Artificial Neural Network Based Adaptive Control of Single Phase Dual Active Bridge With Finite Time Disturbance Compensation

Single phase Dual Active Bridge (DAB) has found numerous applications in modern energy architectures such as direct current (DC) microgrid, electrical vehicle charging and high voltage direct current (HVDC) system. Due to the model complexities of DAB, this work proposes a model free adaptive contro...

Full description

Saved in:
Bibliographic Details
Published inIEEE access Vol. 7; pp. 112229 - 112239
Main Authors Farooq, Zaheer, Zaman, Taimur, Khan, Muhammad Amir, Nasimullah, Muyeen, S. M., Ibeas, Asier
Format Journal Article
LanguageEnglish
Published Piscataway IEEE 2019
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Adaptive control
Artificial neural network
Artificial neural networks
Bridge circuits
Compensation
Control methods
control system
Controllers
dc-dc converters
Direct current
Distributed generation
dual active bridge
Electric bridges
Electric converters
Energy storage
Launching pads
Microgrids
Neural networks
Nonlinear control
Proportional integral derivative
Renewable energy sources
Robust control
Sliding mode control
Switches
Topology
Online AccessGet full text

Cover

More Information
Summary:Single phase Dual Active Bridge (DAB) has found numerous applications in modern energy architectures such as direct current (DC) microgrid, electrical vehicle charging and high voltage direct current (HVDC) system. Due to the model complexities of DAB, this work proposes a model free adaptive control method based on artificial neural network (AANN) which is capable of adjusting the weights online in finite time. The finite time learning property of the proposed controller makes it perfectly robust for the compensation of the disturbances due to source and load side variations. A proportional integral (PI) controller is used to stabilize the nominal dynamics of the system along with the AANN controller. The structure of the proposed controller is as simple as PID controller and as robust as any nonlinear control method. The AANN-PI controller is implemented on TI Launchpad (TMS320F28379D) with a 50 Watts laboratory scale DAB test bench. Finally, the performance of the AANN-PI method is compared experimentally with classical PI and sliding mode controllers.
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2019.2934253