The establishment of strict regulations and developments in engine technologies, particularly the improvement of after-treatment systems to effectively moderate emissions, are responsible for the decrease in real-world emissions. It is important to understand that emission concentrations might fluctuate greatly depending on the driving conditions. Therefore, it is crucial to have a careful understanding of emission characteristics under various operating conditions as well as steady state conditions. Due to poor replication of the driving behavior, laboratory driving cycles cannot capture the real-world driving emissions. In this current study, idle emission and running emission tests were performed to characterize the behavior of carbon monoxide (CO), carbon dioxide (CO₂), non-methane hydrocarbons (NMHC), and oxygen (O₂) emitted from various vehicle categories. Portable emissions measurement system (PEMS) was used to conduct real time emission tests. Idle emission tests were carried out on two-wheelers, three-wheelers, cars, and buses. A total of 409 idle emission samples were collected, and data was further improved for analysis. Multiple linear regression (MLR) and artificial neural network (ANN) models were developed to investigate the emission characteristics for a set of predictor variables. A two-stage least squares (2SLS) model was developed to assess the interdependency between CO₂ emission concentration and emitted O₂ emission concentration. Running emissions were carried for one diesel pickup truck. Virtual BOX (VBOX) was used to collect speed data simultaneously with real time emissions data. Emission factors were calculated and averaged across speed bins. Speed based emission models were developed to investigate the relation. The findings indicated a clear inverse relationship between vehicle speed and emissions, as higher speeds corresponded to lower emission levels.

Keywords: After-treatment systems, real-world emissions, laboratory driving cycles, and VBOX.