Prediction of Soil Temperature Field in Panax Notoginseng Plough Layer Based on PSO-LSTM Neural Network

Abstract

Soi1 temperature in the tillage layer has a significant impact on crop growth, so the accurate prediction of its change trend can help intelligent agricultural systems to make autonomous decisions and ensure the normal growth of plants. In this paper, an accurate prediction model of soil temperature in the tillage layer is established based on PSO-LSTM. First, the particle swarm optimization algorithm is used to optimize the key parameters of the LSTM model, which effectively improves the model performance. Then, kriging interpolation is used to estimate the soil temperature distribution in the tillage layer, and uneven distribution results are obtained. Finally, an experiment is conducted with the soil data actually collected from the Panax notoginseng cultivation layer. The results show that the proposed soil temperature prediction model in this paper has higher accuracy, which can achieve accurate prediction of soil temperature and effectively guide the intelligent agricultural system to make autonomous decisions on soil temperature.