Sensitivity Analysis of a Combat Simulation Using Response Surface Methodology.
Abstract
This thesis provides a methodology to analyze the sensitivity of a combat simulation relative to the capabilities of the modeled weapon systems. Response Surface Methodology (RSM) was the primary technique used for analysis. A ground battle scenario was developed with a commercial PC wargame Empire II: The Art of War(TM). In the computer verses computer mode, the wargame was transformed into a combat simulation'. A screening experiment was used to determine which of an original 23 weapon factors were actually significant to this scenario. After analyzing the remaining five factors, only three ended up as truly significant to the battle outcome. The three factors are used in a regression equation which predicts the outcome of the overall simulation scenario. The sensitivity of the simulation outcome is reflected by the coefficients of each factor. The regression model was then validated against independent simulation trials. The method proved useful in analyzing the underlying behavior of a combat simulation. Combat analysts must isolate those weapon capabilities which have the greatest influence on the simulation model. Modeling funds can be concentrated on those factors. With accurate modeling and timely updates of those capabilities, the simulation model will earn more credibility as a decision aid.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1997
- Accession Number
- ADA324065
Entities
People
- William J. Berg
Organizations
- Air Force Institute of Technology