Research On Garlic Price Forecast Based On Improved EEMD-GRU Combined Model

China's garlic production accounts for more than 70% of the world's total output,and it is the single agricultural product that China exports the most foreign exchange earnings.In recent years,the price of garlic has fluctuated sharply and frequently,which has brought severe challenges to the healthy and sustainable development of the garlic industry.Based on the detailed analysis of garlic price fluctuation characteristics,this study constructed a combined forecasting model based on time-frequency decomposition and neural network to realize the accurate forecast of garlic price.Based on the establishment of the price prediction model,a garlic price prediction system was developed using key technologies such as Java EE and data processing.The specific research content is as follows:(1)Research on the law of garlic price fluctuation.The characteristics of garlic price fluctuations are analyzed by combining the macro and micro methods.First,the garlic price fluctuation cycle is divided by the peak-valley method.The fluctuation cycle can be divided into five,and the length of the five fluctuation cycles is different.Further use the STL trend decomposition method to decompose the garlic price series,and obtain the seasonal item,trend item and residual item.Analyzing each item separately,it can be seen that the price of garlic has very significant seasonal fluctuation characteristics,the fluctuation period is 12 months,and the price fluctuation trend of each period is basically the same;the price of garlic is susceptible to irregular factors such as natural disasters.Influence.Finally,the correlation coefficient between the garlic influencing factors and the garlic daily price series is analyzed.The results indicate that the influencing factors have little influence on the garlic daily price series,and it is still necessary to predict the daily price changes of garlic through volatility characteristics.(2)Research on garlic price forecast.Giving full play to the advantages of a single model,combining different models,and building a combined model for price prediction will significantly improve the accuracy of price prediction.Based on the comprehensive analysis and comparison of price forecasting models,the research constructs the EEMD-GRU garlic price combination forecasting model.First,the EEMD decomposition method is used to decompose the garlic price series on multiple time scales,and 12 IMF eigenmode functions with different time scales are obtained,and their classification and integration are combined to obtain high-frequency,intermediate and low-frequency terms.Analyzing the high-frequency items,intermediate-frequency items,the correlation coefficients of the low-frequency items and the garlic price series and the contribution rate to the garlic price series,it is found that the low-frequency and intermediate-frequency items have a high contribution rate to the garlic price.The combination of the two is 94%.The item is combined with the low-frequency item to obtain the reconstructed garlic price sequence.The reconstructed garlic price sequence is imported into the GRU model,50 training times and 30 hidden neurons are selected for garlic price prediction,and the prediction result is obtained.Comparing the prediction results obtained by the EEMD-GRU combined model with the prediction results obtained by the ARIMA,ARIMA-SVR and LSTM models through evaluation indicators such as MSE,it is found that the trend prediction accuracy of the EEMD-GRU combined model is the highest,the prediction error is the smallest,and the prediction is The effect is significantly better than ARIMA,ARIMA-SVR and LSTM models.(3)Research and development of garlic price analysis and forecasting system.Based on the analysis of garlic price fluctuation characteristics and the research of price prediction models,a garlic price analysis and prediction system has been developed.The construction of the garlic price prediction system can provide efficient and accurate data analysis and forecast services for all links of the garlic industry,assist the relevant personnel in the garlic industry to make scientific decisions,and promote the healthy and sustainable development of the garlic industry.