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.

Data-Driven Moving Horizon Estimation Using Bayesian Optimization

State Estimation For Nonlinear Systems By Using Dynamic Optimization

Primal-dual Estimator Learning: an Offline Constrained Moving Horizon Estimation Method with Feasibility and Near-optimality Guarantees

Adaptive Horizon Size Moving Horizon Estimation with Unknown Noise Statistical Properties

A Moving Horizon Estimation Approach To Constrained Linear System With Uncertain Model

Moving horizon estimation for nonlinear systems with time-varying parameters

Moving Horizon Estimation for Stochastic Descriptor Systems with Inaccurate Noise Covariance Matrices

Moving Horizon Estimation for Stochastic Systems with Unknown Inputs

Moving horizon estimation for stochastic systems with missing measurements

Primal-Dual Estimator Learning Method with Feasibility and Near-Optimality Guarantees

Constrained Nonlinear State Estimation --- A Differential Evolution Based Moving Horizon Approach

Nonlinear moving horizon estimation for robust state and parameter estimation - extended version

Transition Moving Horizon Estimation Using Multiple Linear Models

Robust Bayesian Inference for Moving Horizon Estimation

Back-and-forth nudging moving horizon estimation for discrete-time linear systems

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

Robust Moving Horizon Estimation for System with Uncertain Measurement Output

Parallelizable Parametric Nonlinear System Identification via tuning of a Moving Horizon State Estimator

Robust Moving Horizon Estimation for Constrained Linear System with Uncertainties

A Study on Robust Estimation Via Moving Horizon Strategy

Self-tuning moving horizon estimation of nonlinear systems via physics-informed machine learning Koopman modeling