Attitude Control of a Helicopter

Model description: 

Attached image depicts the model of this helicopter. The dynamics of this model simulate the attitude dynamics of a helicopter. Using Lagrange’s equations, one may readily show that the dynamical model of this system is given by:

$$\begin{align*} T_p &= I_x\ddot{\phi}-(I_y-I_z)\dot{\psi}^2S_{\phi}C_{\phi}+mgAC_{\phi} \\ T_y &= (I_yS_{\phi}^2 + I_zC_{\phi}^2)\ddot{\psi} + 1(I_y - I_z)\dot{\psi}\dot{\phi}S_{\phi}C_{\phi}, \end{align*}$$

where $\phi$ is the pitch angle in radians, $\psi$ is the yaw angle in radians, $I_x, I_y, I_z$ are inertia constants about the point of rotation, m is the total mass of the system, and $T_p, T_y$ are the pitch and yaw control torques, respectively.

The above can be written in general form:

$u = M(q,\delta)\ddot{q}+C(q,\dot{q},\delta)\dot{q}+G(q,\dot{q},\delta),$

where

$M(q,\delta) = \begin{bmatrix} I_x & 0\\ 0 & I_yS_{\phi}^2+I_zC_{\phi}^2\end{bmatrix}$

$G(q,\delta)=\begin{bmatrix}mgAC_{\phi}\\0\end{bmatrix}$

$C(q,\dot{q},\delta)=\begin{bmatrix}0&-(I_y-I_z)\dot{\psi}S_{\phi}C_{\phi} \\ (I_y-I_z)\dot{\psi}S_{\phi}C_{\phi} &(I_y-I_z)\dot{\phi}S_{\phi}C_{\phi}\end{bmatrix} $

$q=\begin{bmatrix}\phi \\ \psi\end{bmatrix}$

$u=\begin{bmatrix}T_p \\ T_y\end{bmatrix}$

The nominal values of the model parameters are $m=0.5719 kg$, $A = 0.0801 m^2$, $I_x = 0.0762 kg\cdot m^2$, $I_y = 3.86\times10^{-4}kg\cdot m^2$, $I_z = 0.0766 kg\cdot m^2,$ and $g = 9.81 m/sec^2$. It is straight forward to show that $[\dot{M}(q)-2C(q,\dot{q})]$ is skew symmetric.

Type: 

Form: 

Time domain: 

Attachment: 

Publication details: 

TitleRobust Control of Uncertain Nonlinear Mechanical Systems Using a High Gain Observer
Publication TypeConference Paper
Year of Publication2000
AuthorsZenieht, Salah, and Elshafe Abdel Latif
Conference NameProceedings of the American Control Conference Chicago
Date Published06/2000
PublisherIEEE
Conference LocationChicago, Illinois
ISBN Number0-7803-5519-9
Accession Number6795603
Keywordsaircraft control, convergence, helicopters, nonlinear systems, observers, robust control, singularly perturbed systems, state feedback, tracking, uncertain systems
AbstractWe consider a general class of nonlinear uncertain mechanical systems. All uncertain terms belong to a known, nonempty compact set. A previously derived $r-\alpha$ robust, tracker that, guaraatees global exponential convergence of the tracling error via state feedback is extended to achieve semi-global tracking using output feedback. The proposed controller has the same structure as the state feedback controller, however, all missing states are estimated using a high gain observer. The high gain observer proposed in this paper generalizes an existing observer in the literature for SISO systems to MIMO systems. The interconnection between the observer dynamics and the nonlinear mechanical dynamics are cast into a singularly perturbed system. This technique proves that the full order system closely approximates the behavior of the reduced order system, in this case a full state feedback. Simulation results for the attitude stabilization of a helicopter model are also included for illustration.
DOI10.1109/ACC.2000.879245