l1 norm‐based recursive estimation for non‐linear systems with non‐Gaussian noises
Yuemei Qin,Jun Li,Shuying Li
DOI: https://doi.org/10.1049/cth2.12700
IF: 2.67
2024-06-12
IET Control Theory and Applications
Abstract:This study presents the state estimation problem of discrete‐time nonlinear stochastic system with non‐Gaussian noise. To further improve estimation accuracy, a recursive nonlinear smoother is put forward on the basis of minimizing l1 norm of model errors. The proposed l1 norm‐based filter and smoother is realized by using unscented transformation to implement statistical linear regression applying to non‐linear models. This study addresses the state estimation problem of discrete‐time non‐linear stochastic systems with non‐Gaussian noises, particularly impulsive noises. Instead of minimizing the mean square error of the state estimate, which tends to excessively focus on outliers caused by non‐Gaussian noises, the l1 norm‐based non‐linear recursive filter (L1KF) is put forward in this paper. Here, minimizing the l1 norm of model errors is actually to pursue the minimum sum of absolute values of all errors, which is equitable to all model errors rather than paying much attention on outliers. To further improve estimation accuracy, a recursive nonlinear smoother (L1KS) is proposed, based on minimizing the l1 norm of model errors. The proposed l1 norm‐based filter and smoother are implemented using unscented transformation for statistical linear regression applied to nonlinear models. Additionally, the computational complexity of the proposed method is analysed. Simulation results of tracking a radar target with impulsive noises demonstrate the effectiveness and robustness of the proposed estimator.
automation & control systems,engineering, electrical & electronic,instruments & instrumentation