Outliers detection in Unmanned Aerial System data

Operation of many dynamic systems is connected with data processing and analysis. Presence of single value that is significantly different from other data in the set is one of the main problems of data processing. Presence of outliers in Unmanned Aerial System (UAS) data indicates about system unreliability that can be considered dangerous for all airspace users and reduce the safety of aviation. A set of simple statistical tests for outliers detection in UAS data was considered in the current study. UAS angular data has a cyclic scale and requires a specific approach for its processing. Parametric and non-parametric methods have been studied to detect outliers in UAS parameters with a cyclic scale. Results of k-sigma test for outliers detection in vehicle heading data gives maximal probability of correct detection and minimal values for false alarm and outlier omission.