Energy Efficiency Optimization of Central Air Conditioning System by GBA-GA Algorithm

In order to improve the energy efficiency ratio (EER) of central air conditioning system in buildings under different operating conditions, a new hybrid algorithm was established to optimize the key operating parameters of a central air conditioning system in this paper. The new hybrid algorithm includes the following: key operating parameters on the EER of a central air conditioning system are selected by Grey Relational Analysis (GRA); an energy efficiency prediction model between key operating parameters and EER is established by a BP Neural Network (BPNN); association rules among the key operating parameters are obtained through the Apriori Association Rule Algorithm (AARA); the energy efficiency prediction model will be used as the fitness function of Genetic Algorithm (GA), while association rules as chromosome constraints in GA. The proposed hybrid algorithm is thus called GBA-GA algorithm. The key operation parameters of central air conditioning system are optimized by GBA-GA algorithm and an ordinary GA algorithm respectively. The results show that the EER optimized by the GBA-GA algorithm is 2.16%-6.89% higher than that by the GA algorithm under different load rates.