中文摘要：

针对空间数据属性值的不精确对评价因子总体排序的影响的问题,该文提出了ELECTRE TRI算法与GIS结合的耕地适宜性评价方法。该方法可以抵消空间数据属性值的不精确对评价因子总体排序的影响,使评价结果更客观真实。在ELECTRE TRI算法的两种分类方法中,悲观分类法的阈值比较按照从高到低排序容易造成属性值的不精确,影响排序结果,而乐观分类法能较好地避免出现类似情况。两种评价结果的评价单元数在三种相应等级下重合评价单元较多。评价结果表明评价结果较为相近,具有一定准确合理性。研究结论对区域耕地利用结构调整、划定基本农田保护区以及促进耕地差别化管理具有一定的参考价值。

英文摘要：

In view of the problem that inaccurate spatial data attribute values impact the overall evaluation factor order, this paper put forward a cultivated land suitability evaluation method based on ELECTRE TRI algorithm with GIS. It could offset the impact of inaccurate spatial data attribute values for the overall evalua- tion factor ordering, making the evaluation results more objective and real. In two classification methods of ELECTRE TRI algorithm, the threshold comparison sequence of pessimistic classification was from high to low, so that it was easy to cause the attribute value to be unaccurate, therefore affected the sorting re- sult. Optimistic classification could avoid the similar situation. And the more overlapping evaluation units showed that the evaluation results had certain rationality when these were closer. The results showed that this method could provide the basis for the cultivated land use structure adjustment, scientific planning the basic farmland protection area and promoting the land differentiated management.