Very short-term load forecasting of local loads based on local shape similarity corrective changing

The level of load is related to the weather and the date type, while local loads are affected by its curriculum schedule. The original similar-day algorithm focuses on the weather and the factor considering whether it is holiday or not. It would cause errors if this algorithm is applied directly in...

Full description

Saved in:
Bibliographic Details
Main Authors: Wu Shun-yu, Liu Shi-rong, Ning Kang-hong, Peng Ming-wei
Format: Conference Proceeding
Language:English
Subjects:
Automation
Corrective changing
Electronic mail
Load forecasting
Local load forecasting
Local shape similarity
Meteorology
Neural networks
Shape
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The level of load is related to the weather and the date type, while local loads are affected by its curriculum schedule. The original similar-day algorithm focuses on the weather and the factor considering whether it is holiday or not. It would cause errors if this algorithm is applied directly in local load forecasting. To solve this problem, a very short-term load forecasting method based on modified local shape similarity was proposed. In this method, a change rate of load is generated by the local similar-day algorithm, another change rate of load is obtained according to the change rate at the same moment in the same work day. Then, a synthetical change rate is calculated by the above two values, and used to forecast the load of the next moment. This approach can reduce the error caused by sudden changes sharply, and improve the performance of global prediction as well as the prediction of the peaks and troughs of a load.
ISSN:1934-1768
2161-2927