Open AccessOptimal tuning of PI controller using system identification for two-phase boost converter for low-voltage applications
Author(s) -
M. A. N. Amran,
Afarulrazi Abu Bakar,
Mohd Hafiz A. Jalil,
Ahmet Ganı,
Erum Pathan
Publication year - 2021
Publication title -
international journal of power electronics and drive systems/international journal of electrical and computer engineering
Language(s) - English
Resource type - Journals
eISSN - 2722-2578
pISSN - 2722-256X
DOI - 10.11591/ijpeds.v12.i4.pp2393-2402
Subject(s) - span (engineering) , control theory (sociology) , controller (irrigation) , duty cycle , transfer function , voltage , mathematics , physics , computer science , engineering , electrical engineering , control (management) , agronomy , civil engineering , quantum mechanics , artificial intelligence , biology
This paper presents modeling and hardware implementations of a two-phase DC-DC boost converter by using the system identification approach. The main objective of this research was to study new methods to obtain the values of the constants for the proportional-integral (PI) controller. Existing methods are time-consuming, since the values of the constants for the PI controller need to be calculated. The system identification approach for the closed-loop boost converter saves more time. This method has the fastest technique to find constants K pand K ifor the closed-loop two-phase boost converter. To model a two-phase boost converter using the system identification approach, input duty cycle and output voltage are collected in the time domain data. In this study, the transfer function (TF) model, the autoregressive moving average with exogenous (ARMAX) model and the output-error (OE) model were used to generate a mathematical model. To perform the closed-loop analysis, constants K pand K iwere obtained based on the generated mathematical model from the system identification approach. The result from the experiment shows that the percentages of overshoot for the TF, ARMAX and OE models were 19%, 25.36% and 24.6%, respectively. The output voltage ripples obtained for all three models were less than 5% of output voltage.