Optimal Chiller Loading by Team Particle Swarm Algorithm for Reducing Energy Consumption

Generally, huge energy consumption is required in the operation of the multi-chiller system in air-conditioning system. Concerning minimizing energy consumption, both Lagrangian method and genetic algorithm have been applied to optimize the partial loading rate in each chiller. As is demonstrated by previous studies, though the Lagrangian method could minimize energy consumption, it could not effectively execute convergence at low demands. And despite its capability to execute convergence at low demands, the genetic algorithm may not get the minimum energy consumption solution as Lagrange method did of solving the optimal chiller loading problem. As an efficient method, the particle swarm algorithm has been proposed to solving continuous parameters optimization problems. This paper applies particle swarm algorithm to minimize energy consumption of multi-chiller system. The objective function is energy consumption and the optimum parameter is the partial loading ratio of each chiller. To further testify the feasibility of the proposed method, the paper adopts two case studies to compare the results of the developed optimal model with Lagrangian method and genetic algorithm. The result of the two case studies shows that the particle swarm algorithm outperforms the genetic algorithm not only in overcoming the divergence of Lagrangian method occurring at low demands, but also in getting the minimum energy consumption solution of solving the optimal chiller loading problem.