Multiobjective optimization of building energy consumption based on BIM-DB and LSSVM-NSGA-II. (20th April 2021)

• Record Type: Journal Article
• Authors: Chen, Bin; Liu, Qiong; Chen, Hongyu; Wang, Lei; Deng, Tingting; Zhang, Limao; Wu, Xianguo
• Is Part Of: Journal of cleaner production. Volume 294(2021)
• Publication Date: 2021-04-20
• Subjects: Building energy consumption -- Thermal comfort -- Multi-objective optimization -- Energy consumption simulation and prediction -- LSSVM -- NSGA-II; Factory and trade waste -- Management -- Periodicals; Manufactures -- Environmental aspects -- Periodicals; Déchets industriels -- Gestion -- Périodiques; Usines -- Aspect de l'environnement -- Périodiques; 628.5
• DOI: 10.1016/j.jclepro.2021.126153
• Languages: English
• ISSNs: 0959-6526

Abstract

Abstract: The building envelope performance significantly influences the building energy consumption, and the major influencing factors of the building envelope include the external wall U-value, external wall solar absorptance, roof U-value, roof solar absorptance, external window U-value, and window-to-wall ratio. This paper proposed a framework that integrates Building Information Modeling (BIM) with LSSVM (least square support vector machine (LSSVM) and non-dominated sorting genetic algorithm-II (NSGA-II) to study the influence of building envelope parameters on building energy consumption and discover the best design. The DesignBuilder (DB) is used to perform energy consumption simulation and orthogonal experiment to obtain the sample data set. The LSSVM is utilized to learn the data set to establish a prediction model between the building envelope and energy consumption, which is used as the fitness function of the building energy consumption in NSGA-II. With the objective of minimizing building energy consumption and maximizing the indoor thermal comfort (measured by predicted mean vote or PMV), an NSGA-II multi-objective optimization model is established. The Pareto front is obtained through calculation, and the optimal combination of building envelope parameters is discovered using the ideal point method. A case of a proposed school building in China is used to demonstrate the feasibility and effectiveness of the developed framework, the results show that: (1) The LSSVM has great ability to predict building energy consumption, achieving an accurate prediction with a goodness-of-fit of 0.9549 and an RMSE of 0.0273; (2) The optimized envelope design parameters of the proposed building are presented, including the External wall U-value is 0.2, the External wall solar absorptance is 0.55, the Roof U-value is 0.2, the External window U-value is 0.4166, the External window U-value is 1, the Window-wall ratio is 0.215. (3) The developed LSSVM-NSGA II hybrid approach can effectively improve the building's energy consumption and thermal comfort with the building energy consumption reduced by 10.6% and the thermal comfort increased by 32.2% than the initial values, respectively. The proposed LSSVM-NSGA II framework is able to optimize building envelope design parameters than its precedent, which can provide effective thinking for similar problems. Highlights: LSSVM-NSGA-II hybrid method can predict and optimize the building energy consumption. The energy consumption prediction surrogate model trained by LSSVM is an efficient NSGA-II fitness function. NSGA can effectively reduce building energy consumption and improve thermal comfort at the same time.