Research On Safety Early-warning For Cold-chain Logistics Of Agricultural Products

With the improvement of people’s living standard and the development of Internet of Things technology, the quality and safety of agricultural products have put forward higher requirements. Although agricultural cold chain logistics is an important part of the agricultural industrialization in our country, there’re still not perfect in the cold chain logistics infrastructure, the information technology and the standard system. Therefore, it is growing in demand for the effective cold chain monitoring, security early warning and traceability management of agricultural products. In this context, the early warning system of agricultural products cold chain logistics is studied. Under the analysis of the effect of agricultural products cold chain logistics security causes, respectively, the early warning index system for fruits and vegetables, meat and dairy products are established. The fuzzy analytic hierarchy process (AHP) and improved support vector machine (SVM) algorithm are combined to build early warning models. The early warning of agricultural products cold chain logistics traceability system is developed to achieve agricultural products cold chain logistics of advance warning and retroactively. The main research contents and results of this paper are as follows:(1) Establish an early warning indicators secure system for cold chain logistics of agricultural products based on Fuzzy Analytic Hierarchy process. According to the source and nature of the hazard arising in the course of agricultural cold chain logistics, and the followed relevant norms of agricultural products in the circulation process, this paper analyses the three aspects of internal factors, environmental factors and logistics factors. A set of indicators that can reflect the reality of the situation was found out. The early warning indicators’weights were calculated by applying Fuzzy Analytic Hierarchy process.(2) Propose and verify the effectiveness of the support vector machine regression prediction model based on parameter optimization in solving the early warning of agricultural products logistics problem. We established and solved the model by using polynomial kernel, RBF kernel and sigmoid kernel function respectively. Through comparing the forecasting accuracy, the RBF kernel function with the best prediction accuracy has been found in support vector machine model. In order to improve the prediction accuracy of support vector machine algorithm, we presented to use adaptive mutation particle swarm optimization algorithm to optimize the parameters of support vector machine, and compared with the BP neural network, support vector machine algorithm (not optimized) and cross validation method of optimizing the parameters of support vector machine algorithm. The results show that BP neural network algorithm in the early warning model has poor stability, and low precision. By contrast, the support vector machine algorithm that is optimized by the adaptive mutation particle swarm optimization has the high stability and best accuracy.(3) Design and develop an agricultural products cold chain logistics security early warning system with EPC Internet of things technology as the background, and the B/S JSP+Servlet+JavaBean model and SQL Server as the background database. The system can realize the early warning and retroactively of the agricultural products in the whole supply chain. It is designed to reduce unnecessary loss of agricultural products, improve the efficiency of the cold chain logistics operation, ensure the quality and safety of agricultural products. In an open, networked and efficient early warning and traceability query environment, the system provide the basic information delivery and the processing channels for the supply chain related enterprises and users with a set of perfect support group collaboration, process control, information query and management functions of the early warning and traceability system.