A modified ANP and its application in simulation credibility evaluation

Abstract

With the size and complexity of simulation applications growing rapidly, many simulation systems cannot be structured hierarchically, involving the interaction and dependence of elements in a hierarchy. To give an efficient credibility evaluation for the intricate simulation systems involving network configuration, this paper proposes a modified fuzzy ANP method to calculate weights of indices for assessing. The method used triangle fuzzy numbers to establish judgment matrix, together with confidence level, expressing Subject Matter Experts' degree of understanding with respect to the importance of each component. Furthermore, a new possibility measurement for fuzzy numbers is proposed, to obtain the relative importance ranking of components for the simplified judgment in dealing with the vagueness defined measurement scales. The approach is used to calculate the credibility of a control and guide simulation system for a type of missile. The application result shows that the method is reasonable, easy and feasible.

Exploring the Context and Practices of Expert Simulation Modellers

“…Other aspects of evaluation such as quality, usability, utility, performance, documentation and maintainability are not usually considered by our participants. Simulation modelling literature also shows that validation and verification is the main focus in model evaluation for the models developed for business and industry [24]. However, other aspects of model evaluation seem to be taken care of in large scale defence simulation [6].…”