Electricity Load Profile Using Selected Clustering Methods on Residential Electricity Load Profiles

Abstract

The interest is increasing in contemplative conduct of electricity users in both the housing and retail-zone with the approach of high solution time-sequence capacity requirement data through best metering; drilling this data could be valuable from the computational aspect. One of the trendy facility is clustering, but build upon on the algorithm the solution of the data can have an relevant control on the resulting clusters. This paper shows how suited solution of power capacity portrait affects the quality of the gather process, the texture of cluster participation (profiles exhibiting similar behavior), and the ability of the clustering process. This work uses both raw data from ordinary expenditure data and counterfeit profiles. The rationale for this work is to growth the clustering of electricity load profiles to help categorize user types for charge pattern and switching, fault and fraud find  requirement-side manage and power ability measures. The vital for benchmark drilling very large data agree is how small information needs to be used to get a stable conclusion while manage solitude and preservation.