基于改进分层激励控制线的多阶段信息集结方法

On Multi-phase Information Aggregation Methods Based on Improved Stratified Incentive Control Lines

针对时序动态综合评价问题,在分层激励方法的基础上,提出了3种改进的分层激励多阶段信息集结方法,即按比例分层的集结方法、按一维聚类分层的集结方法和按诱导变量分层的集结方法,并对其分层模式及信息集结过程进行了分析.改进后的分层方法对被评价对象评价值中包含的隐含信息的分析更为深入,且能够灵活地凸显决策者的激励意图.最后通过一个算例对方法的有效性进行了验证.在实际应用中,决策者可以根据实际问题选择适合的改进方法.

Aiming at sequential dynamic comprehensive evaluation, based on multi-phase evaluation information aggregation,three improved stratified incentive methods of multi-phase information aggregation were proposed,i. e.,the aggregation methods of stratifying according to proportion,of stratifying according to one dimensional clustering and of stratifying according to induced variables. Then their stratified models and the processes of information aggregation were analyzed. The improved methods can analyze the alternatives' implicit information much more thoroughly,and protrude the inspiriting intention of decision makers flexibly. Finally,an example was used to testify the validity of these methods. In practical applications,decision-makers can select appropriate methods according to the needs of specific problems.