摘要
随着人们生活水平的提高,食品和医药安全逐渐成为社会关注的焦点。温度监控是保证物流运输中物品安全、减少经济损耗的关键。尤其是乳制品、血浆、疫苗等温度敏感性物品对运输环境中的温度要求更严格。当前冷藏车温度监控在智能控制方面缺乏较好的方法,无法达到对于温度敏感性物品的有效监测。通过BP神经网络对物品的温度变化进行预测可以达到很好的监控效果。针对BP神经网络中存在的收敛速度慢的问题,文中提出了一种自适应的学习速率的新方法,并将其应用于冷藏车温度预测中,通过Matlab仿真表明该算法具有很好的预测效果。
With the improvement of people＇s living standards,the safety of food and medicine is becoming the focus of attention. Temper-ature monitoring is the key factor to ensure the material safety and reduction of economic losses in the logistics transportation. Especially for dairy products,plasma,vaccines and other temperature-sensitive items,more stringent is required. Currently,the refrigerated trucks that lack of a better way in intelligent control can not be achieved for the effective temperature monitoring. Using BP neural network to predict the change in temperature of the items can achieve good control effect. In this paper,a novel method of BP learning algorithm to improve the convergence rate in BP neural network is proposed. The method is used to predict the cold chain temperature. Matlab simula-tion shows that the algorithm has a fast convergence rate theoretically.
出处
《计算机技术与发展》
2013年第10期180-183,共4页
Computer Technology and Development
基金
江苏省科技成果转化专项资金项目(BA2012024)
南京邮电大学项目(NY210012
NY211107)
关键词
神经网络
学习率
温度
neural network
learning rate
temperature
