Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield
Purpose: This paper develops a high-precision yield fusion prediction model for the sugarcane industry in Chongzuo, Guangxi, based on the trend yield and meteorological yield using the long short-term memory (LSTM) model to cope with the multiple factors affecting sugarcane production. Decision supp...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-10-01
|
| Series: | Agronomy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4395/14/11/2512 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846154753322516480 |
|---|---|
| author | Pengcheng Ma Na Zhang Yunhai Yang Zeping Wang Guodong Li Zhishan Fu |
| author_facet | Pengcheng Ma Na Zhang Yunhai Yang Zeping Wang Guodong Li Zhishan Fu |
| author_sort | Pengcheng Ma |
| collection | DOAJ |
| description | Purpose: This paper develops a high-precision yield fusion prediction model for the sugarcane industry in Chongzuo, Guangxi, based on the trend yield and meteorological yield using the long short-term memory (LSTM) model to cope with the multiple factors affecting sugarcane production. Decision support is provided to agricultural producers, policymakers, and supply chain managers so that they can plan resource allocation, market strategies, and policy directions more effectively. Methods: The paper modeled trend yield and weather yield separately to explore the complex relationship between the two in influencing sugarcane production. Trend yields were predicted using the exponential smoothing and multilayer perceptron (MLP) models, while meteorological yields were modeled using stepwise regression. The predicted yields were used again as input variables into the LSTM deep learning network to fit the nonlinear relationship between the two yields. Results: The results showed that (1) the fusion strategy of meteorological yield and MLP trend yield adopted by the model was superior to the fusion strategy of meteorological yield and exponentially smoothed trend yield, achieving a very low mean square error (MSE) of 0.011 and a goodness of fit as high as 0.979, which indicated that the model prediction was highly in agreement with the actual yield, confirming the validity of the method. (2) The prediction curve is basically consistent with the trend of actual sugarcane yield, which predicts that the sugarcane yield in Chongzuo, Guangxi, is expected to maintain a stable and small growth trend in the next eight years. (3) The fusion prediction model proposed in this study provides an accurate and practical solution for sugarcane yield prediction in Chongzuo, Guangxi, with the unique advantage of effectively analyzing and integrating the natural and socio-economic factors affecting the yield, which is of significant reference value for the prediction of sugarcane yield in the local area and even in similar ecoregions. |
| format | Article |
| id | doaj-art-32c61cfe533e4834b8255b0d9ee10ba6 |
| institution | Kabale University |
| issn | 2073-4395 |
| language | English |
| publishDate | 2024-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agronomy |
| spelling | doaj-art-32c61cfe533e4834b8255b0d9ee10ba62024-11-26T17:44:15ZengMDPI AGAgronomy2073-43952024-10-011411251210.3390/agronomy14112512Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological YieldPengcheng Ma0Na Zhang1Yunhai Yang2Zeping Wang3Guodong Li4Zhishan Fu5School of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004, ChinaSchool of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004, ChinaSchool of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004, ChinaGuangxi Academy of Agricultural Sciences, Nanning 530007, ChinaSchool of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004, ChinaSchool of Mathematics and Computing Science, Guilin University of Electronic Technology, Guilin 541004, ChinaPurpose: This paper develops a high-precision yield fusion prediction model for the sugarcane industry in Chongzuo, Guangxi, based on the trend yield and meteorological yield using the long short-term memory (LSTM) model to cope with the multiple factors affecting sugarcane production. Decision support is provided to agricultural producers, policymakers, and supply chain managers so that they can plan resource allocation, market strategies, and policy directions more effectively. Methods: The paper modeled trend yield and weather yield separately to explore the complex relationship between the two in influencing sugarcane production. Trend yields were predicted using the exponential smoothing and multilayer perceptron (MLP) models, while meteorological yields were modeled using stepwise regression. The predicted yields were used again as input variables into the LSTM deep learning network to fit the nonlinear relationship between the two yields. Results: The results showed that (1) the fusion strategy of meteorological yield and MLP trend yield adopted by the model was superior to the fusion strategy of meteorological yield and exponentially smoothed trend yield, achieving a very low mean square error (MSE) of 0.011 and a goodness of fit as high as 0.979, which indicated that the model prediction was highly in agreement with the actual yield, confirming the validity of the method. (2) The prediction curve is basically consistent with the trend of actual sugarcane yield, which predicts that the sugarcane yield in Chongzuo, Guangxi, is expected to maintain a stable and small growth trend in the next eight years. (3) The fusion prediction model proposed in this study provides an accurate and practical solution for sugarcane yield prediction in Chongzuo, Guangxi, with the unique advantage of effectively analyzing and integrating the natural and socio-economic factors affecting the yield, which is of significant reference value for the prediction of sugarcane yield in the local area and even in similar ecoregions.https://www.mdpi.com/2073-4395/14/11/2512fusion prediction modelLSTMtrend yieldmeteorological yieldforecast of sugarcane yield |
| spellingShingle | Pengcheng Ma Na Zhang Yunhai Yang Zeping Wang Guodong Li Zhishan Fu Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield Agronomy fusion prediction model LSTM trend yield meteorological yield forecast of sugarcane yield |
| title | Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield |
| title_full | Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield |
| title_fullStr | Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield |
| title_full_unstemmed | Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield |
| title_short | Forecast of Sugarcane Yield in Chongzuo, Guangxi—LSTM Model Based on Fusion of Trend Yield and Meteorological Yield |
| title_sort | forecast of sugarcane yield in chongzuo guangxi lstm model based on fusion of trend yield and meteorological yield |
| topic | fusion prediction model LSTM trend yield meteorological yield forecast of sugarcane yield |
| url | https://www.mdpi.com/2073-4395/14/11/2512 |
| work_keys_str_mv | AT pengchengma forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield AT nazhang forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield AT yunhaiyang forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield AT zepingwang forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield AT guodongli forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield AT zhishanfu forecastofsugarcaneyieldinchongzuoguangxilstmmodelbasedonfusionoftrendyieldandmeteorologicalyield |