Surrogate-Assisted Multi-objective Particle Swarm Optimization for Building Energy Saving Design

摘要

Evolutionary optimization has been successfully used in building energy saving design due to its global search capability. However, existing algorithms generally need high computation cost because they need to evaluate individuals/solutions by a time-consuming energy consumption software. In view of this, this paper proposes a surrogate-assisted multi-objective particle swarm optimization (PSO) method for building energy efficiency design. Firstly, a management strategy of surrogate model is developed to effectively update the surrogate model; then, a multi-objective evolutionary simulation platform for the energy saving design of buildings is established by integrating the proposed multi-objective PSO algorithm with the building energy simulation software, EnergyPlus. The proposed algorithm is applied in an office building at Beijing of China. Experimental results show that it can obtain highly competitive solutions on the basis of significantly reducing the running cost.