Load forecasting for smart grid using non-linear model in Hadoop distributed file system

The conventional electrical grid structure is evolving in recent years, enhancing with new technology. The new structure of power system ‘Smart grid’ is trying to get solution for the problems in conventional grid. In a smart grid environment, every end user is connected with the market. The control...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Cluster computing Jg. 22; H. Suppl 6; S. 13533 - 13545
Hauptverfasser: Arun Jees, S., Gomathi, Venugopal
Format: Journal Article
Sprache:Englisch
Veröffentlicht: New York Springer US 01.11.2019
Springer Nature B.V
Schlagworte:
Algorithms
Artificial intelligence
Big Data
Classification
Clustering
Communications systems
Complexity
Computer Communication Networks
Computer Science
Decision trees
Electric power grids
Electric power systems
Electrical loads
Electricity
Energy industry
Forecasting
Machine learning
Neural networks
New technology
Operating Systems
Processor Architectures
Regression analysis
Smart grid
Support vector machines
Time series
Variables
k-means clustering
Forecasting model
Support vector machine
Hadoop framework
ISSN:1386-7857, 1573-7543
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The conventional electrical grid structure is evolving in recent years, enhancing with new technology. The new structure of power system ‘Smart grid’ is trying to get solution for the problems in conventional grid. In a smart grid environment, every end user is connected with the market. The control and data signals will flow in both directions from consumer to market and market to consumer instantaneously to provide a reliable market operation. Data plays a main role for control operation, load scheduling and market operation. Market operator will find hard to maintain a stable operation because with a present infrastructure huge number of instantaneous data cannot be manipulated for reliable operation. Load forecasted data helps ISO operator to decide the forecasted market clearing price for power market and it will also help the consumer and producer to choose the profitable asset in time period. Load forecasting method can be used for any duration, but the complexity increases with decreased duration. Every consumer in smart grid is a live participant in market operation need to connect through smart communication system to send data and receive control from the operator. When the smart grid structure increases number of data will also increase drastically, operator should store and analyse the huge number of incoming data. Unlike in conventional grid load forecasting is very complex in smart grid with the huge number of data vulnerability. Smart grid tends to solve this data congestion using the big data technology. Load forecasting is obtained using k means algorithm, grey correlation degree, decision tree algorithm and support vector machine to predict an accurate forecasting result in Hadoop framework.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1386-7857
1573-7543
DOI:10.1007/s10586-018-1997-2