Adaptive Neural Network for a Stabilizing Shunt Active Power Filter in Distorted Weak Grids

Asadi, Yousef; Eskandari, Mohsen; Mansouri, Milad; Chaharmahali, Sajjad; Moradi, Mohammad Hassan; Tahriri, Mohammad Sajjad

doi:10.3390/app12168060

Cited by 11 publications

(9 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are numerous techniques to surmount the raised issues, we highlight PLL based on RST controllers [17], fuzzy logic [18], neural networks [19], or Adaline networks [20]. All these techniques fulfill a settlement between good dynamics and network voltage disturbances insensitivity.…”

Section: The Suggested Bpmvf Pllmentioning

confidence: 99%

Robust Voltage Vector-Controlled Three-Phase SAPF-based BPMVF and SVM for Power Quality Improvement

Essoussi,

Moutabir,

Bensassi

et al. 2024

IJRCS

View full text Add to dashboard Cite

The multiplication of nonlinear loads leads to significant degradation of the energy quality, thus the interconnection network is subject to being polluted by the generation of harmonic components and reactive power, which causes a weakening efficiency, especially for the power factor. In three-phase systems, they can cause imbalances by causing excessive currents at the neutral. This research treats the operation of robust voltage-oriented control (VOC) for a shunt active power filter (SAPF). The main benefit of this technique is to guarantee a decoupled control of the active and reactive input currents, as well as the input reference voltage. To sustain the DC voltage, a robust PI-structure-based antiwindup is inserted to ensure active power control. Besides, a robust phase-locked loop (PLL)-based bandpass multivariable filter (BPMVF) is used to improve the network voltage quality. Furthermore, a space vector modulation (SVM) is designed to replace the conventional one. A sinusoidal network current and unitary power factor are achieved with fewer harmonics. The harmonics have been reduced from 27.98% to 1.55% which respects the IEEE 519-1992 standard. Expanded simulation results obtained from the transient and steady-state have demonstrated the high performance of the suggested control scheme.

show abstract

Section: The Suggested Bpmvf Pllmentioning

confidence: 99%

Robust Voltage Vector-Controlled Three-Phase SAPF-based BPMVF and SVM for Power Quality Improvement

Essoussi,

Moutabir,

Bensassi

et al. 2024

IJRCS

View full text Add to dashboard Cite

show abstract

“…x , v Ã y , and v Ã , are estimated by considering sector (1) in Figure 8 and are represented as follows 125 :…”

Section: Switching Statesmentioning

confidence: 99%

“…A function named “ceil” rounds a real number up to the nearest whole number. The components of the expected voltage vector, including

v_{x}^{*}, v_{y}^{*}

, and

v^{*}

, are estimated by considering sector (1) in Figure 8 and are represented as follows 125 :

({, \begin{array}{l} v_{x}^{*} goodbreak= ∣ v^{*} ∣ ((), \cos ((), θ) goodbreak- \frac{1}{\sqrt{3}} \sin ((), θ)) \\ v_{y}^{*} goodbreak= \frac{\sqrt{3}}{2} ∣ v^{*} ∣ \sin ((), θ) \end{array})

italicif v_{x}^{*} < {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc} and v_{y}^{*} < {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc}, and ((), v_{x}^{*} goodbreak+ v_{y}^{*}) < {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc}; r_{1}

italicif v_{x}^{*} > {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc}; r_{2}

italicif v_{x}^{*} < {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc} and v_{y}^{*} < {0.5}^{*} \sqrt{\frac{2}{3}} . V_{dc}, and ((), v_{x}^{*} goodbreak+ v_{y}^{*}) > {0.5}^{*}

…”

Section: Pdpc‐svm Strategymentioning

confidence: 99%

Predictive direct power control with phase‐locked loop technique of three‐level neutral point clamped inverter based shunt active power filter for power quality improvement

Debdouche,

Benbouhenni,

Deffaf

et al. 2024

Circuit Theory & Apps

View full text Add to dashboard Cite

SummaryIn this study, an improved anti‐windup proportional‐integral predictive direct power control strategy with a phase‐locked loop (PLL) is proposed and applied to a three‐level shunt active power filter (SAPF) neutral point clamped (NPC) inverter using space vector modulation (SVM) technique. The SAPF has to ensure sinusoidal waveforms and unity power factor in the grid side, by eliminating harmonic currents and compensating reactive power of the nonlinear loads. Active and reactive powers as well as the DC bus voltage controls are performed using anti‐windup proportional and integral actions with a predictive algorithm, without a need for an exhaustive mathematical model. The control law is modulated using the SVM technique allowing fixed switching frequency and low power ripples. The performances of the suggested control have been verified through an electronic STM32F407 card under diverse and severe conditions. The obtained results confirmed the expected objectives in terms of current and voltage quality, robustness, and dynamic response.

show abstract

“…We can see the connection between these two in Figure 5. Yi [33] represents the value produced by the neuron, which is formed by first adding up all of its inputs and then passing that total through the activation function.…”

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

TLBO trained an ANN-based DG integrated Shunt Active Power Filter to Improve Power Quality

Venkata Anjani Kumar Gaddam,

Manubolu Damodar Reddy

2024

ARASET

View full text Add to dashboard Cite

In terms of power quality, the rising number of nonlinear loads in modern use has caused warning signs for power system and power engineering professionals. Every day, utilities have to deal with harmonic distortion caused by a growing number of non-linear power electronic equipment. To keep the system's power supply in good condition, a shunt active filter is used to filter out unwanted harmonics in the signal. This study presents a practical and low-cost method for reducing harmonics and enhancing distribution network power quality by means of the use of PV-integrated Shunt Active Power Filters (SAPF). With a teaching-learning-based optimized artificial neural network controller (TLBO-ANN) and the required DC power is extracted from the PV module. SAPF's TLBO-ANN algorithms are intended to increase system performance by reducing total harmonic distortion (THD). Here, the research work was performed in three stages to mitigate grid current harmonics. The first-stage SAPF system comprises a three-prong voltage source converter and uses DC power derived from photovoltaic panels. The P&O algorithm is used to get the maximum power out of a photovoltaic array. In the second stage, the BBO algorithm is used to fine-tune a conventional PI controller, resulting in values for and that increase the controller's performance. Furthermore, it is intended to use the BBO-PI controller's input and output values as training data for the ANN controller. This ANN controller is currently being tuned with the TLBO algorithm to find optimal values for the weight and bias parameters. In the third stage, the converter in PV-SAPF will inject the active power required by the load by using active current control theory, which means the inverter of SAPF is working like DG as well as the active power filter. Employing MATLAB simulations, we concluded that the proposed method is extremely adaptable and highly efficient in lowering harmonic currents that are brought on by non-linear loads.

show abstract

Adaptive Neural Network for a Stabilizing Shunt Active Power Filter in Distorted Weak Grids

Cited by 11 publications

References 32 publications

Robust Voltage Vector-Controlled Three-Phase SAPF-based BPMVF and SVM for Power Quality Improvement

Robust Voltage Vector-Controlled Three-Phase SAPF-based BPMVF and SVM for Power Quality Improvement

Predictive direct power control with phase‐locked loop technique of three‐level neutral point clamped inverter based shunt active power filter for power quality improvement

TLBO trained an ANN-based DG integrated Shunt Active Power Filter to Improve Power Quality

Contact Info

Product

Resources

About