Adaptive Control Based on LMS Algorithm for Grid-Connected Inverters

Abstract

The grid-connected inverter is the key component for the reliable and safe operation of grid-interconnected renewable energy systems. In order to ensure that the inverter output current is in-phase synchronized with the grid voltage, a highly efficient and fast control strategy is required. This paper proposes an algorithmically driven approach for designing the adaptive controller for a grid-connected DC/AC inverter. The controller consists of two adaptive IIR filters based on the LMS algorithm and has two modes of operation. During the learning mode, the controller uses the filters to estimate the coefficients of the inverse transfer function of the PWM-driven inverter at several grid frequencies within a specific range. During the online mode, the controller generates the driving signal for the inverter block using a set of learned coefficients corresponding to the current grid frequency. By constantly monitoring the grid voltage and reconfiguring the controller accordingly, a fast adaptation of the inverter output current to grid frequency variations is achieved. For grid frequency variations in the range of 48 to 52 Hz, the MATLAB simulation results show that the phase difference between the grid voltage and inverter output current waveforms becomes less than 1 degree after at most four periods.

Keywords:

• DC/AC inverter
• pulse width modulation (PWM)
• adaptive filter
• system identification
• inverse modeling

Additional information

Funding

This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia.

Notes on contributors

Goran S. Nikolić

Goran S. Nikolić received his BS degree in communication engineering, and his MS and PhD degrees in electronic engineering from the Faculty of Electronic Engineering, University of Nis, Serbia, in 2003, 2010, and 2019, respectively. He is an assistant professor with the Department of Electrical Engineering at the Faculty of Electronic Engineering, University of Nis, Serbia. His research interests include fault-tolerant and low-power embedded system design, power electronics and wireless sensor networks.

Tatjana R. Nikolić

Tatjana R. Nikolić received her BS degree in communication engineering, and her MS and PhD degrees in electronic engineering from the Faculty of Electronic Engineering, University of Nis, Serbia, in 2000, 2005, and 2010, respectively. She is currently a full professor with the Department of Electrical Engineering at the Faculty of Electronic Engineering, University of Nis, Serbia. Her research interests include embedded systems, fault-tolerant on-chip communications, low-power system-on-chip design, and reconfigurable hardware architectures for application-specific acceleration.

Goran Lj. Djordjević

Goran Lj. Djordjević received his BS degree in computer science, and his MS and PhD degrees in electronic engineering from the Faculty of Electronic Engineering, University of Nis, Serbia, in 1989, 1994, and 1998, respectively. He is a full professor with the Department of Electrical Engineering at the Faculty of Electronic Engineering, University of Nis, Serbia. He has been involved in several research projects, focusing on parallel and distributed computing, reconfigurable computing, and wireless sensor networks. His current research interests include wireless sensor networks, embedded systems, networks-on-chip, and reconfigurable system-on-chip design.