Researchers optimize building energy costs using Bayesian optimization for MPC controllers

Researchers have developed a method to automatically tune Model Predictive Control (MPC) systems for minimizing electricity costs in buildings. By employing Constrained Bayesian Optimization (CONFIG), the system significantly outperforms traditional controllers. In a case study, the optimized MPC reduced electricity expenses by 26.90% compared to a rule-based approach and 17.46% versus a manually tuned MPC. The study also indicated that optimal selection of demand-side management programs could lead to monthly savings of up to 20.18%. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Automated optimization of building energy systems could lead to significant cost savings and improved grid management.

RANK_REASON This is a research paper detailing a new method for optimizing building energy control systems. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• arXiv
• Constrained Bayesian Optimization
• CONFIG
• Model Predictive Control
• MPC
• Jiarui Yu

• paper
• other