Application Of Improved Grey-Markov Model In Peanut Yield Prediction

With the acceleration of urbanization,the structure of land and labor has changed,which has impacted the yield of crops.Peanut is an important oil crop in China,and accurate prediction of its yield has important practical significance for guiding agricultural production.In this paper,the influencing factors of peanut production capacity in Henan Province were studied,and the Grey-Markov model of metabolism was established to predict its yield.Firstly,the main factors affecting peanut production in Henan Province were selected,and the grey correlation analysis method was used to analyze the correlation degree.Secondly,the background value of the traditional grey model is optimized,and the model is optimized twice by Markov process,and an improved Grey-Markov model is established to fit the peanut production.Finally,the original data were metabolized,and the peanut yield in Henan Province in the next 5years was predicted using the above Grey-Markov model.The research contents and achievements are as follows:(1)The first six factors with strong correlation with peanut yield in Henan Province were: sown area,urbanization rate,rural electricity consumption,irrigation area,amount of chemical fertilizer converted into pure fertilizer and amount of agricultural plastic film.This shows that the yield of peanut mainly depends on natural factors.The rational use of scientific and technological means can improve the planting efficiency to a certain extent,such as improving irrigation facilities,increasing the input and use of mechanical equipment,and alternately applying organic fertilizer and chemical fertilizer to make peanut in a more suitable growth environment as far as possible.(2)Based on the peanut yield data of Henan Province in recent 10 years,an improved grey model was established.Firstly,the modeling data were processed by moving average,and then the background value construction sequence of the model was optimized to eliminate the error caused by internal parameters.The results show that the average relative errors of the model before and after the improvement are 2.04%and 1.89% respectively,which shows that the optimization parameters can only improve the model accuracy to a certain extent.In order to make the model accuracy level excellent,it is necessary to use other methods for secondary optimization.(3)Markov process was introduced to optimize the improved grey model for the second time,and the fitting results were revised year by year by using the transition probability between various states in the system.Finally,the average relative error of the model was 0.69%,reaching the excellent level.It was believed that the model could well fit the change trend of peanut yield in Henan Province.The above model was used to predict the peanut yield in the next five years.During this period,the original data series were continuously metabolized.The predicted peanut yield in Henan from 2021 to 2025 was 6202.2,6382.3,6699.6,6893.3 and 7246.3 thousand tons respectively,with an annual growth rate of about 4.03%.