Artificial Intelligence
Test data as the basis for Artificial Intelligence
Testing space propulsion and hydrogen technologies is a challenging task. Artificial Intelligence (AI) is becoming increasingly important in order to accelerate the development of new generations of propulsion systems and hydrogen systems. The DLR Institute of Space Propulsion is pooling the expertise of its research departments to achieve this goal.
The potential applications of machine learning algorithms, a sub-area of AI, are particularly promising. In simple terms, machine learning means automated learning based on sample data. Thanks to decades of data collection through the operation of test benches, the DLR site in Lampoldshausen has an impressive database and is therefore ideally suited for the use of AI. These measurements are supplemented by data from simulations that are generated using computer clusters, among other things.
More efficient and safe engine development through the use of AI
AI methods not only contribute to the analysis of the test data, but can also be used to create optimal test sequences and to detect deviating sensor and system behavior in real time. Machine learning makes it possible to create an additional "employee" in the control room who supports the DLR team before and during the tests. This significantly increases the efficiency and safety of the tests.
Increasing the resilience and service life of an engine through AI control
The use of AI algorithms during the tests enables a more efficient realization of different operating points of an engine. The system can react intelligently to changes in the propulsion system and thus also minimizes the risk of damage to the engine. By integrating fatigue models for engine components into the control system, the service life of the engine can also be extended. Significant improvements can be expected in terms of safe operation, reusability and thrust control of the engines.
Knowledge transfer from the engine test benches to hydrogen research
The AI methods used in engine development will also be applied in hydrogen research on site. For example, the AI-supported optimization of tests in terms of information gain can help to make test campaigns for new technologies related to green hydrogen more efficient and sustainable.
