Determining predictor variables of HC, CO, and CO2 emissions using decision tree models. (November 2019)

• Record Type: Journal Article
• Authors: Larasati, A; Chen, Y W; Hajji, A M; Mahardika, A F P; Darmawan, V E B
• Is Part Of: IOP conference series. Volume 669(2019)
• Publication Date: 2019-11
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/669/1/012032
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: HC, CO, and CO2 are three gasses emitted from heavy duty construction equipment, such as backhoe that contain carbon and hugely affect the environment. This research aims to determine predictor variables of HC, CO, and CO2 emissions level using decision tree. The emissions level is categorized into three classes: low, medium, and high. The predictor variables decision tree are related to the backhoe operation and specification, including backhoe type, engine technology tier, RPM, MAP, horsepower, backhoe age, and intake temperature. This study runs 12 cycles for each model to classify HC, CO, and CO2 level. The results indicate that each gas has a different order of the important level for the predictor variable. Decision tree model to classify HC emission level show the top three most important predictor variables are RPM, temperature, and MAP. On the other hand, the top three most important variables to classify CO are backhoe type, RPM, and temperature. The last decision tree model to classify CO2 level show RPM, MAP, and backhoe type as the most important predictor variables.