Nonlinear Estimation

Some of  the fundamental questions our research seeks to answer are how to effectively estimate and propagate (quantify) the evolution of the uncertainty of a dynamical system. We are particularly interested in nonlinear and non-Gaussian systems and we investigate accurate and efficient approximations of the Fokker-Plank-Kolmogorov equation, as well as the Bayesian update rule. We research approximations using regression points, polynomials, sequential Monte Carlo methods, and Gaussian Mixture Models.

Back to Research