马尔科夫链与GM(1,1)模型在岩溶地下水四氯化碳污染预测中的应用对比

Comparative analysis of forecasts in karst groundwater pollution based on Markov Chain and GM(1, 1)

为选择更优的岩溶地下水水质预测模型,以七里沟岩溶地下水源地四氯化碳污染区为研究区域,选用马尔科夫链与GM(1,1)模型分别对四氯化碳污染进行预测研究,并将预测值与实测值进行对比分析,探讨两种预测模型的适用性及差异性。结果表明:研究区四氯化碳污染监测数据具有一定的波动性,马尔科夫链模型预测结果拟合程度更好,GM(1,1)适用于具有一定变化趋势的水质监测数据。

In order to choose a better prediction model for the karst underground water quality, Markov chain and GM （1, 1） were taken separately to predict the level of CCl4 in contaminated area of Qiligou water source. And the applicability and difference of the two prediction models were discussed through the comparison and analysis between measured values and predicted values. The results indicated that Markov chain prediction model showed higher fitting for the measured values with regular fluctuation, and GM （1, 1） was suitable to the water quality monitoring data with certain tendency of change.