A New Approach Based on Wavelet Design and Machine Learning for Islanding Detection of Distributed Generation

Alshareef, Sami M.; Talwar, Saurabh; Morsi, Walid G.

doi:10.1109/tsg.2013.2296598

Cited by 88 publications

(32 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For each combination of the parameters a and b there is a corresponding wavelet function W. To determine the most appropriate wavelet function W * that closely matches the target signal X, the approach presented in [21] and [22] using procrustes analysis (PA) [23] is used in this paper. In PA, the matching wavelet W is transformed to obtain the comparison signal Z…”

Section: Wavelet Design Proceduresmentioning

confidence: 99%

Nonintrusive Load Monitoring Using Wavelet Design and Machine Learning

Gillis

Alshareef

Morsi

2016

IEEE Trans. Smart Grid

118

View full text Add to dashboard Cite

This paper presents a new concept based on wavelet design and machine learning applied to nonintrusive load monitoring. The wavelet coefficients of length-6 filter are determined using procrustes analysis and are used to construct new wavelets to match the load signals to be detected, unlike previous work which used previously designed wavelet functions that are special cases of Daubechies filters to suit other nonpower system applications such as communications and image processing. The results of applying the new concept to a test system consisting of four loads have shown that the newly designed wavelet can improve the prediction accuracy compared with that obtained using Daubechies filter of order three while keeping the prominent features of the pattern in the detail levels.Index Terms-Load signature, machine learning, nonintrusive load monitoring (NILM), wavelet design.

show abstract

Section: Wavelet Design Proceduresmentioning

confidence: 99%

Nonintrusive Load Monitoring Using Wavelet Design and Machine Learning

Gillis

Alshareef

Morsi

2016

IEEE Trans. Smart Grid

118

View full text Add to dashboard Cite

show abstract

“…The simulation results indicated that the technique possessed accuracy of 98% and above [35]. Similarly, application of support vector machine for enhancement in the performance of passive islanding detection technqiue has been proposed in [36][37][38]. Apart from these, probabilistic neural network [39] and artificial neural network [40] are also utilized in the passive islanding detection techniques to enhance their performance.…”

Section: Passive Islanding Detection Techniquesmentioning

confidence: 99%

An islanding detection strategy for distribution network connected with hybrid DG resources

Laghari

Mokhlis

Karimi

et al. 2015

Renewable and Sustainable Energy Reviews

View full text Add to dashboard Cite

Average rate of change of reactive power Bio-Mass Mini-hydro a b s t r a c tThe exponential growth in electricity demand has driven the Distributed Generation (DG) technology a boost in the power system. The use of DG is beneficial to power utilities, DG owners', and end-users in terms of reliability, power quality, and economics. However, to fully utilize the benefits of DGs, some technical issues need to be addressed. Islanding condition is one of the most important issue in this context. Until now, several islanding detection techniques have been proposed for detecting the islanding condition. This paper presents a brief overview of existing islanding detection techniques with their relative merits and demerits. Apart from this, the paper presents an islanding detection strategy suitable for hybrid DG resources of mini-hydro and Bio-Mass. The proposed strategy uses average rate of change of reactive power and load shift strategy to detect islanding of the distribution network. The performance of proposed strategy is validated on various islanding and non-islanding events. The results of proposed strategy are compared with other existing techniques in terms of fast islanding detection, and non-detection zone. The simulation results show that the proposed strategy is effective in detecting islanding phenomenon possesses fast detection and negligible non-detection zone region compared to existing islanding detection techniques.

show abstract

“…However, the proposed method's sensitivity to outliners and the noisy conditions is considerable [26]. Support vector machine (SVM) with wavelet transform has been utilized to detect islanding [27]. The results of the proposed method show that although as signal processing tool the wavelet transform has suitable time-frequency localization ability, it faces barriers, e.g., batch processing step, non-uniform frequency sub-bands, less flexibility and detection failure during noisy conditions [27].…”

Section: Introductionmentioning

confidence: 99%

“…Support vector machine (SVM) with wavelet transform has been utilized to detect islanding [27]. The results of the proposed method show that although as signal processing tool the wavelet transform has suitable time-frequency localization ability, it faces barriers, e.g., batch processing step, non-uniform frequency sub-bands, less flexibility and detection failure during noisy conditions [27]. Different methods based on the combination of artificial neural network and fuzzy logic are presented in [25,28,29].…”

Section: Introductionmentioning

confidence: 99%

An Anti-Islanding Protection Technique Using a Wavelet Packet Transform and a Probabilistic Neural Network

et al. 2018

View full text Add to dashboard Cite

This paper proposes a new islanding detection technique based on the combination of a wavelet packet transform (WPT) and a probabilistic neural network (PNN) for grid-tied photovoltaic systems. The point of common coupling (PCC) voltage is measured and processed by the WPT to find the normalized Shannon entropy (NSE) and the normalized logarithmic energy entropy (NLEE). Subsequently, the yield feature vectors are fed to the PNN classifier to classify the disturbances. The PNN is trained with different spread factors to obtain better classification accuracy. For the best performance of the proposed method, the precise analysis is done for the selection of the type of input data for the PNN, the type of mother wavelet, and the required transform level which is based on the accuracy, simplicity, specificity, speed, and cost parameters. The results show that, by using normalized Shannon entropy and the normalized logarithmic energy entropy, not only it offers simplicity, specificity and reduced costs, it also has better accuracy compared to other smart and passive methods. Based on the results, the proposed islanding detection technique is highly accurate and does not mal-operate during islanding and non-islanding events.

show abstract

A New Approach Based on Wavelet Design and Machine Learning for Islanding Detection of Distributed Generation

Cited by 88 publications

References 18 publications

Nonintrusive Load Monitoring Using Wavelet Design and Machine Learning

Nonintrusive Load Monitoring Using Wavelet Design and Machine Learning

An islanding detection strategy for distribution network connected with hybrid DG resources

An Anti-Islanding Protection Technique Using a Wavelet Packet Transform and a Probabilistic Neural Network

Contact Info

Product

Resources

About