Abstract
On account of the strong time series feature of smart grid load data, this paper presents a data cleaning method based on similarity measurement of time series, which detects the abnormal data of load data and fills the vacancy value. In this paper, an up-and-coming symbolic method symbolic aggregate approximation(SAX) is applied to the similarity study of 96-point load data. The Euclidean distance algorithm is used to measure the similarity of time series, and the load data are cleaned according to the fitted curves obtained by adjusting similar sequences weighted by similarity. The experimental results show that the method has adequate accuracy and low computational complexity.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
