An Examination of Circular Error Probable Approximation Techniques
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Pagination or Media Count:
Several approximation techniques are currently used to estimate CEP. These techniques are statistical in nature, based on the bivariate normal distribution of the crossrange and downrange miss distances of sample impacts of weapon systems. This thesis examines four of the most widely used approximation techniques Grubbs-Patnaikchi-square, Grubbs-Paternik Wilson-Hilferty, modified RAND-234, and correlated bivariate normal, compares their results with the results and computational effort required by established numerical integration techniques, determines the relative accuracy of each technique in various regimes of the biasellipticity parameter space. Included in this report is a tutorial on the subject of CEP meant to serve as a general introduction to how to calculate CEPs with some of the popular approximation techniques. In general it was found that each of the approximation techniques is best in some regime of the parameter space with the Grubbs-Patnaikchi-square technique being the most reliable estimator. For fast calculations of CEP, the correlated bivariate normal and the exact method may not be feasible because both are computationally rigorous and require from 2 minutes to several hours of computer time on a personal computer to give an estimate of CEP.
- Statistics and Probability
- Fire Control and Bombing Systems