A study on finite-time particle swarm optimization as a system identification method

Author(s):  
Manuel A. Fernández ◽  
Jen-Yuan Chang
Author(s):  
Manuel A. Fernández ◽  
Jen-Yuan (James) Chang

Abstract This paper presents a comparison between different system identification techniques, namely Least Squared Estimation, Total Least Squares, Linear Sequential Estimation, the Gauss-Newton method, and Particle Swarm Optimization. A DC motor model was simulated in Simulink, with arbitrarily selected parameters, and the input and output values were used to test the effectiveness of these system identification techniques.


2013 ◽  
Vol 448-453 ◽  
pp. 2511-2515
Author(s):  
Wen Sun ◽  
Xiang Yu Kong ◽  
Qun Yang ◽  
Fang Zhang

A parameter identification method for generator speed governor system, which combines decoupling parameter identification and overall recognition with measured data, was proposed in the paper. The method bases on particle swarm optimization, and takes parameter identification as a parameters optimization problem under evaluation function. According to an intelligent optimization algorithms evolutionary strategy, the individual's status is continuously adjusted until the identification system and actual system output error is sufficiently small. Case studies show that the proposed method can be applied to the measured parameters and model validation work.


Sign in / Sign up

Export Citation Format

Share Document