在传统灰色模型的基础上，遵循数据“重近轻远”的原则，运用了灰色新陈代谢GM（1，1）预测模型对我国大气环境发展趋势进行了预测，及时考虑了系统发展过程中的扰动因素，在补充新信息的同时去掉因时间推移使信息意义降低的老信息，比常规的GM（1，1）模型更好的反映出了系统当前的特征。通过预测可以得出，实行节能减排后，二氧化硫、烟尘和工业粉尘排放量分别降低了34．48％、28．61％和38．71％，大气环境质量有明显好转；其次，对残差序列进行分析，并利用残差周期修正对残差序列进行了修正补偿，通过残差检验、关联度检验和后验差检验，得出基于残差周期修正的新陈代谢GM（1，1）模型精度较高，适合于大气环境发展趋势的预测。最后，进行了滚动检验，也称为事后检验，得出滚动精度较高，可信度较大。%On the basis of traditional grey model ,the principle of data“weighs nearly far” was followed by using the grey metabolism GM ( 1,1 ) prediction model predicts the development trend of atmospheric environment in China, considering the disturbance factors in the process of system development, timely supplement the new information and remove due to time reduce of the old information , the meaning of information than conventional GM(1,1) model better reflects the characteristic of the current system. Through the forecast it can be concluded that, after the implementation of energy conservation and emissions reduction , and industrial dust emissions of sulfur dioxide and fuel dust are reduced by 34 . 48%, 28 .61% and 38 .71%, atmospheric environment quality is improved obviously;Secondly , by analyzing the residual sequence , and by using the residual cycle correction residual sequence revised compensation by residual test,correlation test and posterior difference test, correction was got based on residual cycle metabolism GM(1,1) model precision,suitable for prediction of atmospheric environmental development trend.Finally,by carrying on the rolling test,also called a post-mortem examination,it is concluded that rolling precision and credibility were high.
展开▼
