Small Rotor Uav Control Maneuvers

Compared to its full-size counterpart, small Unmanned Aerial Vehicle (UAV) is much cheaper, easier to carry and more agile, and especially the rotary-wing class, with its ability of hovering, taking off and landing vertically, is very suitable for flying in low-altitude and cramped space, so it is expected to be used widely both in military and civilian domain in the future.This paper is mostly concerned with controller design for UAV based on its six degree of freedom dynamics. First, it gives a brief introduction to the main parts comprising UAV platform and the function of each part, and then analyzes the characteristics of the system and difficulty in designing a controller. After that, a comprehensive review and comparison is made on various modeling and control techniques.About the control law design, a control structure with inner and outer loop separation is introduced. Outer loop and inner loop is respectively about the control of position and attitude, also the time scale of the former is slower than the latter. Using this structure, Double Predictive Proportional and Integral (DPPI) controller is selected both for the control of inner loop and outer loop, afterwards, Triple PPI, which is proposed instead of DPPI for outer loop, does not need speed signal, but can only achieve high precision set-point tacking and shows low accuracy in following dynamic path. Considering the importance of speed signal, a detailed discussion is carried out on how to design Linear Extended State Observer (LESO) to estimate speed, internal and external disturbance in presence of measurement noise, Active Disturbance Rejection Control (ADRC) based on LESO is then used for outer loop control. At last, a control structure without time scale separation is introduced, and adaptive Command Filtered Backstepping (CFBS) is devised to estimate the mass of UAV, stability proven using Lyapunov Function.All the proposed control laws are tested in MATLAB/SIMULINK and the results show, on the whole, they can effectively deal with mass variation, wind disturbance and input delay with good tracking performance, although each approach has its own flaws and can be chosen according to characteristics of the real platforms. The LESO designed is able to estimate speed and disturbance with good accuracy while attenuating noise. Unlike many precious methods, this paper manages to maintain a relatively simple design and tuning procedures while solving the problems above.