The aim of this study is to present an optimal method for load allocation in a multi-chiller system. By using the processed data from the historical operation data of the system

the Apriori frequent item set algorithm is used to mine the association rules among the operation variables and minimum energy consumption of each chiller under different operation conditions in stages. An optimal load allocation strategy combining a particle swarm optimization algorithm and association rules is proposed to minimize the total operation energy consumption of the chillers. The simulation results show that the optimal load allocation strategy can effectively reduce the total energy consumption of the chillers by optimizing the on-off and load ratio of the chiller. Compared with the original operation strategy

the total operation energy consumption of the system can be reduced by approximately 12.5%

and it is confirmed that the optimal load allocation strategy can save energy effectively.