Using grey incidence to analyze the energy audit reports and rough set for rule extraction. Issue 7 (1st August 2016)

• Record Type: Journal Article
• Authors: Karimi, Tooraj; Forrest, Jeffrey Yi-Lin
• Is Part Of: Kybernetes. Volume 45:Issue 7(2016)
• Page Start: 1024
• Page End: 1035
• Publication Date: 2016-08-01
• Subjects: Rough set -- Energy audit -- Grey incidence -- Iran Oil Ministry; Cybernetics -- Periodicals; Systems engineering -- Periodicals; 003.505
• DOI: 10.1108/K-02-2016-0022
• Languages: English
• ISSNs: 0368-492X

Abstract

Abstract : Purpose: The purpose of this paper is to analyze the energy audit reports in order to define the most favorable factors affecting energy consumption of buildings. Since energy audit of buildings includes assessment of occupants comfort level in addition to the technical data of buildings so some rules are extracted to model the employees thermal comfort level in organization. Design/methodology/approach: Some tools of RST and GIA are used in this research to analyze the energy consumption of official buildings. "Average energy consumption of building per year" is selected as a system characteristic in GIA and as a decision attribute in RST to show the behavior of buildings energy consumption. Ten technical sequences of buildings are chosen as relevant factors of behavior and conditional attributes in GIA and RST. In order to model the employees thermal comfort level in organization by RST, ten technical attributes of buildings are selected as condition attributes and thermal comfort level of employees is selected as decision attribute. Due to the different algorithms of data complement, discretization, reduction, and rule generation, four rule models are constructed. Cross-validation is used for evaluation of the model results and the best model is chosen with 62 rules and 99.8 percent of accuracy. Findings: According to the results of GIA and RST, "Uncontrolled area of the building" has been diagnosed as the most important factor between other relevant factors/attributes and it has the greatest effect on energy consumption of building. Four rule models have been extracted from deferent decision tables in order to describe the thermal comfort level of employees in organization. The maximum number of rules relates to the conditional combination/GA model with 1263 rules and average accuracy of 99.7 percent and the minimum number relates to the conditional combination/Janson model with 62 rules and average accuracy of 99.8 percent. Research limitations/implications: The total observations for rule extraction is 81 and the results can be improved by further samples. Originality/value: It shows that "Uncontrolled area of the building" is the most important factor/attribute to define the consumption of buildings and thermal comfort level of employees in organization.