Regional Soil Nutrient Content Prediction Model Based on Big Data

作者：Yaqi Cui, Yue Qin · 发表于：Journal of Physics Conference Series · 年份：2023 · DOI：10.1088/1742-6596/2555/1/012005 · 被引用次数：1 · 研究领域：E-commerce and Technology Innovations

Abstract Soil nutrients play a decision-making role in environmental management, and the prediction of soil nutrients can be used to achieve precise fertilization and regulate production. In response to the problem of low accuracy of soil nutrient content prediction by traditional prediction models, this paper designs a prediction model of soil nutrient content based on big data statistics. Soil nutrient content data are collected using a spectral collector, and the data are smoothed, standardized, and orthogonalized to eliminate the data that affect the prediction accuracy. The processed spectrograms are analyzed to summarize the big data law of soil nutrients and clarify that different bands at the abrupt change of reflection curve correspond to different contents of soil nutrients. After repeating several times, the calibration of model accuracy is completed, and the design of the soil nutrient content prediction model based on big data statistics is realized. Through comparison experiments with the traditional model, it was verified that the designed model can improve the prediction accuracy by about 3 times and is more suitable for the prediction of soil nutrient content.