Page 73 - IJOCTA-15-3

FPGA design and implementation of fuzzy learning control: Application on DC motor position control
about ±0.012. This error is practically inferior to Acknowledgments
that obtained with the PID controller, as shown This paper is derived from a research grant funded
in Figure 18. by the Research, Development, and Innovation
The experimental results obtained show good
Authority (RDIA), Kingdom of Saudi Arabia,
position tracking and stability, with a lower er-
with grant number 13382-psu-2023- PSNU-R-3-
ror compared to the results of the PID controller.
1-EI-. This research is supported by Automated
These results are explained on the one hand by
Systems and Computing Lab (ASCL), Prince Sul-
the use of the knowledge base modifier, which
tan University, Riyadh, Saudi Arabia. The au-
modifies the membership functions in real time
thors would like to thank Prince Sultan Univer-
according to the DC motor dynamics, and on the
sity, Riyadh, Saudi Arabia for supporting this
other hand by the use of the Zedboard FPGA,
work.
which provides very short sampling times and par-
allel computation, accelerating the computation Fundings
time to quickly converge to the desired perfor-
This paper is funded by Prince Sultan University,
mance.
Riyadh, Saudi Arabia. The authors would like
6. Conclusion to thank Prince Sultan University for paying the
article processing fee for this paper.
This paper presents the implementation of a fuzzy
logic controller FMRLC on a Zedboard FPGA to Conflict of interest
control the position of a DC motor. The idea was The authors declare that have no conflict of in-
to implement a fuzzy controller with learning on terest.
an FPGA board. The use of this configurable cir-
cuit, which offers parallel data computation at a Author contributions
high operating frequency, aims at accelerating the
Conceptualization: Mohand Achour Touat,
computations, in particular the learning process,
which requires a significant computation time to Hocine Khati, Hand Talem, Ahmad Taher Azar
converge to the desired performance. The de- Formal analysis: Mohand Achour Touat, Arezki
Fekik, Rabah Mellah, Ahmad Taher Azar, Saim
sign and implementation of the controller were
Ahmed
carried out using the MATLAB-Simulink environ-
Methodology: Mohand Achour Touat, Hocine
ment. This methodology allows great design flex-
Khati, Arezki Fekik, Ahmad Taher Azar
ibility by using simple blocks on Simulink with-
Writing – original draft: Mohand Achour Touat,
out resorting to low-level programming languages
Hocine Khati, Arezki Fekik, Ahmad Taher Azar,
(VHDL/Verilog), thus reducing the design time of
Saim Ahmed
the algorithm. The fixed-point data type is used
to optimize the use of hardware resources on the Writing – review & editing: Arezki Fekik, Ahmad
FPGA. This binary representation can affect the Taher Azar, Rabah Mellah, Saim Ahmed
accuracy of calculations, but the design results in Availability of data
low power and resource consumption compared to
floating-point based designs. 34 The original contributions presented in the study
The FMRLC controller was first tested using are included in the article/supplementary mate-
the FIL technique in Simulink, then implemented rial, further inquiries can be directed to the cor-
on the Zedboard and applied to the experimental responding author/s.
device. The FIL simulation results showed good References
tracking performance and good robustness to dis-
turbances. The experimental results showed the 1. Kemal U, Beyza Nur A. Adaptive MIMO fuzzy
efficiency of the proposed controller compared to PID controller based on peak observer. Int J Op-
the conventional controller (PID), which can be tim Control: Theor Appl. 2023;13(2):139–150.
explained by the controller adapting the parame- 2. Masjudin, Alimuddin, Aisah SN, Wiryadinata R.
Dc motor speed control based on fuzzy adaptive
ters of the membership functions to each different
with fuzzy model reference learning control (fm-
situations. The design methodology used to im-
rlc) algorithm. In: 2020 2nd International Con-
plement the algorithm on the Zedboard FPGA
ference on Industrial Electrical and Electronics
board allowed to obtain a fast and accurate algo-
(ICIEE). IEEE; 2020: 79-83.
rithm with an optimal consumption of hardware 3. Devashish J, Arifa A, Sanatan K, Debanjan R.
resources. The co-simulation results and the ex- Fuzzy-PID and interpolation: a novel synergetic
perimental results demonstrated the effectiveness approach to process control. Int J Optim Control:
of this design methodology. Theor Appl. 2024;14(4):355-364.
445