Credibility Assessment of Simulation Models using Flexible Mapping Functions

Simulation applications are used more and more for scheme evaluation, strategy formulation, problem solving, military training, and critical decision making in practice. The credibility of simulation models is crucial for the successful utilization of simulation results. In order to ease subject matter experts' repetitive work and reduces subjectivity in the credibility assessment of simulation models, a series of mapping functions are defined in this paper to transform discrepancy measures of crisp values or time series data into credibility scores in an automatic procedure. These mapping functions are predefined by only three adjustable parameters but can capture various applications and preferences. Based on these mapping functions, an adaptable, flexible and automatic approach for credibility assessment of simulation models is proposed. A didactic example for assessing the credibility of two ballistic missile simulation models with three indicators is presented to demonstrate the applicability of these mapping functions for credibility assessment.