Road Test Emulation – A Data-Driven Calibration Approach

The resource-intensive process of road testing constitutes an elementary part of the development of vehicle software. A significant proportion of explorative tests and adjustments for use in service are conducted during the vehicle test phase. However, the observed trends of decreasing development cycles and increasing system complexity generate a field of conflicts. In order to address this issue, this paper proposes road test emulation as a data-driven approach for continuously adapting vehicle software to the evolving overall system. Therefore, test scenarios equivalent to in-service conditions are determined based on customer data and test drives. These test scenarios enable an emulation of road testing and the analysis of the system in a real-world operational context from the early stages of the product development process. Sensitivity analyses are utilised to generate system-specific data for the vehicle under development, which is then employed to define initial parameter settings of the software. The example used is that of the operating strategy of a performance hybrid vehicle. This process results in the definition of an efficiency- and comfort-driven parameter set for the electric drive control, which is compared to a reference in various test scenarios. The method presented enables the data-driven generation of customer-oriented initial parameter sets and the quantification of the effects of comfort adjustments on the efficiency of the system in real-world operational contexts. The utilisation of software-in-the-loop simulations facilitates continuous implementation and iterative adjustments of the parameter sets and the vehicle software throughout the product development process. The early consideration of in-service operations facilitates a shift-left in development activities. This ensures the availability of sufficient resources during road testing for the purposes of final calibration and testing of extreme conditions. Keywords: Shift-Left, Customer Data, Operating Strategy, Hybrid Vehicle