基于灰色马尔科夫模型的平顶山市空气污染物浓度预测

The Prediction of Air Pollutant Concentration in Pingdingshan Based on Grey Markov Model

选用平顶山市2005—2009年各空气污染物浓度作为原始数据序列,建立灰色马尔科夫预测模型,对未来10年的污染因子浓度进行预测.模型检验结果表明:均方差比值和小误差概率均为一级;运用灰色关联分析法计算各污染物原始数据序列与预测数据序列之间的关联度,定量描述灰色马尔科夫预测模型对于空气质量预测的精确度,平均精度达到99.9%,表明灰色马尔科夫预测模型对于空气质量预测有很高的实用性.

By selecting the air pollution concentration of Ping ding shan in 2005-2009 as the original time series,establish a gray markov prediction model,the predication of the concentration of the polluted factor in the next 10 years was made.The result shows：the mean square deviation ratio and the minimum mean squared error are 1 level.By applying correlation analysis of grey system theory,the original time correlation between the data and forecasting data sequences was calculated.It is used to describe the accuracy of the gray Markov prediction model for air quality prediction,the average accuracy is 99.9％.Precision result shows that the grey Markov prediction model for air quality prediction is highly practical.