Iterative parameter identification algorithms for the generalized time‐varying system with a measurable disturbance vector

Abstract

This article deals with the parameter identification of the generalized time-varying systems. The time-varying parameter vector can be expressed as a coefficient matrix multiplied by a measurable disturbance vector, the common identification methods cannot be used to estimate the parameters of the generalized time-varying systems directly. This motivates us to develop new iterative identification algorithms. The gradient-based iterative (GI) algorithm is proposed by means of the iterative technique. Moreover, a moving data window (MDW) is introduced, which can update the dynamical data by removing the oldest data and adding the newest measurement data. The MDW GI algorithm is proposed to improve the parameter estimation accuracy. The numerical simulation is provided and the simulation results show that the proposed identification methods are effective for the generalized time-varying systems.

Singular value decomposition based learning identification for linear time‐varying systems: From recursion to iteration

“…These time-varying systems are prevalent in various industries such as aerospace, integrated circuit manufacturing, and biomedical systems. [1][2][3][4][5] For example, the pure air emergency brake model for high-speed trains changes with time due to factors such as resistance friction, aerodynamic resistance, and nonlinear braking characteristics. 6 Although time-invariant models can characterize such systems over short intervals, the time-varying behavior must be considered for proper modeling and control in cases of fast variations.…”

“…The nonstationary characteristics of industrial processes are often caused by the time‐varying behavior of internal components. These time‐varying systems are prevalent in various industries such as aerospace, integrated circuit manufacturing, and biomedical systems 1‐5 . For example, the pure air emergency brake model for high‐speed trains changes with time due to factors such as resistance friction, aerodynamic resistance, and nonlinear braking characteristics 6 .…”

Data filtering‐based parameter estimation algorithms for a class of nonlinear systems with colored noises

“…24 Hierarchical identification methods can be used to decompose a large-scale system into several subsystems with small sizes and to enhance the computational efficiency. [25][26][27][28] The hierarchical identification principle is particularly suitable for solving large-scale nonlinear systems and multivariable systems with high dimensions. [29][30][31] Ding discussed the decomposition based hierarchical multi-innovation stochastic gradient identification method for Hammerstein system.…”

Parameter estimation for a class of time‐varying systems with the invariant matrix

“…To track the time‐varying parameters in rapidly changing environments, several adaptive algorithms have been proposed to identify fast‐varying systems 27‐29 . Approximating the time‐varying parameters by a weighted combination of a finite number of basis functions is the most common method 30 .…”

“…26 To track the time-varying parameters in rapidly changing environments, several adaptive algorithms have been proposed to identify fast-varying systems. [27][28][29] Approximating the time-varying parameters by a weighted combination of a finite number of basis functions is the most common method. 30 For example, Tsatsanis and Giannakis used a wavelet basis which can capture the characteristics of signals at different scales to describe the time-varying parameters and discussed how to choose the optimal wavelet basis for a given system trajectory.…”