The research project Simulation-Agent at the University of Applied Sciences Offenburg is a cooperation project with the company Peter Huber Kältemaschinenbau in Offenburg. The goal of the cooperation project is the development of a simulation-supported controller technology for cross-location process optimization in international production structures.
Huber is regarded in the industry as the technology leader for high-precision temperature control solutions in research and industry. Huber products ensure precise temperature control in laboratories, pilot plants and production processes worldwide. The product range offers solutions for temperature applications from -125 to +425 °C.
In recent years, great progress has been made with regard to high performance production models. With these series systems a solid market coverage is achieved today. With customers' processes becoming increasingly specialized and, in particular, in connection with the exhaustion of processes to technical limits, more and more individualized solutions are being requested. The proportion of "non-series devices" will increase sharply in the coming years. At the same time, the internationalization of the process industry is increasing. Pharmaceutical, biotechnology or chemical companies produce worldwide and in dynamic structures. As a result, process variations often occur locally. Process applications can practically never be copied 1:1 from one device to the other, but must also be individually run in - and, above all, controlled - on site with largely identical systems and identical processes.
In general, the problem is manageable if the control environment is adapted at the same time. This is precisely where an immense effort is required today for the experimental search for optimized thermostat structures and setting parameters. This results in higher costs - both for Huber and for the customer.
One way out would be to model the "thermostat <=> process" interaction. Then the personnel effort is supported or even substituted by a technical solution. For this solution harmonized structures of plant model (Huber) and process model (customer) are required.
The task of the Offenburg University of Applied Sciences is to develop a controller-integrated simulation agent that implements the adapter function between the plant model and the process model. The main link between the database server, the simulated modules and the simulation models is a simulation adapter.
The new solution approach should make it possible:
- to port plant parameters dynamically, i.e. taking into account individual local conditions or other aspects.
- to commission individualized plants efficiently and, if possible, without explicit know-how carriers on site.
- to vary the operating parameters of plant components so that they correlate better with the process - without having to prepare this with experiments.
- to carry out a targeted plant expansion.