Research Of UAV Fault Injection And Diagnosis Real-time Simulation Platform Based On VxWorks

Unmanned aerial vehicle (UAV) is a kind of highly practical aircraft with complex structure. Inboth military and civil fields, it has important application value. However, due to the under-actuatedand strong coupling characteristics of the attitude control system, as well as the problem of datatransmission and fault correlation, thus it is hard to locate and diagnosis fault happened and difficultto guarantee the accuracy and efficiency of the whole flight system. Therefore, how to get theinfluences from fault to UAV flight control system and verify the performance of fault diagnosisstrategy is very important to improve the UAV security. Fault diagnosis and tolerant control have beena hot research topic of UAV system. Taking UAV as reference object, this thesis does some researchesin the applications of fault injection technology in flight control system. The fault injection anddiagnosis simulation platform for UAV is built, the theoretical study of fault diagnosis for UAVsensors and actuators is conducted based on sliding mode observer, the algorithm proposed is firstlyverified in Matlab/Simulink environment and then in the completed simulation platform which canoffer a real-time condition.Firstly, the study background and significance and the development situation and tendency ofthis thesis are reviewed, and then the injection modes and ways of accomplishment of fault injectionand diagnosis simulation platform are introduced with emphasis.Secondly, a fault injection real-time simulation platform as mentioned above is designed. On thehardware side, the communication frame is completed among the model real-time simulationcomputer, control and fault diagnosis, and monitoring computer. On the software side, the design ofsimulation platform and the real-time solution of control algorithm are accomplished. Simulationresults show the effectiveness of both platform and algorithm proposed.Further, aiming at the both the sensor and actuator fault of UAV, taking system disturbances intoconsideration, a fault diagnosis algorithm based on adaptive sliding mode observer is put forward, andthe simulation results show its effectiveness and the disturbance robust.Finally, the comprehensive tests of the entire simulation platform designed are carried on.