基于Holt指数平滑模型的地下水水质预测

地下水水质预测是地下水污染防治研究工作的重要组成部分。本文介绍了Holt指数平滑模型的基本原理及平滑参数的确定方法，在此基础上，以阿拉善腰坝绿洲地下水水质预测为实例对地下水HCO3-、Cl-、SO42-和TDS进行了预测，并与灰色模型预测结果进行了对比。结果表明2006年HCO3-、Cl-、SO42-和TDS浓度值将分别达到198．260、77．290、114．850和508mg／L。应查清地下水各种污染源及各污染源对地下水水质的影响程度，以便采取有效措施，遏制地下水水质进-步恶化。Holt指数平滑模型统-考虑了所有已有时间序列资料对预测值的影响，提高了预测精唐．适用于已有较长时间序列资料前提下的预测。

Groundwater quality prediction based on Holt exponential smoothing model

Groundwater quality prediction is an essential work for groundwater pollution prevention. A Holt exponential smoothing model is proposed in the study, and the method for determining the model parameters is introduced. The concentrations of HCO3-、Cl-、SO42- and TDS of the groundwater over the Alashan Yaoba oasis is predicted using the proposed model, and the predicted results is compared with those predicted by the gray model prediction. The results show that the concentrations of HCO3-、Cl-、SO42- and TDS will increase to 198. 260, 77. 290, 114. 850 and 508mg/L in 2006, which suggests that internal and external factors influencing water quality should be identified in order to take effective measures to prevent further deterioration of groundwater quality. Holt exponential smoothing model takes into account all existing time series data, improves the prediction accuracy. It is optimal in time series prediction when possessing long time series data.