Methods for the handling with fuzzy knowledge and scaling problems in the modeling of technical systems using the example of large industrial brakes
- © Max Sieting
The central aim of the project is the development and investigation of methods for the generation, transformation and transfer of knowledge from small experiments to the modelling of complex real systems. The example is the brake system of commercial vehicles. The reason for the use of this example is: the direct use of design parameters from small experiments for modelling the complex real system is not always possible because of differences between measured values from small test and real brake systems. Many parameter specifications are fuzzy for reaching the research aim that three test facilities should be used. The differences between the test plants are at the level of details. The test facilities are a test bed for parts of brake pads, a hydraulic road simulator and a real vehicle. The project starts with the modelling as the basis for calculation methods for transforming the experimental results in the real system. A comparison between the measured results and the models will be used for optimizing the scaling modelling methods. The tests of this project are the basis of the real investigation of the dynamic of knowledge in technical sciences using a concrete technical problem. The main task in this project will be the systematic investigation of the whole process of designing, modelling and examining brakes and the relevant modelling and scaling laws. Another aim is the search for answers to the following questions:
- In which way do we acquire the knowledge used?
- What is the influence of the system environment on the acquisition of knowledge?
- Which parameters will not be used for the modelling? What is the reason?
- How accurately do the developed models describe the experiments?
- Is there non-measurable expert knowledge which is not used for the development of the test facility and the real brake?
- What is the influence of the non-measurable expert knowledge on the quality of the model? How it is possible to model this non-measurable expert knowledge?
Prof. Dr.-Ing. Henning J. Meyer , department Construction of Machine Systems, TU Berlin
Julius Jenek, M. Sc. , department Construction of Machine Systems, TU Berlin