Easy Does It_ Automated Model Input for Building MPC

Electricity use in commercial buildings is projected to increase 18% by 2040, with buildings in the Municipal, University and School sector making up 28% of that commercial space. This sector is ill-prepared to meet increasing complexity in state and local electricity rates and energy performance standards targeted toward efficiency, grid flexibility and self-consumption of on-site renewables. While Model Predictive Control could capably manage dynamic operations problems within user constraints, model setup for building Model Predictive Controllers is costly, time consuming and often inaccurate. In this Small Business Innovation Research Phase I proposal, the applicant company proposes a process to investigate the feasibility of a Python-based software application that autonomously translates user-friendly prompts and digital information into model parameter inputs for setup and calibration of two open source hybrid grey-box Model Predictive Control frameworks using primary and secondary school building reference models as a starting point. This project proposes a means for accurately and affordably collecting and updating model parameters for a Model Predictive Control to be integrated into a building automation software for small and mid-sized commercial building customers. The applicant company will bring to market a scalable, autonomous, clean building control system of hardware, software and cloud services that decreases the installed cost and complexity of Model Predictive Control for a solution serving small to mid-sized commercial building owners. This will increase energy savings by 5-25%, reduce peak demand by 20-50% and drive adoption of renewables through load shifting.