual value was set up.The least test times scheme satisfing engineering design was found.The prediction model was demonstrated by examples and the prediction results are very convenient for engineering design.