Wind Turbine Multivariate Power Modeling Techniques for Control and Monitoring Purposes

2020 ◽  
Vol 143 (3) ◽  
Author(s):  
Davide Astolfi ◽  
Francesco Castellani ◽  
Francesco Natili
Keyword(s):  
Wind Turbine ◽  
Principal Component ◽  
Data Driven ◽  
Gaussian Kernel ◽  
Support Vector ◽  
Rotor Speed ◽  
Power Modeling ◽  
Ambient Conditions ◽  
Power Monitoring ◽  
Turbine Performance

Abstract Wind turbine performance monitoring is a complex task because the power has a multivariate dependence on ambient conditions and working parameters. Furthermore, wind turbine nacelle anemometers are placed behind the rotor span and the control system estimates the upwind flow through a nacelle transfer function: this introduces a data quality issue. This study is devoted to the analysis of data-driven techniques for wind turbine performance control and monitoring: operation data of six 850 kW wind turbines sited in Italy have been employed. The objective of this study is an assessment of several easily implementable techniques and input variables selections for data-driven models whose target is the power of a wind turbine. Three model types are selected: one is linear (Principal Component Regression) and two are nonlinear (Support Vector Regression with Gaussian Kernel and Feedforward Artificial Neural Network). The models' validation provides meaningful indications: the linear model in general has lower performance because it cannot reproduce properly the nonlinear pitch behavior when approaching rated power. Therefore, it is concluded that a nonlinear model should be employed and the achieved mean absolute error is of the order of 1.3% of the rated power. Furthermore, the errors are kept at the order of 2% of the rated power for the models whose input is the rotor speed instead that wind speed: this observation supports that, in case it is needed because of nacelle anemometer biases, the power monitoring can be acceptably implemented using the rotor speed.

scholarly journals A Study of Wind Turbine Performance Decline with Age through Operation Data Analysis

Energies ◽  
2020 ◽  
Vol 13 (8) ◽  
pp. 2086 ◽  
Author(s):  
Raymond Byrne ◽  
Davide Astolfi ◽  
Francesco Castellani ◽  
Neil J. Hewitt
Keyword(s):  
Wind Turbine ◽  
Gaussian Kernel ◽  
Support Vector ◽  
Test Case ◽  
Data Set ◽  
Turbine Performance ◽  
Performance Decline ◽  
Performance Deterioration ◽  
Stationary Conditions ◽  
Institute Of Technology

Ageing of technical systems and machines is a matter of fact. It therefore does not come as a surprise that an energy conversion system such as a wind turbine, which in particular operates under non-stationary conditions, is subjected to performance decline with age. The present study presents an analysis of the performance deterioration with age of a Vestas V52 wind turbine, installed in 2005 at the Dundalk Institute of Technology campus in Ireland. The wind turbine has operated from October 2005 to October 2018 with its original gearbox, that has subsequently been replaced in 2019. Therefore, a key point of the present study is that operation data spanning over thirteen years have been analysed for estimating how the performance degrades in time. To this end, one of the most innovative approaches for wind turbine performance control and monitoring has been employed: a multivariate Support Vector Regression with Gaussian Kernel, whose target is the power output of the wind turbine. Once the model has been trained with a reference data set, the performance degradation is assessed by studying how the residuals between model estimates and measurements evolve. Furthermore, a power curve analysis through the binning method has been performed to estimate the Annual Energy Production variations and suggests that the most convenient strategy for the test case wind turbine (running the gearbox until its end of life) has indeed been adopted. Summarizing, the main results of the present study are as follows: over a ten-year period, the performance of the wind turbine has declined of the order of 5%; the performance deterioration seems to be nonlinear as years pass by; after the gearbox replacement, a fraction of performance deterioration has been recovered, though not all because the rest of the turbine system has been operating for thirteen years from its original state. Finally, it should be noted that the estimate of performance decline is basically consistent with the few results available in the literature.

scholarly journals Wind Turbine Operation Curves Modelling Techniques

Electronics ◽  
2021 ◽  
Vol 10 (3) ◽  
pp. 269 ◽  
Author(s):  
Davide Astolfi
Keyword(s):  
Wind Speed ◽  
Wind Turbine ◽  
Wind Turbines ◽  
Rotational Speed ◽  
Gaussian Kernel ◽  
Percentage Error ◽  
Support Vector ◽  
Power Monitoring ◽  
Independent Variables ◽  
Error Metrics

