Advanced time-frequency signal processing applications
Vibrations and acoustic signals often provide a clue about the technical condition of a system or the nature of a fault. These clues are frequently hidden, however, and it is not usually possible to evaluate their true meaning by conventional means. With many projects, where the aim is to evaluate the condition of a particular system (typically something like a car transmission), the aim is to extract this hidden information in order to be able to interpret it. Priority is given to the observation of moving parts and complex signal patterns. Ideally, the process is carried out in a realistic production setting. To carry out the assessment, modular software is used with the fundamental aim of expanding the analysis to other areas. By exploiting the full potential offered by advanced methods and combining time-frequency analysis with techniques for classifying events and error categories, there are major opportunities to gain a deeper understanding of damage and monitoring processes.
Development of tomographical processes aimed at analyzing the changing state of fluid systems
The aim in this field is to map temperatures and flow patterns in fluid systems by using tomographical techniques based on acoustics. The contact-free, remote measurement of fluid properties holds tremendous potential from a technical view point, especially when approached methodically. A particular advantage with this approach is that the objects undergoing examination cannot be influenced by sensors in the field of measurement because all testing is carried out remotely. The method used is also uniquely scalable. This makes it possible to investigate and monitor many different types of objects. We are currently working on techniques for assessing problems in internal climate control (temperatures and flow areas in all kinds of enclosed areas). Another important area where this approach can be applied is process monitoring.
Pollution of the environment and human habitats is one of the major challenges of our time. There have been many investigations into the effect of individual kinds of pollution, but little has been said about how all the influences, in combination, will effect the environment in the long term. As a result, a methods-based project was started to develop and test ways to correlate the discernible effect of fine particle pollution in combination with noise pollution. The aim was to classify pollutant sources by using the acoustic recognition of emissions.
The Steinbeis Transfer Center has organized a series of regular seminars on the use of MATLAB as a means of computerized analysis, not just for training purposes, but also to demonstrate different ways to draw on and define overlaps in the process of technology transfer. At regular intervals, the power of the concept is also demonstrated at invitation-only talks.