Prediction of power generation and rotor angular speed of a small wind turbine equipped to a controllable duct using artificial neural network and multiple linear regression

Siavash, Nemat Keramat; Ghobadian, Barat; Najafi, Gholamhassan; Rohani, Abbas; Tavakoli, T.; Mahmoodi, Esmail; Mamat, Rizalman; Mohamed, Mazlan

doi:10.1016/j.envres.2020.110434

Cited by 32 publications

(17 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The MLP model has a better performance than the regression model. Siavash et al [28] predict turbine performance using multiple linear regression and a neural network considering as many as 4 channel opening angles as response variables. The performance of the neural network model is more satisfactory than the multiple regression model.…”

Section: Related Workmentioning

confidence: 99%

Modeling Multioutput Response Uses Ridge Regression and MLP Neural Network with Tuning Hyperparameter through Cross Validation

Nugroho¹,

Handoyo²,

Hsieh³

et al. 2022

IJACSA

View full text Add to dashboard Cite

The multiple regression model is very popular among researchers in both field of social and science because it is easy to interpret and have a well-established theoretical framework. However, the multioutput multiple regression model is actually widely applied in the engineering field because in the industrial world there are many systems with multiple outputs. The ridge regression model and the Multi-Layer Perceptron (MLP) neural network model are representations of the predictive linear regression model and predictive non-linear regression model that are widely applied in the world of practice. This study aims to build multi-output models of a ridge regression model and an MLP neural network whose hyperparameters are determined by a grid search algorithm through the cross-validation method. The hyperparameter that produces the smallest RMSE value in the validation data is chosen as the hyperparameter to train both models on the training data. The hyperparameter in question is a combination of learning algorithms and alpha values (ridge regression), a combination of the number of hidden nodes and gamma values (MLP neural network). In the ridge regression model for alpha in the range between 0.1 and 0.7, the smallest RMSE is obtained for all learning algorithms used. While the MLP neural network model specifically obtained a combination of the number of nodes = 18 and gamma = 0.1 which produces the smallest RMSE. The ridge regression model with selected hyperparameters has better performance (in the RMSE and R2 value) than the MLP neural network model with selected hyperparameters, both on training and testing data.

show abstract

Section: Related Workmentioning

confidence: 99%

Modeling Multioutput Response Uses Ridge Regression and MLP Neural Network with Tuning Hyperparameter through Cross Validation

Nugroho¹,

Handoyo²,

Hsieh³

et al. 2022

IJACSA

View full text Add to dashboard Cite

show abstract

“…In this article, the authors used three performance functions to analyze the target and output accuracy. These functions include the correlation coefficient R, the root mean squared error (RMSE), and the coefficient of determination

R^{2}

, 52–55 which are defined as follows:

R = \frac{y_{o} (i)}{y_{e} (i)}

RMSE = \sqrt{\frac{\sum_{i = 1}^{n} {(y_{e} (i) - y_{o} (i))}^{2}}{n}}

R^{2} = {([], \frac{\sum_{i = 1}^{n} (y_{e} (i) - \overset{true¯}{y_{e}}) (y_{o} (i) - \overset{true¯}{y_{0}})}{\sum_{i = 1}^{n} (y_{e} (i) - \overset{true¯}{y_{e}}) \sum_{i = 1}^{n} (y_{o} (i) - \overset{true¯}{y_{0}})})}^{2}

where “ y e ” and “ y 0 ” are, respectively, the estimated and the observed output, and

\overset{true¯}{y_{e}}

and

\overset{true¯}{y_{0}}

represent their mean values.…”

Section: Back Propagation Neural Network For Desalination‐renewable E...mentioning

confidence: 99%

Design of water‐energy management based on neural networks for an autonomous solar and wind renewable energy system coupled with brackish water desalination

Zgalmi

Cherif

2022

Env Prog and Sustain Energy

View full text Add to dashboard Cite

In this article, smart water-energy management based on an artificial neural network for a photovoltaic-wind system coupled with water pumping and reverse osmosis desalination unit with hydraulic storage is developed. The proposed system was designed to meet freshwater demand for remote areas of Tunisia's south. The authors focus on an artificial neural network-type management strategy, ensuring power-sharing between sources, hydraulic components, and tank storage. The main objective of the management strategy is to deliver and purify as much freshwater as possible by taking advantage of all the energy available according to climatic conditions. A parametric sensitivity algorithm is developed to find the most appropriate neural network architecture. A dynamic simulator for 1 year of operation integrates the energy management of the system is developed. The used water-energy management-based artificial neural network showed good results of power-sharing during a fast time with a correlation coefficient and a coefficient of determination between the actual and estimated values of the three motor pumps electrical power equal to 96% and 98%, respectively. Also, the root mean squared error was acceptable at 0.1476. These results have been proved over a year and can be used by other researchers with a similar system.

show abstract

“…The power coefficient C p is an indicator of the useful power in the wind flow and it is a function of the pitch angle β and λ. The power coefficient can be defined using different methods, including the theory of aerodynamics, blade element momentum (BEM) theory, computational fluid dynamics (CFD), fuzzy logic, or generalized dynamic wake (GDW) models [9], [10], [11]. Nevertheless, a numerical approximation of the aerodynamic power coefficient is often accepted for achieving enough simplicity and accuracy [12], [13].…”

Section: A Mechanical Modelmentioning

confidence: 99%

Degradation of a wind-turbine drive-train under turbulent conditions: effect of the control law

Romero

Martínez

Bérenguer

2021

2021 5th International Conference on Control and Fault-Tolerant Systems (SysTol)

View full text Add to dashboard Cite

This work compares the simulated degradation produced at the drive-train of a given wind turbine when it is functioning at optimal and sub-optimal operating points. The simulation considers different classes of wind conditions and supposes the use of the same Maximum Power Point Tracking (MPPT) algorithm but calibrated with different feedback gains. The dissipation at the drive-train is modeled using contact mechanics principles, and it is intended for modeling the mechanical fatigue due to changes in wind speed and turbulence. The paper presents the proposed model for the drive-train degradation and the obtained power curves for a simulated variable-speed 2M W (100m rotor diameter) wind turbine, with fixed gear box and horizontal-axis.

show abstract

Prediction of power generation and rotor angular speed of a small wind turbine equipped to a controllable duct using artificial neural network and multiple linear regression

Cited by 32 publications

References 34 publications

Modeling Multioutput Response Uses Ridge Regression and MLP Neural Network with Tuning Hyperparameter through Cross Validation

Modeling Multioutput Response Uses Ridge Regression and MLP Neural Network with Tuning Hyperparameter through Cross Validation

Design of water‐energy management based on neural networks for an autonomous solar and wind renewable energy system coupled with brackish water desalination

Degradation of a wind-turbine drive-train under turbulent conditions: effect of the control law

Contact Info

Product

Resources

About