Exploration on the methods of weight calculation in the interval-based AHP

For comparing the accuracies of different methods of weight calculation in interval-based analytic hierarchy process (IAHP), five methods are studied. By comparison and analysis of the results, accuracy of these methods is evaluated and the principle of applying different methods to practical multi-attribute decision-making (MADM) is concluded. In case of off-time MADM, various algorithms can be invoked and the weights with the minimal error will be the final results. In addition, some improvements are made to increase the accuracey of the iteration algorithm. As to random simulation method, the relationship between accuracy and sample capacity and the rules of probability distribution are studied. In addition, the reason that the lower model (LM) algorithm has no solution in some cases is found; and a method to avoid this situation is also proposed. It has been shown that whether the LM has a solution is a new method to determine the consistence of a judgment matrix.