Intelligent mitigation of blackout in real-time microgrids: Neural Network Approach

Zarrabian, Sina; Belkacemi, Rabie; Babalola, Adeniyi A.

doi:10.1109/peci.2016.7459213

Cited by 16 publications

(1 citation statement)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In a protection scheme, the present measuring devices are used to give input data and to provide inclusive station protection. Thus, researchers began to research this zone utilizing numerous techniques, where artificial neural networks (ANNs) appeared as a promising method to detect different types of faults [6], [7], [8]. The ANN technology includes a data processing step that set the input data for the ANN [9], [11], [12].…”

Section: Introductionmentioning

confidence: 99%

Data Processing Method for Artificial Neural Network ANN Based Micro grid Protection Model

Baidya

Nandi

2023

EAE

View full text Add to dashboard Cite

Effective fault detection and isolation technologies are very necessary for uninterrupted power supply and for making a flexible protection scheme. Almost all protection schemes in the power system are based on data exchange among protection units through a strong communication structure. Thus, it is important to deal with a large amount of data. Artificial Intelligence (AI) is one of the key factors in this regard. AI has several sections and Artificial Neural Network (ANN) is one of them. It is suggested to implement the ANN-based models while working with big data. The existing protection models are facing difficulties while trying to deal with big data. Thus ANN-based approaches have come into the front line in advanced power system networks. The performance of the ANN model is depending on the training of the data set. Hence in this work, we are focusing on preparing the data to provide input in the ANN model. The principal component analysis (PCA) method is applied here for reduced the dimension of a large number of data sets. The new data set is used to run the k-means clustering algorithm. It is shown that the clustering is more accurate with the processed data set by PCA. Therefore, the prepared data set is used to run the ANN model that has a smaller size with higher information and minimum computational time. This study shows the data preparation part to train the ANN model.

show abstract