Research On Prediction Method Of Rice Sheath Blight Based On Combination Prediction Model

As one of the three major food crops in the world,rice plays a crucial role in national economic construction,but diseases have always been a key factor restricting rice high yield.If effective prevention and control measures are not implemented in a timely manner,it will cause unpredictable economic losses to the yield of rice.Therefore,adopting a combination of prevention and control methods is more conducive to protecting the high and stable yield of rice.Accurate prediction of rice sheath blight is beneficial for developing effective prevention and control measures,which is of great significance for ensuring the yield and quality of rice.In response to the severe situation of rice sheath blight,this article constructs a combination prediction model based on Random Forest Logistic to effectively predict the degree of rice sheath blight occurrence,thereby reducing the economic losses caused.This thesis mainly includes the following aspects:(1)Selection and processing of data.To ensure the authenticity and reliability of experimental data sources,this article combines the disease and meteorological data from the Anhui Province Crop Disease and Pest Monitoring and Warning System and the Anhui Provincial Meteorological Center to establish a dataset of influencing factors for the occurrence of rice sheath blight.Normalize the data using the maximum minimum method,and perform correlation analysis on the dataset using Pearson correlation coefficient to screen out five key influencing factors for the occurrence of rice sheath blight.(2)The construction of a combination prediction model based on Random Forest Logistic.The PSO algorithm is used to find the optimal weight ratio of random forest prediction model and logistic regression prediction model in the combination prediction,and the Random Forest Logistic combination prediction model is established.The combined model is used to predict the treated rice sheath blight dataset.The experimental results show that the RMSE of the combination prediction model based on Random Forest Logistic is 3.1236,MSE is 9.7566,MAE is 2.3948,and the determination coefficient R2 is0.9193.(3)Design and implementation of a rice sheath blight prediction system.A rice sheath blight prediction system was developed based on the Random Forest Logistic combination prediction model.The system can output predicted values and corresponding occurrence levels of rice sheath blight by inputting meteorological data for predicting the occurrence degree of rice sheath blight.The test results indicate that the system has good performance in predicting rice sheath blight.In summary,this article established a dataset for rice sheath blight through data preprocessing and feature selection;Using PSO algorithm to construct a Random Forest Logistic combination prediction model and predict the occurrence degree of rice sheath blight;Finally,a rice sheath blight prediction system was designed and completed.The research in this article provides reference significance for future research and development of disease prediction,and can to some extent prevent and control rice sheath blight,which is helpful for crop production and management,and reduces economic losses caused by diseases.