(709b) A Generalized Framework for Supervisory Economic Model Predictive Control of Energy Systems

Over the years, practical MPC and EMPC designs have been widely discussed in the literature. Qin and Badgewell [2] provide a comprehensive overview of commercial MPC technologies from the late 1990s, primarily applied in refinery, petrochemical, and chemical industries. They examined vendor approaches to aspects such as output feedback (state estimation), input parameterization, and solution methods. More recently, Badgwell and Qin [3] outlined the steps for successful MPC implementation, including justification, pretest, step test, modeling, control configuration, commissioning, post-audit, and sustainment. Research in MPC for building operations has also advanced. Drgoňa et al. [4] presented a review of MPC applied to buildings, reviewing the state estimators, software for formulating and solving MPC problems, and methods for handling parametric and non-parametric uncertainties in the literature, among others. Kumar et al. [5] developed a method to integrate a PI controller model into EMPC to reduce HVAC energy costs. Additionally, Wenzel et al. developed a hierarchical control strategy for successful optimization of the central plant, battery storage, temperature setpoints, and utility program participation in real-time for campus buildings [6].

Despite over a decade of development and deployment efforts, a systematic and generalized framework for deploying supervisory EMPC systems to energy systems has yet to be presented. Such a framework could elucidate critical concepts and components often overlooked by fundamental EMPC research, demystify the deployment of supervisory EMPC, and provide a more robust foundation for its design and development. Energy systems, in particular, require careful consideration due to the dynamic nature of energy demand, the availability of renewable energy, and economic factors, which could be addressed by incorporating energy demand forecasts, marginal greenhouse gas emissions rate, and time-varying electricity prices [7].

To address these challenges, we present a systematic, generalized framework for supervisory EMPC systems, focusing on energy systems. This framework highlights key design considerations in the data processing pipeline, algorithm execution (e.g., state estimation, EMPC, post-processing of EMPC solutions), and system scalability for practical implementation. This includes the design of data collection, handling, and pre-processing (e.g., data reconstruction and resampling), considering diverse data sources (e.g., external APIs, telemetry data, or previous EMPC solutions) and data types (e.g., change-of-value/event-based or synchronous sampling). We present our extensible data handling approach. We also address communication and processing delays, including data loading latencies and the time required for the EMPC to return setpoints and schedule them to device(s). We employ a schedule-ahead type approach to address these issues. Furthermore, we demonstrate our framework on a large-scale, real-world heat pump water heater study. We compare its performance in terms of electricity cost, greenhouse gas emissions, and occupant comfort under EMPC with that achieved under a conventional operating strategy.

References:

[1] M. Ellis, H. Durand, and P. D. Christofides, “A tutorial review of economic model predictive control methods”, Journal of Process Control, pp. 1156–1178, 2014.

[2] S. J. Qin and T. A. Badgewell, “A survey of industrial model predictive control technology”, Control Engineering Practice, vol. 11, pp. 733–764, 2003.

[3] T. A. Badgwell and S. J. Qin, “Model predictive control in practice”, In: J. Baillieul and T. Samad (eds) Encyclopedia of Systems and Control, pp. 1239–1252, 2021.

[4] J. Drgoňa, J. Arroyo, I. Cupeiro Figueroa, D. Blum, K. Arendt, D. Kim, and E. P. Ollé, J. Oravec, M. Wetter, D. L. Vrabie, and L. Helsen, “All you need to know about model predictive control for buildings”, Annual Reviews in Control, vol. 50, pp. 190–232, 2020.

[5] P. Kumar, J. B. Rawlings, and P. Carrette, “Modeling proportional-integral controllers in tracking and economic model predictive control”, Journal of Process Control, vol. 122, pp. 1–12, 2023.

[6] M. J. Wenzel, M. N. ElBsat, M. J. Ellis, M. J. Asmus, and A. J. Przybylski, “Large scale optimization problems for central energy facilities with distributed energy storage”, in Proceedings of the 5th International High Performance Buildings Conference, West Lafayette, IN, Jul. 9-12, 2018, pp. 1-10.

[7] Y. Zong, G. M. Böning, R. M. Santos, S. You, J. Hu, and X. Han, “Challenges of implementing economic model predictive control strategy for buildings interacting with smart energy systems”, Applied Thermal Engineering, vol. 114, pp. 1476-1486, 2017.