Parameter identification of Hammerstein-Wiener nonlinear systems with unknown time delay based on the linear variable...

Area of Science:

• Control Systems Engineering
• Nonlinear System Identification
• Computational Intelligence

Background:

• Hammerstein-Wiener (H-W) systems are widely used to model complex nonlinear processes.
• Accurate parameter estimation, including time delays, is crucial for effective system control and analysis.
• Existing methods often struggle with simultaneous estimation of parameters and unknown time delays in H-W systems.

Purpose of the Study:

• To develop an effective method for parameter estimation of Hammerstein-Wiener (H-W) nonlinear systems with unknown time delays.
• To address the challenge of simultaneously estimating parameters in linear and nonlinear submodules, along with the time delay.
• To evaluate the performance and accuracy of the proposed method through simulations and a practical application.

Main Methods:

• Formulation of a linear variable weight particle swarm optimization (PSO) algorithm tailored for time-delay systems.
• Transformation of the nonlinear system identification problem into a function optimization problem in the parameter space.
• Leveraging PSO's parallel searching capabilities and iterative identification techniques for simultaneous parameter and time delay estimation.

Main Results:

• The proposed method successfully separates parameters of the linear submodule, nonlinear submodule, and the time delay.
• Simulation examples demonstrate fast convergence speed and high estimation accuracy.
• The method was effectively applied to the identification of bed temperature systems, validating its practical utility.

Conclusions:

• The developed linear variable weight PSO method provides an accurate and efficient solution for parameter estimation in Hammerstein-Wiener systems with unknown time delays.
• The algorithm's ability to simultaneously estimate all parameters and the time delay offers a significant advantage over traditional approaches.
• This research contributes a valuable tool for the modeling and analysis of complex nonlinear systems in various engineering domains.