Which price to pay? Auto-tuning building MPC controller for optimal economic cost

TL;DR

This work tackles suboptimal building MPC performance caused by hyperparameter sensitivity and varying DSM pricing by introducing a closed-loop tuning framework based on CONFIG constrained Bayesian optimization. The approach leverages Gaussian process surrogates to model the unknown closed-loop objective $C_i(\theta,B_i)$ and constraint $g(\theta)$, enabling global optimization under black-box comfort constraints. Applied to a BOPTEST-based single-zone building with a simple QP-MPC (Mask MPC), the method demonstrates that carefully tuned, low-complexity MPC can match or exceed the performance of a delicately designed MIQP MPC while drastically reducing commissioning cost, achieving up to 26.90% monthly electricity bill savings. The study also shows that, across 12 real Belgian electricity contracts, optimal contract selection can yield up to 20.18% additional savings, highlighting the practical value of integrating contract-aware optimization with building control. The framework relies on offline simulations and a digital twin perspective, with potential for online deployment in future work to continuously adapt hyperparameters to changing prices and disturbances.

Abstract

Model predictive control (MPC) controller is considered for temperature management in buildings but its performance heavily depends on hyperparameters. Consequently, MPC necessitates meticulous hyperparameter tuning to attain optimal performance under diverse contracts. However, conventional building controller design is an open-loop process without critical hyperparameter optimization, often leading to suboptimal performance due to unexpected environmental disturbances and modeling errors. Furthermore, these hyperparameters are not adapted to different pricing schemes and may lead to non-economic operations. To address these issues, we propose an efficient performance-oriented building MPC controller tuning method based on a cutting-edge efficient constrained Bayesian optimization algorithm, CONFIG, with global optimality guarantees. We demonstrate that this technique can be applied to efficiently deal with real-world DSM program selection problems under customized black-box constraints and objectives. In this study, a simple MPC controller, which offers the advantages of reduced commissioning costs, enhanced computational efficiency, was optimized to perform on a comparable level to a delicately designed and computationally expensive MPC controller. The results also indicate that with an optimized simple MPC, the monthly electricity cost of a household can be reduced by up to 26.90% compared with the cost when controlled by a basic rule-based controller under the same constraints. Then we compared 12 real electricity contracts in Belgium for a household family with customized black-box occupant comfort constraints. The results indicate a monthly electricity bill saving up to 20.18% when the most economic contract is compared with the worst one, which again illustrates the significance of choosing a proper electricity contract.

Figures (9)

• Figure 1: An overview of the approach for developing the performance-oriented MPC controller tuning in building control
• Figure 2: An overview of the CONFIG Bayesian Optimization method. The point in the flow chart refers to a vector composed of tuning parameters. Data is a point together with its performance metric obtained from the black box function. With obtained Data, CONFIG utilized Gaussian progress (GP) to surrogate black box function behavior with Mean and Confidence. CONFIG computes the lower confidence bound of the GP and choose the next point with the smallest performance metric value on the lower confidence bound.
• Figure 3: Temperature constraints illustration
• Figure 4: Start up cost and concept of the Mask MPC
• Figure 5: Cumulative distribution function constraint of PMV