中文摘要：

针对运输环境中随机振动引起的产品破损问题,以鸡蛋为例建立了具有自学习功能的基于BP神经网络的产品破损评估模型.重点分析了随机振动下产品的破损机理,研究了随机振动信号频率谱分析方法,利用疲劳累积损伤理论给出了运输环境随机振动下产品破损概率分析理论,为模型的建立提供了充分的理论依据.具体分析了鸡蛋在运输过程中的特点,确定了模型的结构及输入层、隐含层、输出层神经元个数及意义.通过模拟运输环境对200箱鸡蛋进行了振动试验统计的样本数据,对模型进行了训练和测试.结果表明该模型具有较高的评估精度和较好的泛化能力,为研究多因素下产品破损评估模型提供了一定的基础.

英文摘要：

According to the problem that product damage caused by random vibration in transportation, a model based on self-learning neural network was established to assess the state of the products with the eggs as an illustration. Sufficient theoretical basis for the model establishment has been provided by focusing on the analysis of damage mechanism of products under random vibration, studying the method to random vibration signal frequency spectrum and giving the result of product damage probability in transportation with fatigue cumulative damage theory. Specifically, the characteristics of the eggs were analyzed in the process of transportation and the model structure, the number and significance of neurons of input layer, hidden layer and output layer were determined. According to the sample data for 200 cases of eggs from the vibration testing by simulating conditions of transport, the model was trained and tested. The results show that the model is featured with high precision and good generalization, hence providing a certain basis for the research on the product damage assessment model with multiple factors