摘要
探讨新疆生态足迹与其主要社会经济影响因子之间的内在关系。计算新疆1990-2006年的人均生态足迹,通过应用偏最小二乘回归(PLS)法和灰色关联熵法,进行比较和印证分析,揭示影响生态足迹的各种因素的重要性程度,筛选出最主要因素,并通过PLS回归和灰色预测拟合效果的比较,选取预测精度更高的模型。研究结果如下:能源消费总量、总人口、人均GDP、城镇居民人均可支配收入、社会消费品零售总额、耕地面积6项因子均属于生态足迹的显著影响因子;PLS回归模型拟合的效果优于GM(1,1)模型,说明该模型预测精度更高,可用于未来生态足迹的动态预测。
For a case study of Xinjiang in China,inner relationship between ecological footprint （EF） and its influence factors were further revealed by the quantitative analysis results. First, the per capita EF and ecological carrying capacity （EC） for Xinjiang was calculated during the period from 1990 to 2006 by employing the quantitative method for EF. Next,after analyzing the concept and the model of the EF, 14 socio-economic influence factors of EF for modeling were selected,and partial least squares （PLS） method and grey relation entropy analysis were used to select the important factors influencing the EF. Quantitative analysis results of the two methods are feasible and credible,and can be referred to one another. It is found that the most important factors influencing the EF are： energy consumption, total population, per capita GDP, annual per capita disposable income of urban households, total retail sales of consumer goods,cultivated areas. Then,the above six factors were chosen as the main influencing factors in this study to be involved in the subsequent analysis. Finally,prediction models of PLS regression and GM（1,1） were Constructed based on the historical time series data, and comparisons between the predicted value and the actual value of the two models were listed. Simulation results demonstrated the accuracy and the reliability of PLS regression model presented in this study.
出处
《地理与地理信息科学》
CSSCI
CSCD
北大核心
2010年第6期70-74,共5页
Geography and Geo-Information Science
基金
上海重点学科建设项目(B410)
关键词
生态足迹
影响因子
偏最小二乘法
灰色关联熵
动态预测
ecological footprint （EF）
influence factor
partial least squares （PLS）
grey relation entropy analysis
dynamic prediction
