Wind Turbine Driving a PM Synchronous Generator Using Novel Recurrent Chebyshev Neural Network Control with the Ideal Learning Rate

Lin, Chih-Hong

doi:10.3390/en9060441

Cited by 5 publications

(4 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While most of the recently published studies using neural models to predict multivariate wind turbine time series employ LSTM, there are also several alternative approaches focusing on other RNN variants. For example, there are several papers on the use of Elman neural networks in forecasting multivariate wind turbine data [23,24]. Kramti et al [20] also applied Elman neural networks, but with a slightly modified architecture.…”

Section: Forecasting Models With Recurrent Neural Networkmentioning

confidence: 99%

Online learning of windmill time series using Long Short-term Cognitive Networks

Morales-Hernández¹,

Nápoles²,

Jastrzębska³

et al. 2021

Preprint

View full text Add to dashboard Cite

Forecasting windmill time series is often the basis of other processes such as anomaly detection, health monitoring or maintenance scheduling. The amount of data generated on windmill farms makes online learning the most viable strategy to follow. Such settings require retraining the model each time a new batch of data is available. However, update the model with the new information is often very expensive to perform using traditional Recurrent Neural Networks (RNNs). In this paper, we use Long Short-term Cognitive Networks (LSTCNs) to forecast windmill time series in online settings. These recently introduced neural systems consist of chained Short-term Cognitive Network blocks, each processing a temporal data chunk. The learning algorithms of these blocks is based on a very fast, deterministic learning rule that makes LSTCNs suitable for online learning tasks. The numerical simulations using a case study with four windmills showed that our approach reported the lowest forecasting errors with respect to a simple RNN, a Long Short-term Memory, a Gated Recurrent Unit and a Hidden Markov Model. What is perhaps more important is that the LSTCN approach is significantly faster than these state-of-the-art models.

show abstract

Section: Forecasting Models With Recurrent Neural Networkmentioning

confidence: 99%

Online learning of windmill time series using Long Short-term Cognitive Networks

Morales-Hernández¹,

Nápoles²,

Jastrzębska³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…This output is used to operate the converter at the maximum power point (MPP). The authors of [22,23] show sensorless MPPT algorithms based on ANN techniques. The inputs are the rotor speed and the output power of the turbine while the output is the optimal rotor speed.…”

Section: Related Workmentioning

confidence: 99%

Limited Power Point Tracking for a Small-Scale Wind Turbine Intended to Be Integrated in a DC Microgrid

Aourir

Locment

2020

Applied Sciences

View full text Add to dashboard Cite

Limited power point tracking (LPPT) is emerging as a new technology for power management controllers for small-scale wind turbines (SSWTs) thanks to its advantages in terms of operation flexibility, economy and system security. LPPT operates in such a way that power requested by the user can be extracted from the wind turbine while respecting constraints. However, operating in LPPT mode still requires a deep understanding to obtain a compromise between minimizing power oscillations and transient response. For that, three LPPT power control strategies for an SSWT intended to be integrated in a direct current (DC) urban microgrid are investigated. These methods concern perturb and observe (P&O) with fixed step size, P&O based on Newton’s method and P&O based on the fuzzy logic (FL) technique. The experimental results highlight that all methods function correctly and reach the limited power point (LPP). The FL method improves dynamic performances with more steady oscillations around LPP compared to fixed step size and Newton’s methods. The sudden variation of wind velocity and power lead us to conclude that the FL method ensures a good balance between reducing oscillation of wind turbine (WT) output power around the operating point and convergence of rising time toward LPP.

show abstract

“…3e sc 3L+2L 1 (13) Equation (13) shows that the derivative of three-phase stator current is only related to the three-phase stator EMF in mode 1, when the system parameters are determined.…”

Section: Analysis Of Input Power Factormentioning

confidence: 99%

“…Furthermore, with the ability to handle the shoot through states, the reliability of ZSI is substantially improved. Moreover, the output voltage distortion of inverter is reduced, because there is no need to set dead time [12][13][14][15][16][17][18]. To improve on the original ZSI, the quasi-Z-source inverter (q-ZSI) has been developed which features several improvements, such as continuous input current and no need input capacitance.…”

Section: Introductionmentioning

confidence: 99%

Stator Current Harmonic Reduction in a Novel Half Quasi-Z-Source Wind Power Generation System

Huang

Zhang

2016

Energies

View full text Add to dashboard Cite

Abstract:The generator stator current gets distorted with unacceptable levels of total harmonic distortion (THD) because impedance-source wind power generation systems use three-phase diode rectifiers. The stator current harmonics will cause increasing losses and torque ripple, which reduce the efficiency and stability of the system. This paper proposes a novel half quasi-Z-source inverter (H-qZSI) for grid-connected wind power generation systems, which can reduce the generator stator current harmonics a great deal. When H-qZSI operates in the shoot-through zero state, the derivative of the generator stator current is only determined by the instantaneous value of the generator stator voltage, so the nonlinear relationship between generator stator current and stator voltage is improved compared with the traditional impedance-source inverter. Theoretically, it is indicated that the stator current harmonics can be reduced effectively by means of the proposed H-qZSI. Finally, simulation and experimental results are given to verify the theoretical analysis.

show abstract

Wind Turbine Driving a PM Synchronous Generator Using Novel Recurrent Chebyshev Neural Network Control with the Ideal Learning Rate

Cited by 5 publications

References 32 publications

Online learning of windmill time series using Long Short-term Cognitive Networks

Online learning of windmill time series using Long Short-term Cognitive Networks

Limited Power Point Tracking for a Small-Scale Wind Turbine Intended to Be Integrated in a DC Microgrid

Stator Current Harmonic Reduction in a Novel Half Quasi-Z-Source Wind Power Generation System

Contact Info

Product

Resources

About