Electrical machines

This field of research focuses on the design and optimization of rotating electrical machines.

Analytical computational models form a key element of the work and are used to establish a rough design. Using numerical simulations, the intricacies of the design of the machine are analyzed at the same time as the rough design. If a machine design development stage results in the creation of a prototype, its electrical, mechanical, and thermal properties are recorded on our own machine test rig and the calculations validated.

The magnetic properties of the materials used are determined using our own testing facility for the selection and calculation of the electrical machines. Alongside research involving conventional machine designs, such as asynchronous machines and permanent magnet synchronous motors, machine concepts such as reluctance machines, cascade motors, polyphase machines and “flux switching” machines, which have recently shifted back into focus, are being investigated in greater detail. 

Mechatronic systems

Modelling and energetic balancing and optimization of the entire drive system, from energy input to the work process.

Comprehensive and holistic optimization makes it possible to improve the efficiency of the system in a way that simply cannot be achieved by the traditional approach of considering the optimum efficiency of each of the individual components.

In order to achieve this, new modelling and simulation procedures are used, together with the electric drive train and the feed-in and working machine or mechatronic installation. The analytical and numerical optimization also results in new criteria and algorithms for the design, operation, and control of the system as a whole.

Using those models as a basis, it is also possible to determine the relevant parameters for the real system by means of parameter identification and to use them for the purposes of adaptation to ensure optimal operation.

Tasks/examples:
Thanks to energetic recovery systems, it is possible to reduce the energy requirement of an elevator drive by up to 25%. The energy yield of a small hydropower plant can be increased by 20-25% through the use of an optimized, directly driven, variable-speed hydropower generator. Model-based status monitoring allows pump drives to be operated at optimum efficiency and up to 50% of operating power to be saved during partial-load operation. 

Model-based system optimization

Modelling and simulation of an electric drive train, development of robust control concepts, model-based holistic optimization of systems

The stated aim of the research is to demonstrate the potential for energy savings and enhanced dynamics by means of optimized control achieved with model-based predictive control (MPC) for electric drive systems on the one hand and, on the other hand, to operate downstream systems (e.g. pumps) at their ideal operating point without any additional sensors. In addition, the intention is to present in detail beneficial properties that are discovered by means of a combination of modelling and consideration of the system as a whole.  

Model-based predictive control has already been used successfully in the chemical process industry for decades. However, due to the high sampling rates and the large computing capacity that is therefore required, its use in power electronics systems and propulsion technology is new and is currently being researched. It is especially important that further improvements are sought with regard to modelling that takes account of realistic assumptions. This involves application-oriented research, which makes use of the smart control and observer models, performs the transition to application and demonstrates the potentials and limitations.

The detailed modelling of the system to be controlled forms a key aspect of model-based predictive control and system optimization; the results of research carried out at the institute can be referred to in this regard. In particular, the loss models from previous research projects, which have already been verified for a range of machine types, are used for this purpose. This makes it possible to include the operating losses in the quality criterion for model-based predictive control and to minimize it in a targeted manner. Alongside the machine models, converter models are also available with a range of detail levels, which can be used to attempt to reduce converter losses within the power semiconductors in an application-oriented manner. In this regard, the modelling of the complete control path also takes account of mechanically coupled systems, such as pump systems. This allows pressure or flow control to be achieved without the need for any additional sensors.

Power electronics

The aim of power electronics is to convert electrical energy into the various forms required for a diverse range of applications and consumers in a highly efficient manner and to control the power flow.

The electrical output varies across an extremely broad range, from typically less than one watt in the case of the voltage regulators used in communications electronics up to several gigawatts for high-voltage direct current (HVDC) installations. They therefore present a basic or cross-cutting technology to act as a link between energy technology and information technology. Due to the high level of efficiency that can be achieved and the excellent flexibility, the importance of power electronics is expected to increase significantly in the coming years, particularly in the field of electrical energy networks. 

Embedded systems

State-of-the-art control algorithms for electric drive systems, such as model-based predictive control, require a significant amount of processing power. The interface between the power electronics and the control procedure developed is provided by computational platforms, typically microcontrollers.

These measure actual values, compute the control in real time, and generate switching signals for the power electronics. In addition to microcontrollers, state-of-the-art system-on-a-chip (“SoC”) platforms are growing in importance. These combine microcontrollers and FPGAs (Field Programmable Gate Arrays) on a single chip. By combining FPGAs and microcontrollers, complex algorithms can be computed much more quickly. This makes it possible to fulfil real-time requests that were not possible just a few years ago. The control concepts developed for power electronics and drive systems are executed in real time through the use of computational platforms. State-of-the-art computational platforms allow for the use of high-performance, CPU intensive control procedures.