PID, LQR and LQR-PID on a quadcopter platform
read more
Citations
Swarms of unmanned aerial vehicles — A survey
Type-2 Fuzzy Logic Trajectory Tracking Control of Quadrotor VTOL Aircraft With Elliptic Membership Functions
A review of quadrotor UAV: control methodologies and performance evaluation
Quadcopter Robust Adaptive Second Order Sliding Mode Control Based on PID Sliding Surface
Dynamics modelling and linear control of quadcopter
References
Modern control engineering
Modern Control Systems
PID vs LQ control techniques applied to an indoor micro quadrotor
Robust pole assignment in linear state feedback
An integral predictive/nonlinear H∞ control structure for a quadrotor helicopter
Related Papers (5)
Frequently Asked Questions (13)
Q2. What are the future works mentioned in the paper "Pid, lqr and lqr-pid on a quadcopter platform" ?
In future work, the results will be applied in a LinkQuad quadcopter recently obtained, which will allow us to make considerations about the actual applicability of the controllers proposed, while also assessing the robustness of the methods.
Q3. Why is it necessary to use the transformation matrix to obtain the response of any movement from a?
Due to the presence of two coordinate systems, it is necessary to use the transformation matrix to obtain the response of any movement from a coordinate system (Earth-fixed frame) to the other (model-fixed frame).
Q4. What is the transfer function of the attitude?
Once obtained the transfer function of the attitude, a controller function can be defined as:C(s) = (k3s2 + k1s+ k2)s , (19)where k3, k2, k1 represents, respectively, the derivative, integral and proportional gains.
Q5. What is the effect of phase delay compensators?
The phase delay compensators are also known for their properties of assisting for a better steady response with the consequence of shifting the poles to the right side of the root locus plane.
Q6. What is the common control technique used in a quadcopter?
There are several controllers already implemented in various models of quadcopters, each one with its peculiarities: in [7] a PID control system is used based on the dynamic model taking into account the bending in the rotor and the propeller.
Q7. How many feedback gains are there in the system?
K = −3.1623× 103 −1.5070× 1031.3062 −1.3062 1.3062 −1.3062 (11)ẋ1 = x2 (12)ẋ2 = −0.0106x3 − 0.0106x4 + 0.0106x5 + 0.0106x6 (13)ẋ3 = 10x3 + 7u (14)ẋ4 = 10x4 + 7u (15)ẋ5 = 10x5 + 7u (16)ẋ6 = 10x6 + 7u (17)When a block diagram of the new expanded equations 12 to 17 is built, and the control law shown in equation 8 is applied, it becomes a closed loop system with six feedback gains and no inputs.
Q8. What is the transfer function of the closed-loop system?
the transfer function of the closed-loop system can be dened as:Fw(s) = k3s2 + k1s+ k2s3 + (10 + k3)s2 + k1s+ k2 . (20)Equalling the coefficients of the characteristic equation 20 with ITAE model optimum coefficients [12], values k1, k2 and k3 were obtained.
Q9. What is the transfer function of the vertical position controller?
Equations 21 and 23 (respectively k1 and k3 gains) were obtained using thesimplified model shown in equation 18 and later applied to the space state matrices given by equations 3 through 5.2) Vertical position controller:
Q10. What is the way to control the plant?
It is known that the LQR controllers are robust and produce a very low steady state error, but with a big transition delay and using six feedback gains, that makes them a bad choice when the system needs fast parameters update and has no direct access to all states of the plant.
Q11. What is the way to control a quadcopter?
In [8] is presented a backstepping control that uses resources from the Extended Kalman Filter, producing good results in a quadcopter designed for indoor flight.
Q12. What is the lqr function for the quadcopter?
The matrices A and B were also described in equations 3 and 4 and represent the particular dynamic model of the quadcopter LinkQuad [10].ẋ = Ax+Bu (6)According to [11] for a Linear Quadratic Regulator Controller tuning it is convenient to know a vector u that minimises the quadratic cost function presented in equation 7 which leads to the linear control law presented in equation 8.J = ∫ ∞ 0 (x.Qx+ u.ru)dt (7)u = −Kx (8)Therefore the vector K described in equation 8 need to be set in order to minimise equation 7, and for this reasonthe parameters k, p, e were obtained applying a lqr function shown in equation 9, where Q is a square matrix of sixth order described in equation 10, adjusted to provide the most efficient values and R is a unitary vector.[k, p, e] = lqr(A,B,Q,R) (9)Q = 10000000 0 0 0 0 00 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 (10)The output of the LQR function was the vector K shown in equation (11) containing the six elements that controls each state of the system individually, this makes a expansion of the state space function necessary.
Q13. What is the step response for the vertical speed controller?
Using matrix Q = diag[0, 10000, 10000] as a parameter of their algorithm, the following gains kPw, kIw and kDw were obtained:kPw = −2.3758× 103 (35)kIw = −1 (36)kDw = −173.6511 (37)Figure 7 shows the step response for the vertical speed controller.