Development of a Numerical Methodology for the Generation of Driving Scenarios for Propulsion System Virtual Testing

Fuel consumption assessments are typically performed with reference to regulatory test driving cycles such as the Worldwide Harmonised Light Vehicles Test Cycle (WLTC). Although these cycles represent a standard at the type-approval level, they do not capture all the real-world driving conditions which may be relevant to consider before introducing new vehicles on the market. Moreover, the European Commission has recently introduced the Real Driving Emissions (RDE) legislation, which sets emission standards to be met during road testing. This represents a massive challenge for the automotive industry, which needs to guarantee that vehicles comply with emissions standards under real driving conditions. To this end, several experimental roadside tests are performed with prototype vehicles in order to assess fuel consumption and emissions but also to validate control systems or to establish durability requirements for particular components. In this way, real-world variables such as type of scenario, road layout, traffic and driver behaviour can be taken into account. On the other hand, conducting experimental tests is obviously time-consuming, expensive, and not feasible in the early stages of vehicle design. Within this framework, this master thesis, developed in collaboration with POWERTECH Engineering S.r.l. and Maserati S.p.a., aims to establish a numerical methodology for generating real-world driving scenarios in order to test the propulsion system in an integrated simulation environment. First of all, routes from experimental roadside acquisitions were reproduced in a virtual environment. In this regard, four types of road layout were studied: highway, extra-urban, mountain and urban scenarios. Several aspects of real-world driving have been taken into accounts, such as speed limits, road signs, traffic congestion level and driver behaviour in terms of aggressiveness. Then, a vehicle model was developed to assess fuel economy and powertrain operating conditions within the previously generated virtual driving scenarios. Moreover, a statistical validation analysis was performed in order to achieve satisfactory statistical reliability between driving cycles coming from experimental acquisitions and virtual generation methodology. Finally, a sensitivity analysis was performed to assess the impact of traffic congestion and driver aggressiveness level within the urban scenario. To conclude, the methodology developed in this thesis showed to be a tool for the early estimation of propulsion system operating points in a virtual real-world driving scenario.