A Proven Realtime Nonlinear Maximum Likelihood Estimation Technique

The solution to the nonlinear estimation problem has eluded realtime estimation practitioners over the past thirty years. The explanation for this is that they have adopted the concepts and mathematics of the Kalman filter, a memory efficient recursive realtime solution to only the linear estimation problem, and then proceeded to modify it to make it a nonlinear estimator. ISC has adopted the concepts and mathematics of the maximum likelihood (ML) estimator, a non realtime solution to the nonlinear estimation problem, and then proceeded to modify it to make it a recursive realtime estimator.According to studies performed on the HVWS BMD program, by replacing two Kalman filters with two realtime ML estimators the miss distance of the entire intercept system was reduced by a factor of two. According to studies performed on the THAD BMD program, by replacing the Kalman filter with a realtime ML estimator, not only were the radar resources required for intercept reduced by a factor of ten, but the ML estimator was more processing efficient. The key to making a realtime estimator is the mathematical concept and software design for the efficient propagation of the Hessian matrix, the inverse of the covariance matrix.