System Analysis & Control

Due to the complex physical processes within the components of the engine and the interactions between the components, liquid rocket engines have a multi-layered, sometimes highly dynamic system behavior. During operation, the conditioning of the engine, the ignition of the pre- and main combustion chambers and the start-up of the turbopumps must therefore be closely coordinated and operating point changes must be carried out precisely. Optimum control and regulation can not only guarantee propulsion performance, but also significantly increase the service life and thus make a major contribution to the cost-efficient operation of reusable engines. Rocket engines are susceptible to a large number of faults. Due to the need to minimize weight, the engines are operated at the limits of what is technically possible, and abnormal behaviour can have devastating consequences. The immense costs associated with the loss of the launcher or a test stand demonstrate the importance of a suitable condition monitoring system. Such a system must detect and assess anomalies in real time based on the available sensor data, for example to trigger a necessary emergency shutdown or reconfiguration.

Modern IT methods are used, in particular machine learning techniques. The focus is on validated physical and data-based models for system description and analysis on the one hand, and on the other, the combination of advanced control technology with artificial intelligence methods for efficient and safe operation. The research work is highly application-oriented and is embedded in a national and international network of leading universities, research institutions and industrial companies.

The activities include in particular:

• Description of the system and individual components on the basis of physical and data-based models,
• Comparative consideration of different architectures and multidisciplinary design optimization,
• Design of optimal control systems and robust controls for the transient and stationary operating range,
• Development of condition monitoring systems and fault-tolerant control.

Among other things, the research group supports the implementation of engine demonstrators such as LUMEN through design optimization, the operation of which in turn forms the basis for model validation and evaluation of the intelligent control systems developed. In addition to being used for rocket engines, the methods developed are also used on the (large-scale) test benches and increase the efficiency and safety of the test infrastructure.