A A Particle Swarm Optimization Algorithm Based System Identification of AC-DC Power System Including a Three-Phase Controlled Rectifier

Abstract

The objective of research are 1) to study the mathematical model of a controlled three-phase rectifier 2) to determine the parameters of the controlled three-phase rectifier using the Particle Swarm Optimization (PSO) technique and 3) to develop a test set for identifying the parameters of the controlled rectifier. This paper presents the application of the artificial intelligence techniques to identify the AC-DC power system parameters. In such systems, DC-link filters are often included. For certain applications, precise parameter values of the filter circuit are crucial for system analysis and design. The tools used in this research include the development of a test set for the controlled three-phase rectifier. The system adjusts the firing angle of the thyristor at 0, 10, and 30 degrees, respectively. Artificial intelligence methods are employed to identify the system parameters, resulting in parameters of R_eq = 0.0864 Ω, L_eq = 0.1277 H, C_dc = 235 µF, and r_c = 2.88 Ω. The findings indicate that artificial intelligence methods can determine parameters close to the actual values, as shown by the objective function value W = 1.4535.

The results reveal that parameter identification using artificial intelligence demonstrates that the obtained parameters effectively respond in transient and steady-state conditions, closely matching real system testing results. Moreover, the resulting parameters are very useful for a stability analysis of the power electronic systems due to a constant power load.