Projects
Battery Energy Storage Arbitrage Optimiser #
Python Pyomo LightGBM Streamlit — Feb 2026
An tool for optimising battery storage dispatch in day-ahead electricity markets. Built a linear programming(LP) optimiser using HiGHS, three forecasting models (naive lag-24, rolling 7-day average, and a LightGBM pipeline with iterative inference), a backtesting engine evaluated on European market data Electricity Maps’ API, and a deployed Streamlit dashboard with live integration across 40+ market zones.
On German day-ahead prices, the ML forecast captured 87.7% of theoretically available arbitrage value. Currently under active mentorship from a researcher in the energy optimisation field.