Data-Driven Modeling of Hybrid Vehicle Vibration on Roads with Low Surface Friction

This research aimed to develop a principle model with the goal of clarifying the mechanism of vibration generated in a hybrid vehicle braking on a road with partially low surface friction. One conventional analysis method involves the application of detailed CAE models that enable the implementation of precise analysis. However, issues of this method include complexity and the time required to measure the characteristics (such as stiffness) of various components. In contrast, another conventional method is to apply simple models. Although the simplicity of these models facilitates the fitting of model parameters with experimental data, such models cannot always express the necessary mechanism and it is unknown which degrees of freedom should be considered. To help resolve these issues, this article applies an experiment-based method for hybrid vehicles that adopt a different approach to these conventional methods. By combining both physical equations and measured data, this approach enables the creation of a principle model with minimum degrees of freedom using the minimum number of measurement variables. As a result, an eleven degree of freedom model can be obtained that expresses the target vibration. The physical structure of the model facilitates analysis, and it was clarified that the model can be applied effectively to the investigation of vibration causes and the design of vibration control systems.