Wind turbines are machines operating in non-stationary conditions and the power of a wind turbine depends non-trivially on environmental conditions and working parameters. For these reasons, wind turbine power monitoring is a complex task which is typically addressed through data-driven methods for constructing a normal behavior model. On these grounds, this study is devoted the analysis of meaningful operation curves, which are rotor speed-power, generator speed-power and blade pitch-power. A key point is that these curves are analyzed in the appropriate operation region of the wind turbines: the rotor and generator curves are considered for moderate wind speed, when the blade pitch is fixed and the rotational speed varies (Region 2); the blade pitch curve is considered for higher wind speed, when the rotational speed is rated (Region 2 12). The selected curves are studied through a multivariate Support Vector Regression with Gaussian Kernel on the Supervisory Control And Data Acquisition (SCADA) data of two wind farms sited in Italy, featuring in total 15 2 MW wind turbines. An innovative aspect of the selected models is that minimum, maximum and standard deviation of the independent variables of interest are fed as input to the models, in addition to the typically employed average values: using the additional covariates proposed in this work, the error metrics decrease of order of one third, with respect to what would be obtained by employing as regressors only the average values of the independent variables. In general it results that, for all the considered curves, the prediction of the power is characterized by error metrics which are competitive with the state of the art in the literature for multivariate wind turbine power curve analysis: in particular, for one test case, a mean absolute percentage error of order of 2.5% is achieved. Furthermore, the approach presented in this study provides a superior capability of interpreting wind turbine performance in terms of the behavior of the main sub-components and eliminates as much as possible the dependence on nacelle anemometer data, whose use is critical because of issues related to the sites complexity.

scholarly journals Multivariate SCADA Data Analysis Methods for Real-World Wind Turbine Power Curve Monitoring

Energies ◽  
2021 ◽  
Vol 14 (4) ◽  
pp. 1105 ◽  
Author(s):  
Davide Astolfi ◽  
Francesco Castellani ◽  
Andrea Lombardi ◽  
Ludovico Terzi
Keyword(s):  
Standard Deviation ◽  
Wind Turbine ◽  
Wind Turbines ◽  
Curve Analysis ◽  
Gaussian Kernel ◽  
Support Vector ◽  
Features Selection ◽  
Power Curve ◽  
Working Parameters ◽  
Extracted Power

Due to the stochastic nature of the source, wind turbines operate under non-stationary conditions and the extracted power depends non-trivially on ambient conditions and working parameters. It is therefore difficult to establish a normal behavior model for monitoring the performance of a wind turbine and the most employed approach is to be driven by data. The power curve of a wind turbine is the relation between the wind intensity and the extracted power and is widely employed for monitoring wind turbine performance. On the grounds of the above considerations, a recent trend regarding wind turbine power curve analysis consists of the incorporation of the main working parameters (as, for example, the rotor speed or the blade pitch) as input variables of a multivariate regression whose target is the power. In this study, a method for multivariate wind turbine power curve analysis is proposed: it is based on sequential features selection, which employs Support Vector Regression with Gaussian Kernel. One of the most innovative aspects of this study is that the set of possible covariates includes also minimum, maximum and standard deviation of the most important environmental and operational variables. Three test cases of practical interest are contemplated: a Senvion MM92, a Vestas V90 and a Vestas V117 wind turbines owned by the ENGIE Italia company. It is shown that the selection of the covariates depends remarkably on the wind turbine model and this aspect should therefore be taken in consideration in order to customize the data-driven monitoring of the power curve. The obtained error metrics are competitive and in general lower with respect to the state of the art in the literature. Furthermore, minimum, maximum and standard deviation of the main environmental and operation variables are abundantly selected by the feature selection algorithm: this result indicates that the richness of the measurement channels contained in wind turbine Supervisory Control And Data Acquisition (SCADA) data sets should be exploited for monitoring the performance as reliably as possible.

