Estimating Parameters of Nonlinear Systems Using the Elitist Particle Filter Based on Evolutionary Strategies

Abstract

In this article, we present the elitist particle filter based on evolutionary strategies (EPFES) as an efficient approach for nonlinear system identification. The EPFES is derived from the frequently-employed state-space model, where the relevant information of the nonlinear system is captured by an unknown state vector. Similar to classical particle filtering, the EPFES consists of a set of particles and respective weights which represent different realizations of the latent state vector and their likelihood of being the solution of the optimization problem. As main innovation, the EPFES includes an evolutionary elitistparticle selection which combines long-term information with instantaneous sampling from an approximated continuous posterior distribution. In this article, we propose two advancements of the previously-published elitist-particle selection process. Further, the EPFES is shown to be a generalization of the widely-used Gaussian particle filter and thus evaluated with respect to the latter for two completely different scenarios: First, we consider the so-called univariate nonstationary growth model with timevariant latent state variable, where the evolutionary selection of elitist particles is evaluated for non-recursively calculated particle weights. Second, the problem of nonlinear acoustic echo cancellation is addressed in a simulated scenario with speech as input signal: By using long-term fitness measures, we highlight the efficacy of the well-generalizing EPFES in estimating the nonlinear system even for large search spaces. Finally, we illustrate similarities between the EPFES and evolutionary algorithms to outline future improvements by fusing the achievements of both fields of research.

An Intelligent Optimization-Based Particle Filter for Fault Diagnosis

“…It has become an effective method to study optimal estimation problems of non-linear and non-Gaussian dynamic systems, and has been widely used in the fields of machine vision, navigation, parameter estimation, target tracking, state monitoring, and fault diagnosis, etc. [20][21][22][23].…”

“…where in Equations ( 16) to (18), RS = 0.0071 Ω and RR = 0.005 Ω represent the stator and rotor per phase resistance respectively; LS = 0.171 H and LR = 0.159 H represent the cyclic stator and rotor inductances respectively; Constant ωR is the mechanical rotor frequency (speed); np = 2 is the number of pole pairs. The mutual inductance is written as Lm equals to 2.9 H and σ is defined by Equations (20).…”

Semi-Blind Source Separation for Nonlinear Acoustic Echo Cancellation

“…However, the nonlinear distortion in many practical applications is not negligible, causing significant deterioration of the linear AEC 3,4 . Various nonlinear AEC (NAEC) methods have been proposed based on Volterra filters 5,6 , function link adaptive filters 7 , particle filters 8,9 , state-space frequency-domain adaptive filters [10][11][12] and kernelized adaptive filters 13 . These methods are designed by combining the adaptive filter with a pre-assumed nonlinear model, which often mismatches the actual nonlinear model and thus challenges the behavior of the adaptive filter.…”

Mobile robot localization based on PSO estimator

“…Although, the performance of this algorithm is superior to PF standard, its computation is expensive. In , PF is combined with an evolutionary strategy to improve the estimation accuracy of PF. A main innovation of this paper is an evolutionary elitist‐particle selection scheme that combines long‐term information with instantaneous sampling from an approximated continuous posterior distribution.…”