Data center chiller plant optimization via mixed-integer nonlinear differentiable predictive control

Ján Boldocký

doi:10.1016/j.conengprac.2026.107063

Data center chiller plant optimization via mixed-integer nonlinear differentiable predictive control

2026-05-20

0

PRE

AI

DOI：10.1016/j.conengprac.2026.107063

Original

Original request for help

Save

Abstract

En 中文

• Develops a control-oriented multi-chiller dynamics model. • Formulates mixed-integer differentiable predictive control (MI-DPC) for multi-chiller plant optimization. • Introduces binary-variance regularization to limit high frequency binary switching. • Performance comparison with rule-based control (RBC) and mixed-integer model predictive control (MI-MPC). • Open-source code for reproducible numerical experiments.

Keywords:

chiller plant optimization

mixed-integer differentiable predictive control

binary-variance regularization

data center cooling

model predictive control

Data center chiller plant optimization via mixed-integer nonlinear differentiable predictive control

Abstract

Journal

Organization