The Research And Application Of The Prediction And Traceability Model Of Vegetable’s Quality And Safety Based On Support Vector Machine

Currently the food safety has become public safety in China, the vegetables is theimportant part of their daily consumption of the largest food quality and safety. Vegetablesafety is the relationship between the people’s life safety and health, economic developmentand national stability, social stability development of major issues. The developed countriesand some developing countries have construction as an important part of vegetable qualityand safety. Vegetable information traceability and quality prediction is related to the twoaspects of vegetable quality and safety. But at the moment of vegetables quality and safetyinformation and quality prediction research is not enough. At present domestic vegetabletraceability information model is simpler, the information acquisition of vegetables aresketchy; At present most of the units and the government only on vegetables qualityprediction for intuitive trend analysis to the naked eye, also not make more mathematicaltheory on vegetables quality prediction based on predictive model. Based on these two pointsin this paper, design the vegetable quality and safety prediction model based on supportvector machine (SVM), to a certain extent, to improve and make up for the present deficiencyin vegetable quality and safety prediction.In this paper, the main innovations include the following aspects:1. Vegetable quality safety traceability model is established. Through the analysis current ourcountry in the aspect of vegetables quality safety traceability, design a set of vegetablesquality and safety traceability model, the specific function of each part are analyzed, and therelationship between the parts. To realize forecast and roots of vegetables quality and safetymonitoring system of vegetables in the logo, made the vegetables tracing the source codingscheme, design has realized the vegetables quality traceability labels. Sign vegetables to makeuse of rfid and bar code technology, information collection and transmission, using thecomponent technology development system of key modules. Put forward a set of integrationof vegetable production source monitoring operation theories.2. Based on support vector machine (SVM) regression algorithm of vegetables time-seriesquality forecasting. By the support vector machine (SVM) regression algorithm, the principle of a vegetable from the history of the past data model, and then to predict the quality qualifiedrate of the specific time concrete vegetables this year, and according to the percent of pass tomake adjustments to the vegetables were and quality inspection. The model with vegetablesquality qualified rate of the actual value and predictive value of contrast, performanceassessment. At last, the imitation of the real and the feasibility of the model was verified.3. Based on support vector machine (SVM) classification algorithm of vegetables area qualityprediction model. Support vector machine (SVM) classification algorithm theoretical basis,Comprehensive pollution index of soil heavy metals and pesticide residues are two majorfactors in data modeling. Using the established model to predict harvest vegetables back anarea or a focus on quality is the focus of attention. Through the predicted classificationaccuracy for performance evaluation model. At last, the imitation of the real and thefeasibility of the model was verified.4. The vegetables quality prediction model is embedded into the information source model,prediction and implements a vegetable quality and safety traceability model. And completedthe several key function of the main modules of the model. Make the model and informationtraceability and quality prediction two major functions.