Abstract:Focusing on identification, this paper develops a class of convex optimization-based criteria and correspondingly the recursive algorithms to estimate the parameter vector $\theta^{*}$ of a stochastic dynamic system. Not only do the criteria include the classical least-squares estimator but also the $L_l=|\cdot|^l, l\geq 1$, the Huber, the Log-cosh, and the Quantile costs as special cases. First, we prove that the minimizers of the convex optimization-based criteria converge to $\theta^{*}$ with probability one. Second, the recursive algorithms are proposed to find the estimates, which minimize the convex optimization-based criteria, and it is shown that these estimates also converge to the true parameter vector with probability one. Numerical examples are given, justifying the performance of the proposed algorithms including the strong consistency of the estimates, the robustness against outliers in the observations, and higher efficiency in online computation compared with the kernel-based regularization method due to the recursive nature.

Identification Methods for Two-Variable Difference Systems

Automated Identification of Differential-Variational Equations for Static Systems

Modified stochastic gradient parameter estimation algorithms for a nonlinear two-variable difference system

Comparison of Three Classes of Identification Methods

Hierarchical Least-Squares Based Iterative Identification for Multivariable Systems with Moving Average Noises

Rapid Identification of Switched Systems: A Data-Driven Method in Variational Framework

On Two Methods for Identifying Dynamic Errors-in-Variables Systems

Expectation Maximization Identification Algorithm for Time-Delay Two-Dimensional Systems

Two-stage Gradient-based Iterative Estimation Methods for Controlled Autoregressive Systems Using the Measurement Data

System identification under non-negativity constraints

Multivariable CAR-like System Identification with Multi-innovation Gradient and Least Squares Algorithms

Delay System Identification Using Global Optimization Methods

Two‐stage gradient‐based iterative algorithms for the fractional‐order nonlinear systems by using the hierarchical identification principle

A new identification method for dual-rate Hammerstein systems

Research on Two Kinds of Parameter Identification Methods for Short-Term Time-Varying System

Joint Estimation And Identification For Stochastic Systems With Two Kinds Of Unknown Inputs

Identification of dual-rate systems based on finite impulse response models

Stochastic Gradient Based Iterative Identification Algorithm for a Class of Dual-Rate Wiener Systems

A Class of Convex Optimization-Based Recursive Algorithms for Identification of Stochastic Systems

Least-Squares Based and Gradient Based Iterative Parameter Estimation Algorithms for a Class of Linear-in-Parameters Multiple-Input Single-Output Output Error Systems

Recursive least squares estimation algorithm applied to a class of linear-in-parameters output error moving average systems.