Prediction Of Vegetable Price In Henan Province Based On ARIMA-BP Neural Network

Vegetables are the necessities of residents' living,and the price of vegetables is stable in relation to the income level of farmers and the quality of life of residents.In recent years,the supply and demand of vegetables in China have continued to grow.As of 2017,China's vegetable planting area has increased by more than 68 million mu,with a total consumption of 508 million tons.At the same time,due to various factors such as weather,transportation,inflation,and asymmetric sales information,the price of some vegetables fluctuated drastically,The constant emergence of words has caused relatively bad public opinion and has attracted the attention of the government and the public.Therefore,it is of great practical significance to study the law of vegetable price changes.Based on the operation law of vegetable price itself and the formation process of influencing factors,this paper first analyzes the characteristics of vegetable price fluctuations,and uses principal component analysis method,combined with its volatility characteristics,to quantitatively study the influencing factors of potato price,and establishes the indicators of potato price influencing factors.The system explores the main factors that influence potato price volatility.Finally,a combined forecasting model based on ARIMA model and BP neural network is established and a predictive analysis of future potato price fluctuations is made.The main findings are as follows:First of all.in order to explore the fluctuation law of vegetable prices,vegetables were divided into three categories:roots,stems,and leafy vegetables according to different vegetable organs.The data of 72 periods from 2013 to 2018 were selected and analyzed separately.The inter-annual and intra-annual fluctuations of the three major types of vegetables.The results showed that the fluctuation characteristics of similar vegetables were basically the same,and the price fluctuations between the years were not large.The vegetable prices in the year showed three different fluctuation laws:"one peak and one valley","two peaks and two valleys" and "peaks and valleys are not obvious''.Combined with the law of vegetable price fluctuation,an index system affecting vegetable price fluctuation was established.The principal component analysis method was used to study the specific influence degree of residents' income level,social development level and climatic conditions on vegetable price fluctuation.The results showed that the price of vegetables was affected by many factors.The production cost,climate factor,cucumber price and eggplant price had a greater impact on vegetable prices.The factors such as the price of the previous period and the level of urbanization had less impact on vegetable prices.Finally,the price of potato in Henan Province from December 2011 to August 2018 was selected.From August 2011 to August 2017 as a training set,the monthly data from September 2017 to August 2018 was used as a verification set.A combined forecasting model based on ARIMA model and BP neural network and a predictive analysis of future vegetable price fluctuations.From the results,the established combination forecasting model has a better prediction effect,which can well fit the fluctuation characteristics of vegetable prices.The maximum error of the predicted value is only 0.03 yuan/kg,and the relative error is only 2.3%,and the predicted value is Historical prices reflect the same regularity and have certain rationality.