Robust optimization based optimal chiller loading under cooling demand uncertainty

• Published in: Applied thermal engineering Vol. 148; pp. 1081 - 1091
• Main Authors: Saeedi, Mohammadhossein, Moradi, Mahdi, Hosseini, Meysam, Emamifar, Armin, Ghadimi, Noradin
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 05-02-2019; Elsevier BV
• Subjects: Cooling; Cooling demand uncertainty; Decision analysis; Decision making; Demand; Deterministic method; Energy conservation; Energy consumption; Heat conductivity; Modelling; Nonlinear programming; Optimal chiller loading; Optimization; Parameter robustness; Power consumption; Risk; Robust control; Robust optimization approach; Robust scheduling of electrical power consumption; Uncertainty; Robust scheduling of electrical power consumption; Deterministic method; Cooling demand uncertainty; Robust optimization approach; Optimal chiller loading
• ISSN: 1359-4311; 1873-5606
• DOI: 10.1016/j.applthermaleng.2018.11.122

•Optimal scheduling of electrical power consumption in multi-chiller system is studied.•Robust optimization approach is proposed for uncertainty modeling of cooling demand.•Robust scheduling of electrical power consumption in multi-chiller system is obtained.•Robust optimization based results are compared with deterministic method.
Optimization of electrical power consumption in multi-chiller system leads to save more energy in building or industrial locations. Also, this optimization problem is one of the most important issues in multi-chiller system. Furthermore, uncertainty modeling of cooling demand is necessary because of variation in cooling demand should be considered. Therefore, this work proposes a robust optimization approach for uncertainty modeling of cooling demand in order to obtain robust chiller loading in the uncertain environment which cooling demand is supplied by multi-chiller system. Minimizing of electrical power consumption in multi-chiller system is considered as objective function. The proposed robust scheduling of multi-chiller system is modeled as non-linear programming which is solved via CONOPT solver under General Algebraic Modeling System (GAMS) optimization software. The proposed optimization model is studied in the deterministic and robust optimization strategies and obtained results are compared with each other. Also, the effects of changes in the robust control parameter are analyzed on optimal chiller loading which decision maker can select the decision without risk as risk-neutral strategy via deterministic method or the most robust decision as risk-averse strategy via robust optimization approach. Comparison results show that capability of proposed approach in the uncertain environment.