Automated Methodology for Startability Optimization and Emissions Reduction

The objective of this thesis is to evaluate the effectiveness of automated methodologies for calibration map correction, with particular focus on startability optimization and emissions reduction. To achieve this goal, a MATLAB-based tool was developed to import calibration files and data acquisition files. The tool processes time-varying signals recorded during vehicle tests and automatically corrects the values of the calibration maps with respect to a user-defined lambda target map, generated within a dedicated section of the software. After development, the tool was validated through a vehicle testing campaign. It was subsequently applied to optimize the startability calibration of the Ducati V2 engine with the aim of reducing pollutant emissions. To enable an objective comparison with the traditional calibration approach, the original homologation calibration file was imported into the tool. The startability maps were then optimized exclusively using the numerical values automatically generated by the algorithm based on the lambda target map, without any manual intervention from the calibrator. Once a sufficiently robust calibration capable of accurately tracking the lambda target trajectory was obtained — defined to maximize the effectiveness of the Secondary Air Injection strategy — WMTC emissions tests were conducted according to current motorcycle homologation regulations (Euro 5+). The tests were first performed using the original homologation calibration and subsequently using the optimized calibration obtained through the automated methodology. The results showed significant reductions in the main regulated pollutants, including HC, CO and NOx, during the startability phase. Overall, the study demonstrates that algorithm-based calibration map correction represents an effective approach for improving calibration efficiency, reducing development time and costs, and minimizing subjectivity in the calibration process.