A method for determining the weight of objective indoor environment and subjective response based on information theory. (January 2022)

• Record Type: Journal Article
• Authors: Zhang, Zhongchen; Geng, Yang; Wu, Xiaoying; Zhou, Hao; Lin, Borong
• Is Part Of: Building and environment. Volume 207:Part B(2022)
• Publication Date: 2022-01
• Subjects: Information gain ratio -- Subjective response -- Objective parameters -- Indoor environment -- Weight -- KANO model; Buildings -- Environmental engineering -- Periodicals; Building -- Research -- Periodicals; Constructions -- Technique de l'environnement -- Périodiques; Electronic journals; 696
• DOI: 10.1016/j.buildenv.2021.108426
• Languages: English
• ISSNs: 0360-1323

Abstract

Abstract: Buildings are supposed to create a healthy and comfortable environment for people. Considering subjective response, the problem of multi-objective optimization of building environment control becomes more complicated. The weight method can achieve the unification of optimization goals, but the core challenge is to obtain accurate weights based on a small amount of measured data. Information theory provides a way to quantify the importance of parameters. This paper established a method to calculate the weights of subjective and objective parameters quickly based on the information gain ratio, and took the difference in occupants' feedback under satisfied and unsatisfied conditions into account. Then this method was applied to a case study, two office buildings located in Shanghai, China, in which 120 samples (satisfied 104; unsatisfied 16) in Case 1 and 715 samples (satisfied 537; unsatisfied 178) in Case 2 were collected. The weight calculation time was less than 0.1s, and thus the weight of all parameters can be updated quickly when data is accumulated. Among all the parameters, the thermal satisfaction (with a weight of 0.145 in Case 1 and 0.243 in Case 2) and light satisfaction (with a weight of 0.146 in Case 1 and 0.171 in Case 2) were found to have the greatest impact on overall satisfaction. The weight calculation results under satisfied (0.57 and 0.90 of subjective parameters; 0.43 and 0.10 of objective parameters) and unsatisfied conditions (0.55 and 0.70 of subjective parameters; 0.45 and 0.30 of objective parameters) were found to be different, thus, comprehensive consideration of the difference is of great significance to engineering applications. Highlights: Proposed a calculation method of building environment control weight based on information gain ratio. Combined subjective response and objective environment parameters. Balanced the impact of satisfied and unsatisfied conditions. Reduced the weight calculation time for 715 samples to less than 0.1s. Realized weight calculation with low data and solved the problem of cold start.