Prediction of Soil Moisture Based on BP Neural Network

Abstract

Grassland is the largest green ecological barrier in China, and reasonable grazing policies are key to ensuring people's livelihoods. The soil moisture data in this article is time series data and is related to multiple factors. On the premise of maintaining the same grazing strategy, first expand the collected data and use it as input. The L-BFGS algorithm was used to iteratively update the parameters, and the ReLU activation function was used to select the final model by comparing the Goodness of fit of models with different structures. A BP neural network model was established to predict the soil moisture at a depth of 40cm in 2022, providing theoretical reference for the research on grazing strategies of grasslands in pastoral areas in China