Customer phase identification based on smart meter measurements and clustering technique

• Grant number: 25/12881-9
• Support Opportunities: Scholarships abroad - Research Internship - Scientific Initiation
• Start date: December 01, 2025
• End date: February 28, 2026
• Field of knowledge: Engineering - Electrical Engineering - Power Systems
• Principal Investigator: Fernanda Caseño Trindade Arioli
• Grantee: Pedro Arrelaro Ribeiro
• Supervisor: Mariana Resener
• Host Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Institution abroad: Simon Fraser University, Surrey Campus, Canada
• Associated to the scholarship: 25/00865-9 - Regression Analysis Applied to Time-Series Data of Smart Meters for Identifying Energy Theft in Electric Power Distribution Systems, BP.IC

Abstract

Electric power distribution systems are undergoing a paradigm shift, moving from historically passive networks with limited observability toward more active and responsive infrastructures. In this context, this project proposes a data-driven approach to address a key problem in distribution network modeling: the identification of customer-to-phase connectivity. Leveraging real smart meter datasets provided by a Canadian research partner, containing voltage and power measurements at 15-minute and hourly intervals, the project will implement and evaluate clustering algorithms, specifically k-means, to group customer units based on voltage magnitude similarity. The research will be conducted over a three-month period and will involve data processing, algorithm development, result evaluation, and dissemination through technical reporting and academic publication. Beyond its technical contribution, the internship will enhance a related Brazilian research initiative on energy theft detection by integrating real measurement data into its models, thereby improving their accuracy, robustness, and practical applicability. Overall, this project advances the development of smart grid analytics and contributes to the modernization of distribution system monitoring and planning.