Physical-oriented and machine learning-based emission modeling in a diesel compression ignition engine: Dimensionality reduction and regression

2022 ◽  
pp. 146808742110707
Author(s):  
Aran Mohammad ◽  
Reza Rezaei ◽  
Christopher Hayduk ◽  
Thaddaeus Delebinski ◽  
Saeid Shahpouri ◽  
...  
Keyword(s):  
Principal Component Analysis ◽  
Support Vector Machine ◽  
Factor Analysis ◽  
Dimensionality Reduction ◽  
Principal Component ◽  
Component Analysis ◽  
Data Driven ◽  
Support Vector ◽  
Emission Models ◽  
Emission Modeling

The development of internal combustion engines is affected by the exhaust gas emissions legislation and the striving to increase performance. This demands for engine-out emission models that can be used for engine optimization for real driving emission controls. The prediction capability of physically and data-driven engine-out emission models is influenced by the system inputs, which are specified by the user and can lead to an improved accuracy with increasing number of inputs. Thereby the occurrence of irrelevant inputs becomes more probable, which have a low functional relation to the emissions and can lead to overfitting. Alternatively, data-driven methods can be used to detect irrelevant and redundant inputs. In this work, thermodynamic states are modeled based on 772 stationary measured test bench data from a commercial vehicle diesel engine. Afterward, 37 measured and modeled variables are led into a data-driven dimensionality reduction. For this purpose, approaches of supervised learning, such as lasso regression and linear support vector machine, and unsupervised learning methods like principal component analysis and factor analysis are applied to select and extract the relevant features. The selected and extracted features are used for regression by the support vector machine and the feedforward neural network to model the NOx, CO, HC, and soot emissions. This enables an evaluation of the modeling accuracy as a result of the dimensionality reduction. Using the methods in this work, the 37 variables are reduced to 25, 22, 11, and 16 inputs for NOx, CO, HC, and soot emission modeling while maintaining the accuracy. The features selected using the lasso algorithm provide more accurate learning of the regression models than the extracted features through principal component analysis and factor analysis. This results in test errors RMSETe for modeling NOx, CO, HC, and soot emissions 19.22 ppm, 6.46 ppm, 1.29 ppm, and 0.06 FSN, respectively.

scholarly journals Multivariate Wind Turbine Power Curve Model Based on Data Clustering and Polynomial LASSO Regression

2021 ◽  
Vol 12 (1) ◽  
pp. 72
Author(s):  
Davide Astolfi ◽  
Ravi Pandit
Keyword(s):  
Wind Turbine ◽  
Data Clustering ◽  
Performance Monitoring ◽  
Power Curve ◽  
Lasso Regression ◽  
Data Set ◽  
Power Monitoring ◽  
Turbine Performance ◽  
Input Variables ◽  
Power Curves

Wind turbine performance monitoring is a complex task because of the non-stationary operation conditions and because the power has a multivariate dependence on the ambient conditions and working parameters. This motivates the research about the use of SCADA data for constructing reliable models applicable in wind turbine performance monitoring. The present work is devoted to multivariate wind turbine power curves, which can be conceived of as multiple input, single output models. The output is the power of the target wind turbine, and the input variables are the wind speed and additional covariates, which in this work are the blade pitch and rotor speed. The objective of this study is to contribute to the formulation of multivariate wind turbine power curve models, which conjugate precision and simplicity and are therefore appropriate for industrial applications. The non-linearity of the relation between the input variables and the output was taken into account through the simplification of a polynomial LASSO regression: the advantages of this are that the input variables selection is performed automatically. The k-means algorithm was employed for automatic multi-dimensional data clustering, and a separate sub-model was formulated for each cluster, whose total number was selected by analyzing the silhouette score. The proposed method was tested on the SCADA data of an industrial Vestas V52 wind turbine. It resulted that the most appropriate number of clusters was three, which fairly resembles the main features of the wind turbine control. As expected, the importance of the different input variables varied with the cluster. The achieved model validation error metrics are the following: the mean absolute percentage error was in the order of 7.2%, and the average difference of mean percentage errors on random subsets of the target data set was of the order of 0.001%. This indicates that the proposed model, despite its simplicity, can be reliably employed for wind turbine power monitoring and for evaluating accumulated performance changes due to aging and/or optimization.

