Research On Forecast And Early Warning Of Soybean Price Fluctuation

With the progress of the times,people pay more attention to the absorption of nutrients,and soybean,as the agricultural product with the most protein content,occupies an important position in the daily consumption of the people.Although China is a big consumer of soybeans,it relies heavily on imported soybeans.It should pay more attention to the market price of soybeans to balance soybean imports and exports and promote the orderly development of China’s soybean industry.This paper studies the characteristics of the volatility of soybean prices in China,finds out the key factors that affect soybean price volatility,establishes corresponding forecasting and early warning models,and provides a basis for scientific decision-making for soybean growers and soybean downstream industries,and provides a reference for the government to macro-control market prices.This article first analyzes the status quo of soybeans in China,analyzes the production of soybeans in China,including the production area,and then analyzes the relationship between supply and demand and the downstream industries of soybeans,and finally analyzes the import and export situation: it is found that China’s soybeans are heavily dependent on imported soybeans;The analysis result of soybean current situation,further analyze the price of soybean;according to the monthly price data of soybeans in China’s farmer’s market,the overall fluctuation characteristics of the monthly price of soybeans in China are analyzed,and the time series decomposition method and HP filter method are used to analyze the fluctuation of soybean prices.The characteristics are further decomposed,and the seasonal fluctuation characteristics,random factor fluctuation characteristics and trend cycle fluctuation characteristics of soybean prices are obtained;the influencing factors of soybean price fluctuations are analyzed according to the price fluctuation characteristics,and the literature analysis method is first used to summarize the factors that may affect soybean price fluctuations.Then,the lasso algorithm is used to identify the factors that affect soybean price fluctuations,and the factor variables that have little effect on soybean price fluctuations are compressed to 0,so as to select the factors that really affect soybean price fluctuations.To establish a soybean price forecast and early warning model through influencing factors,the basic idea is as follows: build an ARIMA model,apply the ARIMA model to obtain the predicted value of the influencing factors,and then input the historical data of soybean prices and the data of influencing factors into the established BP neural network for learning After training,the forecast model is obtained,and the forecast value of soybean price can be obtained by inputting the forecast value of the influencing factor.The result shows that the true value of the soybean price and the forecast value are fit = 0.933,and the forecast result is good;then an early warning system is established,which is related to the time difference The analytical method calculates the correlation coefficient and establishes a black warning model.Finally,the predicted value of soybeans and the real value are input into the warning model.The result shows that the real value of the warning is exactly the same as the predicted value of the warning.It proves that the established soybean price forecast and early warning model has strong practicability and feasibility.