带有偏好约束的控规单元用地与交通协调DEA模型

DEA model of land use and traffic coordination for regulatory detailed planning unit with preference constraints

为定量评价城市控规阶段交通与土地利用的协调程度,分析了城市控规单元用地与交通协调的相互机理,指出了传统数据包络分析将输入、输出指标影响无差别处理与实际情况不符;针对控规单元特点和实际控制需求,建立了土地利用与交通协调的指标体系;根据数学规划的约束锥理论,借助层次分析法的思想,构建了反映用地指标和交通指标偏好的约束锥模型,将数据包络评价的客观分析与层次分析的主观判断科学结合,以隶属度函数判定土地利用与交通系统互为投入产出情况下的相互协调度,进而定量评价控规单元交通与用地不同规划方案的相互协调效果;以邯郸市南湖管理单元3为背景,设定22个外部交通小区和68个内部交通小区,预测了4类土地利用指标和9类城市交通指标,计算了模型输入指标和输出指标的偏好约束条件。研究结果表明:提出的模型克服了层次分析法过于依赖主观意向的局限,弥补了数据包络分析忽视决策者偏好的不足,缩小了土地利用与交通协调分析结果与实际状况之间的偏差;采用原始模型对6个街区相互协调度的计算结果均为0.80以上,全部为协调状态,采用改进后模型对其中4个街区相互协调度的计算结果为0.32~0.57,为不协调状态,2个街区相互协调度为0.80以上,与实际情况更相符。

To quantitatively evaluate the coordination degree of traffic and land use in urban regulatory detailed planning stage, the mutual mechanism of land use and traffic coordination for urban regulatory detailed planning unit was analyzed. It was pointed out that the traditional data envelopment analysis （DEA） makes an undifferentiated treatment between input indexes and output indexes, which was not consistent with actual situation. In view of the characteristics of regulatory detailed planning unit and actual control demand, the index system of land use and traffic coordination was established. According to the constraint cone theory based on mathematical programming, using the aid of analytic hierarchy process （AHP）, the constraint cone model responsing land index and traffic index preferences was constructed. The objective analysis of data envelopment evaluation was combined with the subjective judgment of analytichierarchy process. The mutual coordination degree of land use and traffic system in the cases of input and output was determined by using membership degree function. Therefore, the mutual coordination effects of the different planning schemes for the traffic and land use of regulatory detailed planning unit were quantitatively evaluated. Based on the No. 3 management unit of Nanhu in Handan City, 22 external traffic zones and 68 internal traffic zones were set up. 4 types of land use indicators and 9 types of urban traffic indicators were predicted. The preference constraints of the input index and output index of proposed model were calculated. Analysis result shows that the proposed model overcomes the limitation of too dependent on subjective intention in AHP, and makes up for the deficiency of ignoring the decision preference in DEA. The deviation between the results of land use and traffic coordination analysis and the actual situation reduces. The mutual coordination degrees of 6 blocks calculated by the original model are all more than 0.80, all of which are coordinated. After improvement, the mutual coordination degrees of 4 blocks are 0.32-0.57, all of which are incongruous. The mutual coordination degrees of the other 2 blocks are over 0.80, which is more consistent with the actual situation. 5 tabs, 4 figs, 26 refs.