Evaluation of Aircraft Sensor Failures Effects Using the Artificial Immune System Paradigm

Abstract

This paper presents the development of simple models for the assessment of flight envelope reduction under aircraft sensor failures within the artificial immune system paradigm. The proposed methodology is expected to facilitate the design of on-board augmentation systems increasing aircraft survivability and improving operation safety. It demonstrates and exploits the capability of the artificial immune system to not only detect and identify aircraft subsystem abnormal conditions but also evaluate their impact and consequences. An artificial immune system built through simulation for a fighter aircraft is used in conjunction with a hierarchical multi-self strategy for estimating ranges of flight envelope relevant variables at post-failure conditions affecting angular rate sensors that are needed within the control augmentation system. Failure-specific algorithms correlated with the characteristics and dimensionality of self projections are developed for roll, pitch, and yaw rate sensors affected by bias. Metrics for the performance evaluation of the proposed approach are defined and used for successful demonstration in a motion-based flight simulator.