A Range Estimation on Priority Vectors in AHP
层次分析的区间估计

In this paper it is shown that due to limitation on information,knowledge and ability in decision-making and discontinuity of comparison scale,there are inevitably some uncertainty and deviation in pairwise comparison matrices Analytic Hiorarohy Process(AHP). Applying linear programming and restraint cone,we present a new model to conduct range estimation on priority weight vectors in AHP in this paper. In this model,Decision-maker's subjective preference is reflected in term,of preference restraint rather than a specific weight priority vector as in conventional models. Comparing with cited methods,the model is more robust and provides more information on decision makers' preference and uncertainty. Numerical example shows the effectiveness and validity to apply this model in subjective decision-making.