During the development of car engines, regression models that are based on machine learning techniques are increasingly important for tasks which require a prediction of results in real-time. While the validation of a model is a key part of its identification process, existing computation-or visualization-based techniques do not adequately support all aspects of model validation. The main contribution of this paper is an interactive approach called HyperMoVal that is designed to support multiple tasks related to model validation: 1) comparing known and predicted results, 2) analyzing regions with a bad fit, 3) assessing the physical plausibility of models also outside regions covered by validation data, and 4) comparing multiple models. The key idea is to visually relate one or more n-dimensional scalar functions to known validation data within a combined visualization. HyperMoVal lays out multiple 2D and 3D sub-projections of the n-dimensional function space around a focal point. We describe how linking HyperMoVal to other views further extends the possibilities for model validation. Based on this integration, we discuss steps towards supporting the entire workflow of identifying regression models. An evaluation illustrates a typical workflow in the application context of car-engine design and reports general feedback of domain experts and users of our approach. These results indicate that our approach significantly accelerates the identification of regression models and increases the confidence in the overall engineering process.

No abstract

Starting with the motivation to invent the new standard SSP ("System Structure and Parameterization") within the Modelica Association and the need to have one more standard beyond the mature Modelica language and the already well established Functional Mockup Interface (FMI) proposed in Modelica Association (Blochwitz et al, 2011), the main use-cases are presented were SSP can help. As SSP relies on XML, the schemas and in consequence the main features for defining system structures and parameterization of models are described. The need to be able to transport complex networks of FMUs between different simulation platforms like MIL, SIL and HIL is emphasized as a motivator for SSP.A variety of prototypes are shown that support the early version of SSP. This gives a good impression how the standard can be used for quite different tasks and proofs, that system structures can be exchanged between them seamlessly.Finally the next steps for the ongoing development of SSP are outlined.

The computational simulation of the vibrational behaviour of running engines at the design stage is one of the most challenging targets in the present design analysis work of automotive engineers. In order to describe oil-lubricated body contacts (piston-liner, shaft-bearing) a detailed model of the geometrical and physical properties of the contact area is necessary. This paper presents a comprehensive simulation tool for modelling multi-body dynamics. The concept is based on dividing the non-linear mechanical system into subsystems with linear elastic behaviour and with non-linearities occurring only at the connections between these subsystems. Therefore, in the simulation model, linear elastic bodies, e.g. piston and liner, and highly non-linear connections, e.g. oil lm at the inner liner wall, are distinguished. The paper describes the mathematical modelling of body structures and the calculation of the non-linear connecting forces resulting from elastohydrodynamic contacts between these bodies and outlines the methodology of the simulation procedure. Results of parametric studies, e.g. the in uence of piston surface pro le on the contact mechanism between piston and liner, are shown. Future aspects of the application of this method to support noise optimization of internal combustion engines as well as minimization of frictional losses and wear are discussed.

No abstract