Very high requirements are placed on the quality of a product during industrial manufacturing of foodstuffs such as dairy products or pharmaceutical products. To this end, raw products are refined in a multi-phase process. Amongst other processes, this is achieved by centrifuges – known as separators – which separate the substances out of the suspension using centrifugal force. Optimal operating conditions are needed to guarantee reliable separation, such as temperature, rotational speed and the composition of the raw products. However, these conditions often do not exist as the centrifuge is tied into a higher-order production process which is subject to fluctuation. To optimise work procedures, it is necessary to have extensive knowledge of the machinery and processes, which is very often only available to a limited degree. Both aspects lead to lower efficiency and yield losses. To increase the reliability and efficiency of the separation process, the separators must be able to autonomously adapt to changing conditions while at the same time having access to the required expert knowledge.
The aim of the innovation project is to develop a virtual system model for designing hardware and software solutions for intelligent centrifuges. In addition, an intelligent sensor system and a database will be devised in order to make expert knowledge available.
This involves compiling the operating conditions, such as the processing temperature and process sequences (e.g. the individual separation processes) for different raw products. These are consolidated into a virtual system model and used to develop modular software and hardware solutions for intelligent separators. On that basis, an intelligent sensor system is devised to analyse the operating conditions in the separator and identify deviations from the target state. The required expert knowledge is represented in the form of mathematical rules in a database, enabling autonomous evaluation of the sensor signals to take place. The project draws on the results of cross-sectional projects in systems engineering, intelligent networking and self-optimization. The sensor system and the database will be validated by means of a demonstration model and integrated into centrifugal separators.
The project will increase the reliability and the efficiency of the separation process. Overall, efficiency is expected to be increased by around 10%. The virtual system model forms the basis for the development of intelligent centrifugal separators and the implementation of other technologies such as remote maintenance. The sensor system and the expert system can be transferred to other industries, such as medical technology.
01 July 2014 - 30 June 2017