【作者】 王辉；

【导师】 赵艳；

【作者基本信息】 河北农业大学 ， 计算机应用技术， 2012， 硕士

【摘要】 食品质量安全是目前人们最为关心的问题。随着中国经济的增长,国民收入的增加,人民生活质量的不断提高,乳制品已成为百姓不可或缺的食品之一。其质量的好坏不仅影响生产企业的声誉,更重要的是影响人们的身体健康。因此,开展计算机技术应用于产品质量预测的研究工作,可以有效地控制产品质量,同时,也可以减少质检的工作量,对于保证乳制品质量具有重要意义。高级人工智能在乳制品生产的过程中进行质量智能控制,是当今乳制品企业质量管理的发展趋势。本文针对乳制品品种繁多,生产工艺复杂等特点,以及生产中每一个环节的控制,可能会成为影响产品最终质量的关键因素,结合目前流行的神经网络技术,对乳制品质量预测进行了仿真实验,并在系统中具体实现。主要工作如下:1.在深入研究乳制品生产过程的基础上,详细分析了乳制品质量预测软件的系统需求,提出了乳制品中小企业质量预测系统的解决方案,对乳制品预测系统进行了模块化设计。2.针对人工神经网络具有自适应、自组织、自学习能力的特征,结合乳制品生产特点,建立了基于人工神经网络的乳制品质量预测模型,为本课题深入研究奠定了理论基础。3.利用归一化方法对样本数据进行处理,以减少对神经网络模型训练的影响。通过建立的BP神经网络和RBF神经网络的两种质量预测模型,应用MATLAB软件对其分别进行了仿真实验。通过实验数据确定了应用于质量预测系统的RBF神经网络模型。4.对系统数据库进行了分析设计与实现,应用新型的SSH框架,即以Hibernate作为持久层, Struts作为表示层,并结合Spring作为业务层与框架整合的方法,实现了B/S模式的乳制品质量预测系统的部分功能。结果验证了该方法应用于乳制品质量预测系统是有效可行的。更多还原

【Abstract】 Nowadays food quality and safety is the most concern of the people. With the growth of Chinese economic, the increase in national income and the continuously improve in people’s life quality, dairy products have become one of the essential food in people’s life. Its quality not only affects the reputation of manufacturing company, but also affects people’s health. Therefore, the study of the computer technology using in product quality prediction can effectively control the product quality and also can reduce the workload of Quality Supervision and Inspection. Meanwhile, it is of great significance for ensuring dairy quality. Using advanced artificial intelligence in the process of dairy production to control quality intelligently is the trend of development in the quality management of today’s dairy companies.Due to the wide variety of dairy kinds and the complexity of the production process as well as every aspect of production control may become the key factor affecting the final product quality, this paper made a simulation of dairy quality prediction with the popular neural network technology and realized it in the system. The main work is as follows:1. Based on the deep study of the dairy production process, the paper analyzed the systematic requirements of the dairy quality prediction software in detail, proposed solution of the quality prediction system for dairy products and made a modular design to a dairy forecasting system.2. Due to the characteristics of artificial neural network, such as self- adaptive, self-organizing and self-learning ability, according to the dairy production characteristics, this paper built a dairy quality prediction model based on artificial neural networks, which has laid a theoretical foundation for a deep study of the topic.3. This paper used the normalization method to process the sample data in order to reduce the influence to the neural network model training. It made simulations separately using the quality prediction models of the BP neural network and RBF neural network by means of MATLAB software. The experimental data shows that RBF neural network model could be used in quality prediction system.4. This paper made analysis, design and Implementation to the system database by means of new SSH framework. That is to say, using Hibernate as the persistence layer, struts as the presentation layer spring as service layer and combined with framework integrated approach, this paper realized some functions of the B / S modeled dairy quality prediction system. The result shows that the method using in dairy quality prediction system is feasible and effective.更多还原