Research On Control Technologies Over Full Flight Envelope For UAV

Nonlinear and unmodeled dynamic characteristics seriously affect flight quality and stability,while UAV flys over the flight envelope. Control law with fixed gains can’t obtain satisfying controleffect. Control theory and method applied to the sample UAV with flight state changing become the keyissues to control the sample UAV steady fly over the flight envelope. Taking the sample UAV as object,this thesis studies flight envelope deviding method and designs fuzzy clustering gain scheduling controllaw, and steady control of the sample UAV over the flight envelope is realized.First, the six DOF mathematical model is built by applying the Newton rigid body dynamics, andthe flight capabilities, handling quality and modality of the sample UAV over the flight envelope areanalyzed, which provide the basis for control law and gain scheduling method design.Second, the characteristic matrix is got in the application of the principle of linear smalldisturbance. Flight envelope of the sample UAV is partitioned automatically and the typical flight states’membership function is got while fuzzy clustering partition algorithm and clustering effectivenessevaluation index are used.Third, the robust servo LQR method and classical root locus method are used to design vertical andhorizontal control law at every typical flight state of the sample UAV. The high quality control of thesample UAV is realized with the control law which takes the angular rate as the control loop and heightand trajectory as the guiding loop.Then, gain scheduling mechanism using fuzzy clustering proved to be effective. According to theflight states membership function which is got through envelope partition, T-S fuzzy clustering gainscheduling method is designed on the based of Sigmoidal profile. The method using T-S fuzzy rulechooses true airspeed and dynamic pressure as schedulablevariable.Finally, the flight simulation over the flight envelope is done in Matlab/Simulink environment,simulation result shows that the requirement on control quality and robust properties is satisfied withT-S fuzzy clustering gain scheduling control law. Safety flight and stability control of the sample UAVare realized over the flight envelope.