An agricultural product yield prediction method and system

Abstract

The invention discloses an agricultural product yield prediction method and system. The method comprises the steps that historical data of acquiring all influence factors influencing the yield of agricultural products, and dividing the influence factors are divided into short-term influence factors and long-term influence factors, wherein the historical data corresponding to the short-term influence factors are daily data or monthly data, and the historical data corresponding to the long-term influence factors are seasonal data or annual data; Predicting future data corresponding to the short-term influence factors by adopting an autoregressive integral moving average model according to the historical data of the short-term influence factors to obtain predicted data of the short-term influence factors; Predicting future data of the long-term influence factors by adopting a grey model according to the historical data of the long-term influence factors to obtain predicted data of the long-term influence factors; Obtaining a neural network model; And inputting the prediction data corresponding to the short-term influence factors and the prediction data corresponding to the long-term influence factors into the neural network model to obtain a yield prediction value of the agricultural product. The method has the characteristic of high prediction precision.

Description

technical field [0001] The invention relates to the technical field of yield prediction, in particular to a method and system for predicting the yield of agricultural products. Background technique [0002] The output of agricultural products is closely related to people's daily life. Using past output data and historical data of a series of influencing factors, we can fully mine the trend of the data itself and the influence relationship and degree of each factor, and use certain methods and techniques to predict future output. Forecasting and pre-judgment will help the country, producers and consumers to better judge the economic situation and make correct decisions. [0003] In the prior art, a single model is often used to predict the output of agricultural products. However, there are many factors that affect the output of agricultural products, and their frequencies are different. If a unified model is used to fit it, each model Different applicability will lead to la...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, a single model is often used to predict the output of agricultural products. However, there are many factors that affect the output of agricultural products, and their frequencies are different. If a unified model is used to fit it, each model Different applicability will lead to large errors

[0004] At present, there are many studies based on mixed-frequency data, but most of them use low-frequency data indicators to analyze and estimate high-frequency data. For missing data, most of them are obtained through difference, substitution, model estimation, etc., so the original data used for analysis and estimation is itself estimated The data is not the actual data, and then the analysis and prediction obtained from this may have errors

Images