Improvement Of Random Consistency Index In AHP And Its Application

The logic defect of AHP method is that the random consistency index of high order(15order or more)is not comprehensive,and the generalization ability of the traditional method of calculating the index is weak,as well as the theory.Based on these,in this paper we present an improved scheme in the big data background for obtaining the random consistency index in AHP consistency test.One is to design the sampling by constructing the uniform table,and then obtain the estimation of the population mean value;the other is analyzing the distribution of the disturbance matrix,and then obtaining the Monte Carlo sampling method which is consistent with the distribution of disturbance matrix,thereby a new random consistency index is obtained.Then conducting different numerical experiments to evaluate and test the proposed scheme in order to prove that Monte Carlo sampling method is more reasonable and effective.Finally,based on the new random consistency index,the improved AHP method is applied to the typhoon disaster analysis model of Hainan Island,combined with the fuzzy multiple attribute decision theory method,for obtaining the analysis conclusion of typhoon characteristics and temporal and spatial characteristics of Hainan Island.This analysis is not only proved the practicality of the AHP method,but also provide the theoretical basis for the typhoon disaster risk assessment in Hainan Island.