Abstract

In this paper, we consider the optimal scheduling problem for a campus central plant equipped with a bank of multiple electrical chillers and a Thermal Energy Storage (TES). Typically, the chillers are operated in ON/OFF modes to charge the TES and supply chilled water to the campus. A bilinear model is established to describe the system dynamics. A model predictive control (MPC) problem is formulated to obtain optimal set-points to satisfy the campus cooling demands and minimize daily electricity costs. At each time step, the MPC problem is represented as a large-scale mixed integer nonlinear programming (MINLP) problem. We propose a heuristic algorithm to search for suboptimal solutions to the MINLP problem based on mixed integer linear programming (MILP), where the system dynamics is linearized along the simulated trajectories of the system. Simulation results show good performance and computational tractability of the proposed algorithm.

Highlights

  • For a campus with a large number of buildings, a central chiller plant is commonly used to serve its cooling loads

  • To generate optimal ON/OFF sequences for chillers, a mixed integer nonlinear programming (MINLP) problem needs to be solved at each time step for the model predictive control (MPC) scheme

  • We propose a mixed integer linear programming (MILP)-based heuristic algorithm to search for the suboptimal solutions of MINLP problem

Read more

Summary

INTRODUCTION

For a campus with a large number of buildings, a central chiller plant is commonly used to serve its cooling loads. A bank of multiple chillers is usually installed at the central plant to meet various cooling demands with high operational flexibility. In many practical retro-fit cases, chiller operation decisions are ON/OFF scheduling sequences For such problems the optimization/control framework proposed in above works would not directly apply. To generate optimal ON/OFF sequences for chillers, a mixed integer nonlinear programming (MINLP) problem needs to be solved at each time step for the MPC scheme. This MINLP problem involves considerable nonlinearity and a large number of integer variables, finding a global optimal solution is nontrivial and computationally expensive.

Nomenclature
Chiller bank model
TES model
Campus load model
Model summary
MODEL PREDICTIVE CONTROL PROBLEM
Deciding TES operation mode profile
Linearizing system dynamics
Formulating an MILP problem
Generating nominal trajectories for linearization
System setup
Two scheduling strategies
Comparison of two strategies
CONCLUSIONS
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call