Local anomaly detection in hybrid rocket combustion tests

Local anomaly detection was applied to image data of hybrid rocket combustion tests for a better understanding of the complex flow phenomena. Novel techniques such as hybrid rockets that allow for cost reductions of space transport vehicles are of high importance in space flight. However, the combustion process in hybrid rocket engines is still a matter of ongoing research and not fully understood yet. Since 2013, combustion tests with different paraffin-based fuels have been performed at the German Aerospace Center (DLR) and the whole process has been recorded with a high-speed video camera. This has led to a huge amount of images for each test that needs to be automatically analyzed. In order to catch specific flow phenomena appearing during the combustion, potential anomalies have been detected by local outlier factor (LOF), an algorithm for local outlier detection. The choice of this particular algorithm is justified by a comparison with other established anomaly detection algorithms. Furthermore, a detailed investigation of different distance measures and an investigation of the hyperparameter choice in the LOF algorithm have been performed. As a result, valuable insights into the main phenomena appearing during the combustion of liquefying hybrid rocket fuels are obtained. In particular, fuel droplets entrained into the oxidizer flow and burning over the flame are clearly identified as outliers with respect to the main combustion process. Graphic abstract