Based on ANN and many-objective optimization to improve the performance and economy of village houses in Chinese cold regions

Juanli Guo,Jian Zhou,Mingchen Li,Siao Lu
DOI: https://doi.org/10.1080/19401493.2023.2183259
2023-03-01
Journal of Building Performance Simulation
Abstract:The number of studies considering building performance optimization (BPO) in the building design phase is steadily growing, but many of the existing studies do not consider the applicability of many-objective optimization algorithms when increasing the objective dimensions. This article first compares the NSGA-II, IDBEA, MSOPS-II, and NSGA-III algorithms. Then, the algorithm most suitable for many-objective optimization is combined with Artificial natural network(ANN) and TOPSIS-AHP to complete the optimization of four dimensions of building energy consumption (EC), useful daylight illuminance (UDI), comfort time ratio (CTR) and energy-saving envelope cost (ESEC) for village houses in cold regions of China. The results show that the NSGA-III algorithm performs well in terms of convergence speed, convergence, diversity, and uniformity when solving many-objective problems compared to the other three algorithms. Finally, four optimization strategies were selected using the TOPSIS-AHP method.
construction & building technology
What problem does this paper attempt to address?