GridSFM predicts optimal power flow efficiently across large grids
GridSFM: A Foundation Model for Solving AC Optimal Power Flow
Machine Learning
Summary
Finding the best way to send electricity through complex power grids is hard because the math can be tricky and solutions change depending on the grid's layout. The authors created GridSFM, a large neural network trained on many grid setups that can quickly predict good solutions for big power grids it has never seen before. They improved it further by combining it with physics knowledge to fine-tune the predictions on new grid layouts, even with very little new data. To handle mathematical challenges caused by disconnected possible solutions, they changed the problem representation so their model can learn well and provide accurate results.
What this means in practice
- •For power grid operators: Generate fast and accurate initial solutions for large-scale electricity flow optimization to reduce computation time and improve solver performance.
- •For energy system software developers: Incorporate pretrained models to enable scalable adaptation of optimization software to new or changing grid layouts with minimal additional data.
Authors
Luke Bhan, Weiwei Yang, Margaret Capetz, Baosen Zhang
Abstract
We introduce GridSFM, a framework that combines a pretrained foundation model across grid topologies with physics-informed fine-tuning for solving AC Optimal Power Flow (AC-OPF) at scale. It is a $15$ million parameter physics-inspired graph neural network pretrained across $54$ topologies of $500$ to $4{,}000$ buses. Our model attains a $2.45\%$ zero-shot generation-cost error on a $10{,}000$ bus case held-out operating conditions with no degradation as system size grows. Building on this, we pair the pretrained backbone with a physics-informed fine-tuning design based on Newton's method for power flow. With only $100$ solved instances, GridSFM adapts to unseen grids up to $10{,}000$ buses. We show it out performs single topology, dedicated neural network models that are trained more data, both in terms of cost and solver iterations when deployed as warm starting points. In designing this foundation model, we overcome the fact that the feasible set for AC-OPF can be disconnected. This is an obstruction that prevents any continuous neural network from approximating the solution map. To do so, we lift the problem and relax its constraints with logarithmically penalized slacks. We prove that the resulting elastic feasible set is contractible, that the AC-OPF minimizers remain minimizers of the elastic problem above an explicit penalty threshold, and that projecting an approximate solution back onto the AC-OPF feasible set is well posed. We release all models, data, and code so that the community can build on a shared starting point for AC-OPF.