Data-driven fault detection methods for detecting small-magnitude faults in anaerobic digestion process

2020 ◽  
Vol 81 (8) ◽  
pp. 1740-1748 ◽  
Author(s):  
Pezhman Kazemi ◽  
Jaume Giralt ◽  
Christophe Bengoa ◽  
Jean-Philippe Steyer
Keyword(s):  
Anaerobic Digestion ◽  
Control Charts ◽  
Volatile Fatty Acids ◽  
Principal Component ◽  
Data Driven ◽  
Detection Methods ◽  
Support Vector ◽  
Statistical Control ◽  
Excellent Performance ◽  
Small Magnitude

Abstract Early detection of small-magnitude faults in anaerobic digestion (AD) processes is a mandatory step for preventing serious consequence in the future. Since volatile fatty acids (VFA) accumulation is widely suggested as a process health indicator, a VFA soft-sensor was developed based on support vector machine (SVM) and used for generating the residuals by comparing real and predicted VFA. The estimated residual signal was applied to univariate statistical control charts such as cumulative sum (CUSUM) and square prediction error (SPE) to detect the faults. A principal component analysis (PCA) model was also developed for comparison with the aforementioned approach. The proposed framework showed excellent performance for detecting small-magnitude faults in the state parameters of AD processes.

Multi-Source Information Fusion Based on Data Driven

2010 ◽  
Vol 40-41 ◽  
pp. 121-126
Author(s):  
Xin Zhang ◽  
Li Yang ◽  
Yan Zhang
Keyword(s):  
Information Fusion ◽  
Rough Set Theory ◽  
Dimensional Space ◽  
System Modeling ◽  
Principal Component ◽  
Data Driven ◽  
Support Vector ◽  
Source Information ◽  
Knowledge Reduction ◽  
Lower Dimensional Space

Take data driven method as theoretical basis, study multi-source information fusion technology. Using online and off-line data of the fusion system, does not rely on system's mathematical model, has avoided question about system modeling by mechanism. Uses principal component analysis method, rough set theory, Support Vector Machine(SVM) and so on, three method fusions and supplementary, through information processing and feature extraction to system's data-in, catches the most important information to lower dimensional space, realizes knowledge reduction. From data level, characteristic level, decision-making three levels realize information fusion. The example indicated that reduced computational complexity, reduced information loss in the fusion process, and enhanced the fusion accuracy.

scholarly journals Adaptable and Explainable Predictive Maintenance: Semi-Supervised Deep Learning for Anomaly Detection and Diagnosis in Press Machine Data

2021 ◽  
Vol 11 (16) ◽  
pp. 7376
Author(s):  
Oscar Serradilla ◽  
Ekhi Zugasti ◽  
Julian Ramirez de Okariz ◽  
Jon Rodriguez ◽  
Urko Zurutuza
Keyword(s):  
Anomaly Detection ◽  
Null Space ◽  
Model Performance ◽  
Principal Component ◽  
Predictive Maintenance ◽  
Data Driven ◽  
Support Vector ◽  
Operational Conditions ◽  
Detection Model ◽  
Cluster Data

Predictive maintenance (PdM) has the potential to reduce industrial costs by anticipating failures and extending the work life of components. Nowadays, factories are monitoring their assets and most collected data belong to correct working conditions. Thereby, semi-supervised data-driven models are relevant to enable PdM application by learning from assets’ data. However, their main challenges for application in industry are achieving high accuracy on anomaly detection, diagnosis of novel failures, and adaptability to changing environmental and operational conditions (EOC). This article aims to tackle these challenges, experimenting with algorithms in press machine data of a production line. Initially, state-of-the-art and classic data-driven anomaly detection model performance is compared, including 2D autoencoder, null-space, principal component analysis (PCA), one-class support vector machines (OC-SVM), and extreme learning machine (ELM) algorithms. Then, diagnosis tools are developed supported on autoencoder’s latent space feature vector, including clustering and projection algorithms to cluster data of synthetic failure types semi-supervised. In addition, explainable artificial intelligence techniques have enabled to track the autoencoder’s loss with input data to detect anomalous signals. Finally, transfer learning is applied to adapt autoencoders to changing EOC data of the same process. The data-driven techniques used in this work can be adapted to address other industrial use cases, helping stakeholders gain trust and thus promote the adoption of data-driven PdM systems in smart factories.

Sign in / Sign up

Export Citation Format

Share Document