An augmented group search optimization algorithm for optimal cooling-load dispatch in multi-chiller plants

• Published in: Computers & electrical engineering Vol. 85; pp. 106434 - 12
• Main Authors: Teimourzadeh, Hamid, Jabari, Farkhondeh, Mohammadi-Ivatloo, Behnam
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Ltd 01.07.2020; Elsevier BV
• Subjects: Algorithms; Augmented group search optimization (AGSO) algorithm; Chillers; Economic cooling-load dispatch; Electrical power consumption; Energy consumption; Multiple-chiller grid; Optimization; Optimization algorithms; Robustness (mathematics); Searching; Augmented group search optimization (AGSO) algorithm; Economic cooling-load dispatch; Multiple-chiller grid; Electrical power consumption
• ISSN: 0045-7906; 1879-0755
• DOI: 10.1016/j.compeleceng.2019.07.020

•AGSO algorithm is proposed and implemented on economic cooling-load dispatch of multi-chiller plants.•Energy-saving capability of AGSO based OCL strategy is proved.•Numerical results of AGSO approach are compared with best solutions of recently published algorithms.•AGSO with fast convergence finds better solutions in lower calculation times. Improper operation of a multi-chiller plant would significantly increase associated electric consumption. This paper introduces a new model to determine the optimal loading point of the chillers. To do so, optimal dispatching of multi-chiller plant is expressed as an optimization problem and associated constraints are all together accommodated in the process of optimization. The problem is tackled by a new version of a metaheuristics approach. The proposed approach is entitled as augmented group search optimization (AGSO) algorithm, which is devised to avoid drawbacks of conventional group search optimization algorithm such as trapping in the local minima. The effectiveness and robustness of the proposed approach in comparison with available methods are studied through three well-known test cases. Numerical results demonstrate that AGSO with its strong exploration capability achieves a lower energy consumption than that of recently published methods with higher convergence speed.