National cotton yield forecasting analysis based on multi-objective locust optimisation algorithm
With the impact of global climate change and agricultural development,forecasting and analyzing cotton yields is crucial for agricultural planning and resource allocation.In order to provide a more accurate prediction of national cotton yield,a multi-ob-jective locust optimal combination forecasting is proposed.Three single models,ARIMA time series model,Least Squares Support Vector Machine LSSVM model,and Recurrent Neural Network RNN model,are firstly applied to forecast the national cotton pro-duction data from 2009 to 2023.Then,a set of optimal solutions were obtained through the multi-objective locust iterative optimiza-tion process,and the single model prediction results were compared with the prediction results of the combined prediction method.It is verified through examples that the combined prediction method using multi-objective locust optimisation predicts results with smaller error and higher fitting degree,which proves that the model has good value in practical application and better reflects the actual changes of cotton production.Finally,using this method to forecast the national cotton production in 2024-2026 can provide a reference for the development of the cotton industry.
multi-objective locust optimisation algorithmcotton yieldcombined predictionLSSVM modelRNN modelARIMA model
万方数据
维普
