The performance of the pattern search direct search method in solving load estimation problems
Description
Accurately estimating load is essential for effective electric distribution planning, assets management, precise power flow predictions, accurate power losses calculations, and efficient integration of distributed energy resources. To facilitate this, a dataset was generated using Matlab to produce various simulations in the Open Electric Power Distribution System Simulator (OpenDSS), a widely used software in the electric distribution industry. These simulations were conducted on three typical distribution feeders (IEEE 13-bus, 37-bus, and 123-bus) that support studies in distribution planning, assets management, power flow predictions, power losses calculations, and distributed resource integration. The dataset includes individual demand profiles of residential, commercial, and industrial consumers specified for the three distribution feeders, comprising at least 96 distinct scenarios. An optimization method was developed using the obtained dataset, which employs the pattern search technique to estimate loads by optimizing specified objective functions and constraints. The load estimation quality was assessed for all three feeders, utilizing estimation quality indices proposed by the authors. These indices evaluated both the initial and proposed load estimation methods across the developed scenarios. Furthermore, the data provided in this article can be utilized for comparison with future load estimation studies, particularly regarding the quality of the method's results.
Files
Steps to reproduce
The steps to reproduce data are in the article submitted to data-in-brief, in the source link of the dataset, and in the github link presented below.
Institutions
Categories
Funding
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
Finance Code 001