LMIs in Control/Click here to continue/Applications of Linear systems/LPV state-feedback controller for Attitude and Altitude stabilization of a mass-varying quadcopter

Introduction An LMI-based state feedback approach that ensures optimum path tracking and improved steady state performance of a quadrotor in both translational and rotational movements. Quadcopter Dynamics The motion of Quad Copter in 6DOF is controlled by varying the rpm of four rotors individually, thereby changing the vertical, horizontal and rotational forces. ASSUMPTIONS: The structure is symmetric, thus the inertia matrices are diagonal. The center of mass corresponds to the origin of the physical coordinate system. A quadcopter is a rigid body. Derivations Roll motion The torque about x-axis is modeled as i.e. rolling moment. Now r (moment arm) is distance of individual forces from x-axis As the rotors 1 and 3 are placed along with body x-axis so their moment arm is zero and they do not contribute in rolling moment so Thus, when rotor 4 applies force it tends to rotate the system about x-axis opposite to curl of fingers so using the right hand rule (RHR) its torque contribution in x-axis is negative. Similarly the rotor 2 produces positive torque about x-axis according to Right Hand Rule. Pitch motion The torque about y-axis is modeled as i.e. pitching moment. Now r(moment arm) is distance of individual forces where r is the distance of individual force form vertical axis so </math> \tau_{y}=d_{f1}-d_{f3}</math> when rotor 3 applies force it tends to rotate the system about y-axis opposite to curl of fingers.so using Right Hand rule its torque contribution in y-axis is negative. Similarly the rotor 1 produces positive torque about x-axis according to Right Hand Rule. Yaw motion Finally the yawing moment is modeled as and its torque about Z axis.As the motors rotate the rotors encounter a reactive torque due to air drag on rotor blade, this reactive torque is modeled as Q.The yaw is accomplished by rotor rpm imbalance between clockwise and counter-clockwise rotating propellers. The reactive torque as a convention is taken positive for counter clockwise rotating motors/rotors and negative for clockwise rotating ones. All the four rotors contribute in yawing moment. Whereas if rotation of the ith rotor is counterclockwise and , if clockwise. So opening the summation becomes: Vertical motion The vertical motion is governed by the total thrust produced Horizontal motion The horizontal motion in the x & y-axes is accomplished by decreasing the rpm of the rotor in whose direction it is intended to move, and subsequently increasing the rpm of the rotor on the opposite side of the same arm. The force imbalance will cause the quad copter to tilt to one side and the horizontal component of the thrust force will impart horizontal linear motion to the quadcopter. If the mass of the quad copter is ‘m’, then referring to (figure 4), the linear acceleration in the horizontal x-direction can be written as: using small angle approximation and as 0 is small , State Space Representation where x(t) is state vector, y(t) is output vector and u(t) is Input or control vector. A is the system matrix B is the input matrix C is the output matrix D is the feed forward matrix Quadcopter modelling with 6 degree of freedom REQUIRED 12 STATES: parameter description x position along x axis y position along y axis z position along z axis(height) x' velocity along x axis y' velocity along y axis z' velocity along z axis Φ Roll angle θ pitch angle ψ yaw angle Φ' Roll rate θ' pitch rate ψ' yaw rate The state vector x is The Input matrix u is, , where U1 is the Total Upward Force on the quadrotor along z-axis ( T-mg) U2 is the Pitch Torque (about x-axis) U3 is the Roll Torque (about y-axis) U4 is the Yaw Torque (about z-axis) The Output matrix y is The State differential equations written in matrix form as, = + The above martices represents the equation =+ The above martices represents the equation LMI for stability Analysis Solving the above LMI yields the unknown coefficients of the feedback control. The system will be then asymptotically stable and path track will be achieved. LPV Attitude State Feedback Controller Design To design a LPV feedback control scheme for the altitude/attitude stabilization of the quadrotor aircraft, Output of the system must track r =, a reference trajectory. To achieve the objective outputs of the integrator are considered as extra state variables as In this case the error signal e=y-r ,for the outputs ,the weight functions are added to the system. The system matrices of weight functions are The dynamic of all the weight functions can be constituted as, where is the state, represents the input The system matrices of the weight function can be deducted as, , ,,and are the exogenous input and exogenous output respectively. The system differential equations with augumented ststes and weight functions are, where, ; Making the closed loop system, LMI for H∞ optimal state-feedback ; i=1 to 16 By solving the LMI,the optimal H\inf state feed-back controller with the smallest attenuation level \gamma >0 for the attitude/altitude subsystem of the mass-varying quadcopter is Conclusion This LMIs can be used model the state-feedback controller for attitude/altitude stabilization of a mass-varying quadcopter Implementation This LMI can be used in a problem and can be solved using the solvers like Yalmip,sedumi,gurobi etc,. Return to Main Page: