Abstract
Potato, as a crucial global staple crop, plays a pivotal role in ensuring global food security. In China, both the cultivation area and yield of potatoes rank among the highest globally, highlighting its significance in agricultural production. Accurate prediction of potato yield is essential for guiding cultivation management and making related decisions. The objective of this study is to develop an improved potato yield prediction model to help address supply gaps in the fresh potato market domestically, particularly between northern and southern regions. Considering the limitations of traditional Backpropagation (BP) neural networks in terms of prediction accuracy and robustness, this study optimizes the BP neural network using the Whale Optimization Algorithm (WOA). By analysing meteorological factors, field hydrothermal factors, and potato yield data collected from field Internet of Things (IoT) systems between 2010 and 2022, we constructed and compared three different models: a traditional BP neural network model, a BP neural network model optimized by Genetic Algorithm (GA), and a BP neural network model optimized by WOA. The research findings indicate that the WOA-BP model significantly outperforms the other two models in prediction accuracy, with an R2 (coefficient of determination) value of 0.9764. Moreover, the high degree of fit between predicted and observed values validates the scientific validity and accuracy of the WOA-BP model in potato yield prediction.
Graphical Abstract
