Numerical Simulation and Optimization Comparison of energy consumption in mountain hotel by genetic algorithm and Particle swarm optimization

Creative City Design

مقالات آماده انتشار، پذیرفته شده، انتشار آنلاین از تاریخ 25 مهر 1402

نوع مقاله: Original Article

شناسه دیجیتال (DOI): 10.30495/ccd.2023.1989519.1217

نویسندگان

emad behrouzifard¹؛ mansour nikpour²؛ marjan ilbeigi^{* 3}؛ mahyar jafarpour shalmani⁴

¹PhD Candidate, Department of architecture, Bam branch, islamic azad university, Bam, Iran.

²Department of architecture, Bam branch, islamic azad university, Bam, Iran.

³Department of architecture, chalous branch, Islamic azad university, chalous, Iran.

⁴department of architecture, higher school of economics

چکیده

The aim of this research is to integrate optimization by considering effective parameters in order to decrease energy-use in hotel buildings in Damavand Mountain located in Tehran, Iran. Consequently, the EnergyPlus software is employed for the purpose of assessing energy usage and analyzing essential factors through numerical means. Subsequently, a robust artificial neural network (ANN) is developed, utilizing a multi-layer perceptron model (MLP), and undergoes a comprehensive process of training and testing to replicate building energy consumption. To assess the precision of the ANN model, two performance metrics, namely mean square error (MSE) and the coefficient of correlation (R), are meticulously scrutinized. This research used grasshopper and Lady-bug and Honeybee plugins for simulating the building which has been taken the benefit of EnergyPlus motor for evaluating energy consumption. Also, in order to perform GA optimization, Galapagos plugin has been used. Number of people, lighting density, ventilation, equipment load rate, infiltration rate, wall U-value, roof U-value and window U-value were considered as variable parameters for performing optimization. Finally, results indicated that by changing effective parameters, energy-use decreased 21.4% by GA and 15.11% by PSO. GA shows a significant performance compared to PSO algorithm. The outcomes of the calculations demonstrated that the MLP model, which was trained as part of this research, can reliably forecast energy consumption within the building. This research helps designers to integrate energy optimization for their design in order to decrease energy consumption. Further studies about the most effective parameters on optimization and reducing energy consumption is suggested.

کلیدواژه‌ها

Energy؛ optimization؛ ANN؛ genetic algorithm؛ PSO