Ballistic Trajectory Tracking Using Constrained Estimation

Nonlinear regression augmented with a constrained intercept parameter is investigated and a new nonlinear estimation algorithm is developed. The batch estimation process entails careful modelling of the nonlinear measurement situation, inclusion of the intercept parameter to address the truncation effects of linearization, and incorporation of a constraint akin to the ridge regression concept from statistics that balances linearization induced truncation error with measurement noise induced equation error. Simulations show consistent improved estimation performance over conventional iterative least squares estimation without an increase in the estimation error covariance. In addition, the process expands the measurement geometry envelope where good estimation is achieved