FPGA implementation of NN based LMS–LMF control algorithm in DSTATCOM for power quality improvement

• Published in: Control engineering practice Vol. 98; p. 104378
• Main Authors: Malathi, S., Jayachandran, J.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.05.2020
• Subjects: Field programmable gate array; Neural network; Power quality; Shunt active power filter; Total harmonic distortion; Total harmonic distortion; Field programmable gate array; Neural network; Power quality; Shunt active power filter
• ISSN: 0967-0661; 1873-6939
• DOI: 10.1016/j.conengprac.2020.104378

Neural Network (NN) based Least Mean Square (LMS)–Least Mean Fourth (LMF) control technique is developed and implemented in Field Programmable Gate Array (FPGA) for Shunt Active Power Filter (SAPF) to enhance Power Quality (PQ) in three-phase four wire distribution system. The PQ enhancement comprises the improvement of power factor, mitigation of harmonics and compensation of reactive power. The proposed control strategy is simulated in MATLAB/Simulink to determine the reference source currents and weights of active and reactive components. The simulation results of LMS–LMF control are compared with the other three control strategies. The NN control is developed as individual modules in FPGA. Thus it can be configured to any other research area which has identical requirements. The FPGA implementation overcomes the drawback of conventional controllers with a fast and accurate response. The FPGA based proposed control strategy for SAPF is developed as a laboratory prototype and the results obtained are analyzed. The results attained adhere to power quality standards. •Neural network based LMS–LMF control approach for DSTATCOM.•FPGA hardware implementation of the proposed control algorithm.•Reduction in THD, compensation of reactive power and unity power factor for varying source and load conditions.•A comparative study carried out reveals the superior performance of the proposed controller.•The simulation and experimental results of DSTATCOM with the proposed control technique comply with IEEE 519 standards.