Identification and Correction of Outliers in Wind Farm Time Series Power Data

Yang, Xi; Lu, Zongxiang; Qiao, Ying; Min, Yong; O’Malley, Mark

doi:10.1109/tpwrs.2015.2512843

Cited by 59 publications

(36 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Fault data shows that wind power is low when the wind speed is high, which may be caused by the loss of connection and data transmission error of some wind turbines. Outage data shows that the output power is 0 when the actual wind speed is greater than the starting up wind speed of the wind turbine, which is the wind speed that can make the wind turbine start to rotate, and the reason may be that wind power plant is abandoning the wind or overhauling [13].…”

Section: Setting the Parameters Of Dbscan Algorithmmentioning

confidence: 99%

Short-Term Wind Power Prediction Using GA-BP Neural Network Based on DBSCAN Algorithm Outlier Identification

et al. 2020

View full text Add to dashboard Cite

Accurately predicting wind power plays a vital part in site selection, large-scale grid connection, and the safe and efficient operation of wind power generation equipment. In the stage of data pre-processing, density-based spatial clustering of applications with noise (DBSCAN) algorithm is used to identify the outliers in the wind power data and the collected wind speed data of a wind power plant in Shandong Province, and the linear regression method is used to correct the outliers to improve the prediction accuracy. Considering the important impact of wind speed on power, the average value, the maximum difference and the average change rate of daily wind speed of each historical day are used as the selection criteria to select similar days by using DBSCAN algorithm and Euclidean distance. The short-term wind power prediction is carried out by using the similar day data pre-processed and unprocessed, respectively, as the input of back propagation neural network optimized by genetic algorithm (GA-BP neural network). Analysis of the results proves the practicability and efficiency of the prediction model and the important role of outlier identification and correction in improving the accuracy of wind power prediction.

show abstract

Section: Setting the Parameters Of Dbscan Algorithmmentioning

confidence: 99%

Short-Term Wind Power Prediction Using GA-BP Neural Network Based on DBSCAN Algorithm Outlier Identification

et al. 2020

View full text Add to dashboard Cite

show abstract

“…(2) The dimension equivalent of each dimension of the collected data is calculated according to Equations (1-10) to avoid the influence of different dimensions. (3) According to Equations (11)(12)(13)(14)(15)(16)(17)(18)(19), the dimensionality of the different indicators in the NWP data is reduced to reduce the redundancy of the phase data. (4) The effectiveness of the preprocessing is verified by using the LSTM network for power prediction.…”

Section: Numerical Weather Data Acquisition and Processing Stepsmentioning

confidence: 99%

“…Zhao et al [17] studied the feature reduction analysis of wind-induced anomaly data, and integrated the quadrilateral method and density-based clustering method to eliminate sparse outliers. Ye et al [18] used the adjacent spatial correlation to establish an outlier identification algorithm based on the probabilistic wind farm power curve for the missing data problem in wind farm time series power data.…”

Section: Introductionmentioning

confidence: 99%

Wind Farm NWP Data Preprocessing Method Based on t-SNE

Xie

Zhang

2019

Energies

View full text Add to dashboard Cite

The operation prediction of wind farms will be accompanied by the need for massive data processing, especially the preprocessing of wind farm meteorological data or numerical weather prediction (NWP). Because NWP data are strongly correlated with wind farm operation, proper processing of NWP data could not only reduce data volume but also improve the correlations of wind farm operation predictions. For this purpose, this paper proposes a data preprocessing algorithm based on t-distributed stochastic neighbor embedding (t-SNE). Firstly, the data collected were normalized to eliminate the influence caused by different dimensions. The t-SNE algorithm is then used to reduce the dimensionality of the NWP data related to wind farm operation. Finally, the wind farm data visualization platform is established. In this paper, 22 index variables in NWP data were taken as objects. The t-SNE method was used to preprocess the NWP historical data of a wind farm, and the results were compared with the results of the principal component analysis (PCA) algorithm. It outperformed PCA in error precision; in addition, t-SNE dimension reduction preprocessing also had a visual effect, which could be applied to big data visualization platforms. A long short-term memory network (LSTM) was used to predict the operation of the wind farm by combining the preprocessed NWP data and the operation data. The simulation results proved that the effect of the preprocessed NWP data based on t-SNE on the wind power prediction was significantly improved.

show abstract

“…Então diante da importância do histórico dos dados, a qualidade e a integralidade das medições deve ser buscadas, pois para as atividades de operação, manutenção e previsão de séries temporaiś e importante uma integralidade mínima dos dados medidos (Ye et al, 2016).…”

Section: Introductionunclassified

“…Diversas metodologias para detecção e correcão de outliers têm sido desenvolvidas para diversos domínios de aplicações como engenharia, economia, metereologia, dentre outros. E para cada domínio de aplicação tende a ser desenvolvido um métodoúnico de acordo com o tipo de série temporal e o tipo de outlier (Ye et al, 2016). Por exemplo, em (Valverde and Terzija., 2011)é utilizada uma técnica de estimação de estados qué e eficiente, desde que o modelo da série temporal seja preciso e que obedeça algumas propriedades estatísticas (Zhang et al, 2011).…”

Section: Introductionunclassified

Detecção e Correção de Outliers em Curvas de Potência Utilizando Técnica Híbrida de Inteligência Artificial

Andrade

Villanueva

2019

Anais Do 14º Simpósio Brasileiro De Automação Inteligente

View full text Add to dashboard Cite

The distribution substations power curves are often affected by outliers: very discrepant measurements of the curve standard behavior. If the presence of outliers is very large, the internal studies and analysis of the power utilities developed from the history of the data collected may be compromised. In order to detect and correct atypical data, a complementary module for the Supervisory System was developed. In the previous work of the authors, two techniques were developed based on artificial intelligence: fuzzy logic and artificial neural networks. In this work is presented a hybrid technique that uses the best performances of the previous ones. Also developed in this paper, is an outlier detection method based on the standard behavior of the curve and the distribution of the collected samples. The performances of the techniques are compared using real data from a substation over 6 years.

show abstract

Identification and Correction of Outliers in Wind Farm Time Series Power Data

Cited by 59 publications

References 26 publications

Short-Term Wind Power Prediction Using GA-BP Neural Network Based on DBSCAN Algorithm Outlier Identification

Short-Term Wind Power Prediction Using GA-BP Neural Network Based on DBSCAN Algorithm Outlier Identification

Wind Farm NWP Data Preprocessing Method Based on t-SNE

Detecção e Correção de Outliers em Curvas de Potência Utilizando Técnica Híbrida de Inteligência Artificial

Contact Info

Product

Resources

About