GT01 - Forecasting Solar Power Output Using Deep Learning
Develop and compare models like LSTM or CNN to predict short-term (hourly) solar PV generation.

Gabriel Alves Ribeiro <gar@poli.br>,
Luan Lima da Silva <lls4@poli.br>,

PROVA-1 = Nota 10,0

​

GT02 - Wind Power Generation Forecasting with Machine Learning
Apply random forests or gradient boosting to improve short-term (hourly) wind energy forecasts.

JOSE ALVES CABRAL JUNIOR <jacj@poli.br>,
GABRIEL HENRIQUE LINS RAMOS <ghlr@poli.br>,

PROVA-1 = Nota 10,0

​

GT03 - Optimizing BESS Dispatch in Microgrids Using Reinforcement Learning
Model a grid environment where an RL agent learns to schedule charging/discharging for cost minimization.

​

GT04 - Load Forecasting in Hybrid Generation Systems (PV + Wind + BESS)
Use supervised learning to predict future demand and optimize local energy generation planning.

KAYLANE GEORGINA SILVA DE ABREU <kgsda@poli.br>,

ALBERES CESAR GUEDES DE AMORIM <acga1@poli.br>,

PROVA-1 = Nota 7,0

obs: a aluna Kaylane não compareceu na PROVA-1

​​

GT05 - Energy Storage Sizing for Renewable Plants Using AI-Based Optimization
Use genetic algorithms or neural nets to optimize BESS size and cost for hybrid renewable plants.

MARCOS CARNEIRO FERREIRA NETO <mcfn@poli.br>,​

PROVA-1 = Nota 5,0

​

GT06 - BESS Life Cycle Cost Estimation Using Supervised Learning
Develop a model to forecast the total cost of ownership for BESS under different operational profiles.

​

GT07 - Short-Term Price Forecasting in Electricity Markets Using Machine Learning
Compare ARIMA, XGBoost, and LSTM for forecasting day-ahead market prices.

CARLOS GABRIEL MONTEIRO DIAS <cgmd@poli.br>,

GABRIEL ROGER DE LIRA LIMA <grll@poli.br>,

PROVA-1 = Nota 6,0

obs: o aluno Carlos Gabriel não compareceu na PROVA-1

​

GT08 - Electricity Arbitrage Strategy Using AI in BESS-Operated Systems
Use AI to decide when to buy/store or sell electricity for profit from a BESS-equipped asset.

ADRIEL FELIPE DE LIMA <afl2@poli.br>,

MARIA CLARA BARBOZA CAVALCANTI <mcbc@poli.br>,

PROVA-1 = Nota 10,0

​

GT09 - Customer Load Profile Clustering for Retail Tariff Design
Use unsupervised learning (e.g., K-means) to identify consumer types and propose tailored tariffs.

JULIANNA MARIA DOS REIS MURTA DA SILVA <jmdrms@poli.br>,

JOAO VICTOR EMANUEL DE MELO NOGUEIRA <jvemn@poli.br>,

PROVA-1 = Nota 7,0

​

GT10 - Reinforcement Learning for BESS Participation in Frequency Regulation Markets
Train agents to maximize revenue by providing ancillary services.

​

GT11 - Real-Time Bidding Strategy for Virtual Power Plants Using Deep RL
Simulate a VPP with multiple assets and use Deep Q-Learning to develop competitive bidding policies.

LUIS GUSTAVO GERMANO MATOS <lggm@poli.br>,

​

GT12 - Personalized Plans to Sell Energy for Retail Clients Using AI
Train individualized consumption models using historical smart meter data.

MILLANY DA SILVA COUTINHO <msc7@poli.br>,
CAIO JOSUEL DE LIMA <cjl@poli.br>,

PROVA-1 = 6,0 Profissional Externo = SIM

​