农产品质量监管与追溯系统设计

发布时间：2018-12-14 19:24

【摘要】：随着人民生活质量的改善,对农产品的质量安全意识越来越高,农产品的种类、安全因素和流通环节复杂,建立全过程的监管与追溯非常重要,而大数据技术的迅猛发展更是为农产品质量监管追溯系统提供了新的平台。在分析国内外农产品质量监管方面和追溯系统技术的基础上,利用大数据技术,基于Hadoop平台架构了农产品质量监管与追溯系统。通过对SVM和BP神经网络算法的深入分析,构建了基于SVM算法的农产品区域质量监管预测模型,在选择最佳的惩罚因子和核函数参数时,将原始数据中的农药污染指数和重金属污染指数平均分为K组,每组数据分别做一次验证集,剩下的K-1组数据作为训练集,用验证集的分类精度的平均值作为分类器最终的交叉验证精度,用最大精度对应的惩罚因子和核函数参数进行训练,实现对需要重点监管的农产品区域的预测,与BP神经网络算法进行对比,SVM算法的分类准确率提高了10%;为了实现对农产品腐败率数据未来一段时间的预测,构建了基于SVR算法的农产品时序质量预测模型,将农产品近期的腐败率数据分为两组,最佳的惩罚因子和核函数参数选取与上一模型相同,用第一组数据训练得到的模型进行预测,预测数据与第二组真实数据进行绝对误差和相对误差分析,与BP神经网络算法进行对比,在预测值和真实值的相关系数方面SVR算法比BP神经网络算法提高了近5%,更加逼近真实值。从Web端和Android端的角度对农产品质量监管与追溯系统进行设计。搭建测试环境,分别对Hadoop监管追溯平台、Web端和Android端的主要功能进行了测试,测试结果表明该系统在农产品质量监管与追溯方面具有一定的实用价值。
[Abstract]:With the improvement of the people's quality of life, the awareness of the quality and safety of agricultural products is becoming higher and higher, and the types, safety factors and circulation links of agricultural products are complicated. It is very important to establish the supervision and traceability of the whole process. The rapid development of big data technology also provides a new platform for the traceability system of agricultural product quality supervision. Based on the analysis of domestic and foreign agricultural product quality supervision and traceability system technology, using big data technology, the agricultural product quality supervision and traceability system is constructed based on Hadoop platform. Based on the analysis of SVM and BP neural network algorithm, the prediction model of agricultural product regional quality supervision based on SVM algorithm is constructed. When selecting the best penalty factor and kernel function parameter, The pesticide pollution index and heavy metal pollution index in the original data were divided into two groups: group K, each group of data made a verification set, and the remaining group of K-1 data as a training set. The average value of the classification accuracy of the verification set is used as the final cross validation accuracy of the classifier, and the penalty factor corresponding to the maximum accuracy and the kernel function parameters are used to train the forecast of the agricultural product area which needs the key supervision. Compared with the BP neural network algorithm, the classification accuracy of the SVM algorithm is improved by 10%. In order to predict the corruption rate of agricultural products for some time in the future, a forecasting model of agricultural product time series quality based on SVR algorithm is constructed, and the recent data of agricultural corruption rate are divided into two groups. The best penalty factor and kernel function parameters are chosen the same as the previous model. The model trained by the first group of data is used to predict, and the absolute error and relative error between the predicted data and the second real data are analyzed. Compared with the BP neural network algorithm, the correlation coefficient between the predicted value and the real value is improved by nearly 5% by the SVR algorithm compared with the BP neural network algorithm, and the real value is closer to the real value. From the point of view of Web and Android, the system of agricultural product quality supervision and traceability is designed. The main functions of Hadoop supervisory traceability platform, Web terminal and Android terminal are tested in the test environment. The test results show that the system has certain practical value in agricultural product quality supervision and traceability.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52

【参考文献】