Nonlinear simultaneous input and state estimation with application to flow field estimation

This paper studies the problem of simultaneous input and state estimation for a nonlinear dynamical system. A Bayesian paradigm is proposed to provide statistical derivation of joint input and state estimators. Using the Bayesian paradigm, a Maximum a Posteriori (MAP) based estimation scheme is developed as a joint estimator. The scheme involves nonlinear MAP optimization, which is addressed by a classical Gauss-Newton method. The effectiveness of the proposed scheme is illustrated via a simulation based study on ocean flow field estimation using submersible drogues that can measure position and acceleration intermittently.