Immunity-Based Sensor and Actuator Abnormal Condition Evaluation on a UAV Using the Partition of the Universe Approach

A general framework established for the evaluation of abnormal conditions is applied to generating an artificial immune system using an alternative clustering approach, the partition of the universe approach. The designed artificial immune system is capable of evaluating abnormal conditions affecting both actuator and sensor on a simulated unmanned aerial vehicle. The targeted actuator malfunctions affect the three primary control surfaces, where the actuator can either be locked at a particular deflection or damaged such that the actuator is partially or completely missing. The faults affecting the sensors are angular rate biases of angular rate sensor outputs which are used in the trajectory tracking control laws. The results obtained through simulation show the efficacy of the proposed methodology for evaluating abnormal conditions affecting a highly coupled dynamic system, both quickly and with a high degree of success.