Abstract:The accuracy of moving horizon estimation (MHE) suffers significantly in the presence of measurement outliers. Existing methods address this issue by treating measurements leading to large MHE cost function values as outliers, which are subsequently discarded. This strategy, achieved through solving combinatorial optimization problems, is confined to linear systems to guarantee computational tractability and stability. Contrasting these heuristic solutions, our work reexamines MHE from a Bayesian perspective, unveils the fundamental issue of its lack of robustness: MHE's sensitivity to outliers results from its reliance on the Kullback-Leibler (KL) divergence, where both outliers and inliers are equally considered. To tackle this problem, we propose a robust Bayesian inference framework for MHE, integrating a robust divergence measure to reduce the impact of outliers. In particular, the proposed approach prioritizes the fitting of uncontaminated data and lowers the weight of contaminated ones, instead of directly discarding all potentially contaminated measurements, which may lead to undesirable removal of uncontaminated data. A tuning parameter is incorporated into the framework to adjust the robustness degree to outliers. Notably, the classical MHE can be interpreted as a special case of the proposed approach as the parameter converges to zero. In addition, our method involves only minor modification to the classical MHE stage cost, thus avoiding the high computational complexity associated with previous outlier-robust methods and inherently suitable for nonlinear systems. Most importantly, our method provides robustness and stability guarantees, which are often missing in other outlier-robust Bayes filters. The effectiveness of the proposed method is demonstrated on simulations subject to outliers following different distributions, as well as on physical experiment data.

Moving-Horizon Estimators for Hyperbolic and Parabolic PDEs in 1-D

Iteratively Preconditioned Gradient-Descent Approach for Moving Horizon Estimation Problems

Observer-based Stabilization of an Unstable Parabolic PDE Using the Pseudospectral Method and Sturm-Liouville Theory

Backstepping Synthesis for Feedback Control of First-Order Hyperbolic PDEs with Spatial-Temporal Actuation

Multiple Window Moving Horizon Estimation

Event-triggered moving horizon estimation for nonlinear systems

Robust stability of moving horizon estimation for continuous-time systems

MHE under parametric uncertainty -- Robust state estimation without informative data

Generalized Moving Horizon Estimation for Nonlinear Systems with Robustness to Measurement Outliers.

Sub-Optimal Moving Horizon Estimation in Feedback Control of Linear Constrained Systems

Robust Bayesian Inference for Moving Horizon Estimation

Moving horizon estimation with multirate measurements and correlated noises

Robust Data-Driven Moving Horizon Estimation for Linear Discrete-Time Systems

Gaussian Process-Based Nonlinear Moving Horizon Estimation

A Lyapunov function for robust stability of moving horizon estimation

Moving Horizon Estimation With Unknown Inputs Under Dynamic Quantization Effects

Finite‐dimensional, output‐predictor‐based, adaptive observer for heat PDEs with sensor delay

Iterative distributed moving horizon estimation of linear systems with penalties on both system disturbances and noise

$H_2$-Optimal Estimation of a Class of Linear PDE Systems using Partial Integral Equations

Robust Stability of Gaussian Process Based Moving Horizon Estimation