engleski

INTERACTIVE VISUAL ANALYSIS AS A SUPPORT OF OPTIMIZATION AND ANALYSIS OF INTERNAL COMBUSTION ENGINES

Increasing complexity and a large number of control parameters make the design and understanding of automotive engines (and all other complex systems) impossible without simulations today. Strict emission rules and regulations force manufacturers to design improved engines in a very short time. In order to meet those requirements, car manufacturers use simulations as a cost-efficient, and often the only possible way to design systems with desired characteristics. During the design process in car industry many different aspects of new designs are checked using simulation long before a new car is manufactured. Examples include mixture formation and combustion, engine cooling and filter regeneration, air conditioning in the passenger cabin and front shield deicing, and many others. The increasing complexity of automotive subsystems, e.g., the power train, the intake and exhaust system, or the fuel injection subsystem, also require simulation for optimization. Tuning of an injection system, for example, for modern cars is an example of a multiparameter optimization process. The operation of the injection system depends, in a very indirect way, on several parameters. Therefore, optimization by experience and intuition is usually not possible. The common practice today is to define a model, run the simulation, and study results. Engineers often use diagrams or 2D charts in order to depict various results. In case of 3D CFD, 3D visualization is also used to illustrate results. However, these visualizations are rarely interactive ; they are rendered once and explored then. If engineers have to optimize a model, numerical optimization of the underlying mathematical model is usually used. A clear hypothesis is necessary for such an approach. As data amount and complexity grows, automatic analysis methods are often not sufficient any more. In order to efficiently cope with huge and complex data sets, interactive visual analysis (IVA) tries to balance human cognition and automatic analysis. The power of human cognition is used to guide analysis. The IVA approach provides an interactive discovery framework. It helps the user in getting insight, in understanding the data and complex, often hidden, correlations and interplay between certain data dimensions. The visual information-seeking mantra – overview first, zoom in, details on demand – as defined by Shneiderman summarizes the main idea. Interactive visual analysis is much more than presentation only ; it supports users in the analysis of complex and heterogeneous data sets. It is necessary as our ability to understand and analyze data cannot cope with the always increasing pace of data generation and simulation.

interactive visual analysis, common rail injection, chain drive, EHD Bearing

nije evidentirano