Research And Application Of Vegetable Quality Safety Prediction Model Based On Support Vector Machine Regression

Vegetables account for a large proportion in the dietary consumption of residents,and the vegetable industry also plays an important role in the agricultural product economy,among which the quality and safety of vegetables has become the focus of public attention and permeates the whole industrial chain from raw material acquisition to final consumption.Afterwards,the loss of greater than the advance warning,using a large amount of data of vegetable market testing,of vegetables quality qualified rate to forecast the short-term control,to effectively control of vegetable quality qualified rate change rule,thus to reasonable control allocation of vegetable market,at the same time to the agricultural market stable development and the health of the residents living is of great significance.In this paper,based on the field of artificial intelligence technology,on the basis of full investigation of prediction research technology,three prediction algorithms are used to establish different prediction models of vegetable quality safety,and experimental comparative analysis is conducted to select the optimal prediction model.The main work completed in this paper includes the following aspects:(1)processing and analysis of vegetable quality qualification rate dataFirstly,the existing data and information of the laboratory platform were sorted out,and the types of enterprise users of the platform and the distribution of vegetable test data in each city were analyzed,so as to grasp the key factors of vegetable quality and safety judgment standards.The study of the experimental data from the bureau of agriculture and rural areas in hefei in 2009-2019,the main market in hefei vegetable pesticide residues detection data report,based on the results of screening analysis,data normalization processing and facilitate late prediction model building,to improve the training speed and precision of the model.(2)multiple models of vegetable quality and safety based on statistical learning are preferredBased on the analysis and processing of the data,the regression algorithm of support vector machine was used to train the model of the data samples,and the model parameters of the support vector machine were determined by the grid search method.In order to verify the stability and practicability of SVM regression model,BP neural network and logistic regression algorithm were selected in this paper to predict the qualified rate of vegetable quality.Finally,by comparing and analyzing the advantages and disadvantagesof the prediction methods and the performance of the model,the accuracy and stability of the support vector machine regression model are higher than those of the other two prediction models,and the reliability of the support vector machine regression model in predicting the qualified rate of vegetable quality in the short term is verified again.(3)design and implementation of vegetable quality and safety prediction systemThe vegetable quality safety prediction system was designed and implemented based on support vector machine regression model.The system consists of three modules: platform management,data operation and visual display,with functions of data acquisition and processing,data information display and visualization,etc.It can predict the fluctuation of vegetable quality qualification rate in the short term in the market in real time,dynamically and quantitatively.The application of the system in real life can effectively understand the local vegetable quality and safety environment,and timely predict the problems brought by the government supervision department to the quality and safety of agricultural products,so as to develop solutions to deal with emergencies,as well as provide the basis for the prevention and treatment of local testing departments.