Application of Data Analysis Techniques in the Evaluation of SIN Operation

• Grant number: 25/10042-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: August 01, 2025
• End date: July 31, 2026
• Field of knowledge: Engineering - Electrical Engineering - Power Systems
• Principal Investigator: Daniel Dotta
• Grantee: Bruno Christopher Aira Soria
• Host Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil
• Associated research grant: 21/11380-5 - CPTEn - São Paulo Center for the Study of Energy Transition, AP.CCD

Abstract

In the past decade, data analysis techniques and artificial intelligence (AI) have become a significant area of research across virtually all fields of knowledge: engineering, science, education, medicine, business, finance, law, among others. Moreover, there has been a noticeable increase in the demand for, use of, and development of AI-based tools. This surge in the application of AI techniques has been made possible by the growing availability of data, ubiquitous connectivity, high-performance computing, and a wide array of algorithms, adding a new level of efficiency and sophistication to existing technologies.Artificial intelligence is divided into several subfields according to its application type. Among them, machine learning stands out, as it aims to automate the learning process by machines. This learning occurs through algorithms that adjust the parameters of a machine's decision model to improve decision accuracy. Machine learning, in turn, is subdivided into two major groups: supervised and unsupervised learning, which differ in the degree of human involvement in the learning process.A fertile application field for AI in general, and machine learning in particular, is modern Electric Power Systems (EPS). The availability of operational data enables broader and more detailed observation of system phenomena. However, the widespread deployment of these new systems has brought with it data volumes several orders of magnitude greater than previously available.This context, known as big data, is observed in various modern systems and introduces a new challenge: the analysis and processing of such massive data volumes. Like all complex systems, EPS are operated by specialists trained over many years to develop the necessary skills for decision-making, event identification, and classification. In addition to the already demanding and complex workload handled by these specialists, there is the added introduction of inverter-based generation sources, whose dynamics are still maturing and whose behavior is not yet fully understood. Their planning and operational strategies are also under development, placing additional burdens on system operators.The development of computational techniques and tools capable of processing massive amounts of data and extracting the necessary information for operators is, therefore, an ideal field for the application of AI. In particular, machine learning offers advantages, as large datasets tend to improve algorithm accuracy, and system modeling becomes secondary in the development of such tools.