Dr Ralph Evins | Energy Systems and Sustainable Cities group
The BESOS platform
Building and Energy Simulation, Optimization and Surrogate-modelling
The Building and Energy Simulation Optimization and Surrogate-modelling (BESOS) platform is a cloud-based portal that will make modular, reusable software components for integrated energy systems analysis available to researchers. Buildings, renewable energy generation and storage technologies and associated energy systems all pose complex, interacting design and operational challenges. Finding high-performing solutions to these problems requires a new generation of computational tools, blending aspects of simulation, optimization, machine learning and visualization.
The core BESOS modules will include:
- An 'energy hub' modelling framework (building on PyEHub) that solves mixed-integer linear programmes describing demand and supply balancing.
- A machine learning toolbox (implemented with TensorFlow) for fitting 'surrogate models' that approximate the behaviour of complex simulators.
- An optimization framework (based on Platypus).
- EnergyPlus model execution.
BESOS modules will be accessible in three ways:
- As open-source Python code in a repository, for users to run locally.
- Via an API for integration in existing workflows, executed on ComputeCanada hardware.
- Via Jupyter Notebooks directly within the portal, giving ease of use for non-programmers.
The project is led by Dr Ralph Evins (Imperial College London, ETH Zurich), who's Energy Systems and Sustainable Cities research group is pioneering the use of advanced computational techniques to deliver the low-energy buildings, cities and energy systems of the future. BESOS builds on the previous Holistic Urban Energy Simulation (HUES) platform. Research teams who will collaborate on developing components of the platform include:
- Dr Jean Duquette (Carleton) will apply the portal to district systems, renewable energy integration, waste heat recovery and energy storage problems.