Bayesian Network For Assessing EKV Lethality

"The Ground-Based Midcourse (GMD) Defense system must operate within demanding requirements - our lives depend upon its success. The end-game of the engagement sequence is planned to execute flawlessly to achieve hit-to-kill lethality. Many threat,system, and environmental parameters influence the success or failure of the engagement, e.g. aspect angle between the threat and kill vehicle at beginning of end-game. Evaluating the complex relationship between the influencing factors and probability ofkill is "success critical". The product of this SBIR is a Bayesian Belief Network (BBN) model of the GMD end-game. The analysis achievable through the BBN approach is a risk mitigator. The use of test and simulation in conjunction with our powerfulmachine learning algorithms define the complex relationship among the factors. The BBN is both a predictive model and a diagnostic tool wherein the conditions that yield a high probability of kill can be readily identified. The BBN is both an analysisaid to test planning (identifies where additional data is needed) and decision aid for the system battle manager (plans optimal engagement parameters to maximize the opportunity for successful intercept). The EKV Lethality BBN approach offers riskmitigation to the government's Missile Defense Agency and their hardware developers in three critical GMD system engineering problem areas: Enhanced GMD performance verification by ability to derive m