Application of a negative selection algorithm to anomaly detection for aircraft electrical power system

The future aircraft, whether it is civil one or military one, is developing towards more electric aircraft or all electric aircraft. The electrical power system has become the most important one in various systems of aircraft. Its operation state directly affects the flight security. Detecting the anomaly of aircraft electrical power system is a general way for finding the anomalous state of this system. To trouble-shoot the latent fault and therefore guarantee the flight security, the anomaly detection is very important in the maintenance procedure. Referring to the artificial immune system idea appeared in computer security area, the Relative Hamming distance is utilized for matching calculation to generate self set and valid detector set, a real-valued negative selection algorithm is implemented for the anomaly detection of aircraft electrical power system. And a simulation calculation of the actually measured data is done. The main attention is focused on the feasibility and practicability of this algorithm applied to the anomaly detection of aircraft electrical power system. The initial results indicate that this algorithm can be feasible and effective for detecting the anomaly of aircraft electrical power system.