Building a Predictive Model to Estimate NOx Emission Pollutant of Backhoe Equipment

. NOx is one of emission pollutant resulted from Backhoe equipment. This research aims to build a predictive model to estimate NOx pollutant released by backhoe equipment using Support Vector Machine model. Two type of kernel types (radial basis function and linear kernel types) are compared. The study runs the model several time to maximize the accuracy of SVM by finding the optimized parameter, which includes C, ε, and γ. The results show that radial basis function kernel type provide higher accuracy than linear kernel type. In addition, this study also conclude that higher C and γ parameter results in much lower mean absolute error value. However, it requires much longer calculation time. The SVM predictive model also show that the significant factors to predict NOx emission are MAP, RPM, backhoe type and the intake temperature.