Gradient-Based Line-Search Optimization for Moving Horizon Estimation

This paper investigates the application of the gradient descent method in moving horizon state estimation for discrete-time nonlinear systems with line-search optimization based on a reduce number of iterations. Conditions guaranteeing the stability of the estimation error are established for single- and multi-iteration schemes to minimize a least-squares cost function based on the most recent batch of information. Numerical results demonstrate the effectiveness of the proposed approaches and highlight the enhanced performance through the combination of descent algorithms and line-search methods.