The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON POWER SYSTEMS
Transactions ID Number: 28-794
Full Name: Jose Fidalgo
Position: Assistant Professor
Age: ON
Sex: Male
Address: Rua Dr. Roberto Frias, s/n 4200-465 Porto
Country: PORTUGAL
Tel: 225083370
Tel prefix: +351
Fax: +351225081440
E-mail address: jfidalgo@inescporto.pt
Other E-mails: jfidalgo@fe.up.pt
Title of the Paper: Estimation of Load Diagrams in MV/LV Substations
Authors as they appear in the Paper: J. N. Fidalgo
Email addresses of all the authors: jfidalgo@inescporto.pt
Number of paper pages: 10
Abstract: Efficient power systems planning and exploration require the estimation of load diagrams at the different levels of the distribution networks. In particular, the planning of new MV/LV substations requires the assessment of their expected load curves under different exploration scenarios. The deregulation process and environment preservation constraints also compel the need of higher efficiency levels in network investments. This paper describes the methodologies adopted for the estimation of load curve diagrams in MV/LV substations. The assessment process is based on billing data (monthly energy consumption, hired power contracts, activity codes and weekday type), which is the unique information generally available at this level of the network. The most common approaches use measurements in typical classes of consumers defined by experts to construct inference engines that, most of the times, only estimate peak loads. In this paper, two different approaches are te!
sted. The first one is based on the definition of classes using a clustering algorithm and uses Artificial Neural Networks (ANN) for the estimation of the MV/LV substation load curve. In the second one, an ANN is trained to output directly the load diagram estimated for each individual consumer. This article describes the adopted methodologies and presents some representative results. Performance attained is discussed as well as a method to achieve confidence intervals.
Keywords: Load estimation, Clustering, Artificial neural networks, Distribution networks
EXTENSION of the file: .pdf
Special (Invited) Session: Load Curve Estimation For Distribution Systems Using ANN
Organizer of the Session: 605-503
How Did you learn about congress: Joao Paulo Saraiva (jsaraiva@fe.up.pt)
IP ADDRESS: 193.136.33.149
