基于ARIMA-LSTM的地下水位预测模型OA
Groundwater Level Prediction Model Based on ARIMA-LSTM
针对地下水位预测过程中线性与非线性特征难以同时兼顾的问题,提出了一种基于差分自回归移动平均模型和长短时记忆网络(ARIMA-LSTM)的组合预测模型.该模型首先使用 ARIMA 模型处理序列中的线性趋势,然后利用 LSTM 模型强大的非线性拟合能力来修正 ARIMA 模型的预测误差(残差).最终,通过合并这两部分的输出,实现对地下水位预测值更全面、更准确地把握.实验结果表明,基于 ARIMA-LSTM 的地下水位预测模型的 MAE、MSE 和 RMSE 分别为 0.011 m、0.000 3 m 和 0.017 m,均显著低于单一的 ARIMA 和 LSTM 模型.ARIMA-LSTM 模型误差较小,为地下水资源的科学评价与合理利用提供了可靠依据.
In order to address the problem of simultaneously considering linear and nonlinear characteristics in groundwater level prediction,a combined forecasting model based on the Autoregressive Integrated Moving Average Model and the Long Short-Term Memory Network(ARIMA-LSTM)is proposed.Firstly,the model employs the ARIMA model to handle the linear trends in the sequence.Then,it leverages the LSTM model's strong nonlinear fitting capability to correct the prediction errors(residuals)of the ARIMA model.Finally,by integrating the outputs of these two parts,more comprehensive and accurate groundwater level predictions are achieved.Experimental results show that the MAE,MSE,and RMSE of the groundwater level prediction model based on ARIMA-LSTM are 0.011 m,0.000 3 m,and 0.017 m,respectively,which are significantly lower than those of the single ARIMA and LSTM models.The smaller error of the ARIMA-LSTM model provides a reliable basis for the scientific evaluation and rational utilization of groundwater resources.
张静
河南工业贸易职业学院,河南 郑州 450053
信息技术与安全科学
地下水位预测模型时间序列ARIMALSTM
groundwater levelprediction modeltime seriesARIMALSTM
《现代信息科技》 2026 (8)
144-147,152,5
水利部重大科技项目(SKS-2022142)2025年度河南省高等学校重点科研项目指导性计划(25B510010)水利部黄河下游河道与河口治理重点实验室开放课题基金项目(2025007)
评论