Edition #03

18/06/2026

Home > Analytics Report > Edition #03

Artificial intelligence is rapidly becoming a core capability in energy analytics, not only as a tool for automation, but as a new layer for modeling, reasoning, software development, and large-scale computation. This issue of the Analytics Report highlights how PSR is positioning itself at the forefront of AI applied to energy systems, combining climate scenario generation, stochastic optimization, reinforcement learning, agentic workflows, MCP-based interaction with analytical tools, AI-assisted coding, and GPU-based optimization. Together, the articles show how AI and advanced computing can support more robust methodologies, faster studies, and new ways of developing, operating, and interpreting complex energy models.

Publications

Explore the exclusive content of this edition.

Especial

Luiz Carlos da Costa, Dsc

PSR User Meeting 2026

Full coverage of the PSR User Meeting 2026, held in Foz do Iguaçu, Brazil, which gathered participants from four continents around the event’s two central themes: artificial intelligence in energy analytics and the increasing complexity of power systems. Highlights include the evolution of the SDDP Platform, client cases from around the world, and a memorable technical visit to the Itaipu hydroelectric plant — plus an invitation to the 2027 edition.

inDepth

Gabriel Vidigal, Bsc

SDDeeP: Blending optimization and reinforcement learning for hydrothermal dispatch

SDDP has been the central tool for hydrothermal dispatch planning for over three decades, but its structural assumptions shape the modeling choices available in practice. This article explores what becomes possible when reinforcement learning is paired with optimization to relax those constraints.

inBrief

Carolina Abdu, David Parrini, Msc e Ricardo Perez, Msc

Expanding the SDDP platform experience through AI: Chatbots, MCPs, and AI reasoning

A practical look at how RAG-based assistants, MCP-connected agents, and reasoning workflows can extend the SDDP Platform experience, supporting users from documentation queries and case operations to simulation execution and result analysis.

inDepth

Rafael Benchimol Klausner, Bsc

AI reasoning in math and coding: a stochastic generation-expansion solver

Translating a research paper into production software typically takes months of engineering effort. This article reports a deliberate experiment in AI-assisted development, using a stochastic generation-expansion solver as a test case for what frontier coding agents can deliver today.

inSight

Luiz Augusto Barroso, Dsc, Julio Alberto Dias, Dsc e Mario Veiga Pereira, Dsc

Integrated AI-based climate modeling and stochastic optimization

Climate change makes historical records an increasingly unreliable guide to the future. PSRCast was designed to address this challenge, generating probabilistic climate scenarios conditioned on global circulation models for energy planning, agricultural insurance, and beyond.

inDepth

Rafael Benchimol Klausner, Bsc, Raphael Sampaio, Msc e Joaquim Dias Garcia, Dsc

GPU-based algorithms for large-scale optimization

As power systems grow in complexity, the size of optimization problems is outpacing what traditional workflows were designed to handle. This article examines how GPU-based algorithms can change the computational envelope for large-scale energy optimization, with benchmarks on Brazil’s dispatch LP.

inSight

Raphael Sampaio, Msc

How AI reasoning unlocked the next generation of software engineering

From the Transformer to autonomous coding agents, this article traces eight years of AI milestones and their growing implications for software engineering practice, examining what reasoning models make possible and what disciplined prompt and context engineering demands from engineers.

inSight

Julio Alberto Dias, Dsc e Mario Veiga Pereira, Dsc

Assessing the capability of agentic AI in a Stochastic Hydrothermal Scheduling case study

Could a reasoning AI model develop effective hydrothermal operating strategies by exploring a simulation environment, without relying on traditional optimization algorithms? This article describes the experimental setup, the agent’s reasoning process, and what the results suggest.

Check out other editions

Edition #02

16/03/2026

Energy systems are evolving rapidly, and so are the analytical frameworks required to understand them. This edition of the Analytics Report brings together studies and methodological discussions on some of the key issues currently shaping the sector: uncertainty representation, storage valuation, transmission flexibility, hydrological constraints, and the growing role of advanced computational approaches in planning and decision-making. Taken together, these contributions demonstrate how analytical innovation is becoming essential to address increasingly complex energy challenges. We invite you to continue reading and explore the analytical perspectives presented throughout this edition.

Edition #01

01/07/2025

The Analytics Report (AR) is a publication by PSR that presents the latest advances in energy analytics, with a focus on models (methodologies and applications), high-performance computing, and research. Each edition provides an in-depth perspective on how PSR continuously innovates in the field of energy analytics, showcasing the development and application of advanced tools and techniques. The report serves as a platform for sharing cutting-edge insights and research that support decision-making and deepen the understanding of the energy sector.

Edition #03

Publications

Check out other editions

Read content by category