Tools and prototypes for automating customised electricity contracts using large language models. Includes integration with power-system studies, feasibility checks, contract negotiation workflows, and case studies for residential, SME, and transmission-level maintenance planning.
This repository contains code for the paper:
Cremer, Jochen L. "Customising electricity contracts at scale with large language models." IEEE Transactions on Power Systems (2026). arXiv preprint arXiv:2505.19551
Cite this paper when using this code.
Tools and examples for household-level electricity contract planning:
- Electricityconsumption.nb — Estimates an hourly household electricity load profile based on appliance usage, EV charging, and behavioral parameters.
- checkphsicalfeasiblecontract.nb — Validates whether a proposed household contract profile is physically feasible given network constraints.
- Energysupplier.nb — A Persona that orchestrates the residential workflow, combining estimation and feasibility feedback into a contract suggestion.
- Example conversation electricity contract.nb — A complete demonstration of an interaction with a residential customer.
Tools for feasibility checks of commercial/industrial connection requests using AC power-flow simulations on the Oberrhein MV network:
- smeOberrhein.py — Python backend performing AC load-flow calculations on the 179-bus Oberrhein MV network to evaluate the feasibility of a 24-hour SME profile at a given bus location.
- MVcontractplanner.nb — Planner notebook that communicates with the Python backend, submits load profiles, and processes feasibility feedback.
- DSO.nb — Persona notebook modelling the Distribution System Operator; guides the user through required inputs and displays feasibility results.
- example conversation.nb — Full interaction example with an SME user requesting connection approval.
Tools for multi-stage security-constrained outage feasibility checking using the IEEE-118 test system:
- lineoutage.py — Python backend that evaluates planned generator outages using multi-stage SCOPF logic across the requested outage window.
- IEEE118New.xlsx & IEEE118_Load_Wind2.xlsx — Network and time-series data files used by the outage-planning model.
- outageplanner.nb — Planner notebook that calls the Python tool, providing feasibility decisions and alternative outage windows.
- TSO.nb — Persona notebook modelling the Transmission System Operator; collects user inputs and presents the resulting schedule recommendations.
- example conversation.nb — Example TSO workflow interaction exploring outage window feasibility.
This work is licensed under a
