METHOD IN WHICH SMALL FIXED-WING UNMANNED AERIAL VEHICLE FOLLOWS PATH AND LGVF PATH-FOLLOWING CONTROLLER USING SAME

Abstract

Provided is an LGVF path-following controller including: an LGVF control unit that is provided with a heading angle command for a wing-fixed unmanned aerial vehicle and guidance commands from the outside, and is provided with a computed estimation disturbance speed from a nonlinear disturbance control unit; a heading angle computation control unit that computes a final heading angle of the wing-fixed unmanned aerial vehicle using a difference between the heading angle of the wing-fixed unmanned aerial vehicle, which is computed by the LGVF control unit, and a heading angle of the wing-fixed unmanned aerial vehicle in an ideal environment where a disturbance is not present; and a nonlinear disturbance control unit that computes the estimation disturbance speed using the final heading angle provided from the heading angle computation control unit and pieces of sensor data on the wing-fixed unmanned aerial vehicle, which are provided from a sensor.

Claims

1. An LGVF path-following controller comprising: an LGVF control unit that is provided with a heading angle command for a wing-fixed unmanned aerial vehicle and guidance commands, such as an airspeed and an altitude, from the outside, and is provided with a computed estimation disturbance speed from a nonlinear disturbance control unit; a heading angle computation control unit that computes a final heading angle of the wing-fixed unmanned aerial vehicle using a difference between the heading angle of the wing-fixed unmanned aerial vehicle, which is computed by the LGVF control unit, and a heading angle of the wing-fixed unmanned aerial vehicle in an ideal environment where a disturbance is not present; and a nonlinear disturbance control unit that computes the estimation disturbance speed using the final heading angle provided from the heading angle computation control unit and pieces of sensor data including a position, posture, and speed of the wing-fixed unmanned aerial vehicle, which are provided from a sensor.

2. The LGVF path-following controller according to claim 1, wherein the heading angle of the wing-fixed unmanned aerial vehicle in the ideal environment where the disturbance is not present is computed using the following equation:
{dot over (x)}=V.sub.α cos ψ+W.sub.x
{dot over (y)}=V.sub.α sin ψ+W.sub.y
ψ=u  Equation where V.sub.a denotes a flight speed of an unmanned aerial vehicle, ψ denotes a heading angle of the unmanned aerial vehicle, u denotes an input command that is a turn rate of the unmanned aerial vehicle, W denotes wind speed, W.sub.x denotes wind speed in the x-axis direction, W.sub.y denotes wind speed in the y-axis direction, {dot over (x)} denotes a speed in the x-axis direction of the unmanned aerial vehicle, and {dot over (y)} denotes a speed in the y-axis direction of the unmanned aerial vehicle.

3. The LGVF path-following controller according to claim 1, wherein The LGVF control unit is provided with the pieces of sensor data, including the position, posture, and speed of the wing-fixed unmanned aerial vehicle, from the sensor.

4. The LGVF path-following controller according to claim 3, wherein the heading angle computation control unit computes the final heading angle using the following equation: $\begin{array}{cc}{\psi }_d={\tan }^{-1}\left(\frac{{\stackrel{.}{y}}_{\mathrm{dn}}}{{\stackrel{.}{x}}_{\mathrm{dn}}}\right)& \mathrm{Equation}\end{array}$ where ψ.sub.d denotes the final heading angle, {dot over (x)}.sub.dn denotes a new input speed in the x-axis direction, which results from a disturbance computed from a disturbance observer being considered for an LGVF, and {dot over (y)}.sub.dn denotes a new input speed in the y-axis direction, which results from the disturbance computed from the disturbance observer being considered for the LGVF.

5. The LGVF path-following controller according to claim 1, wherein the heading angle computation control unit is provided with a disturbance speed reflecting wind speed.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] FIG. 1 is a block diagram illustrating operation of an LGVF path-following controller based on a nonlinear disturbance observer (NDO) according to the present invention;

[0014] FIG. 2 is a diagram illustrating a geometric structure of a tangent vector field according to the present invention;

[0015] FIG. 3 is a diagram illustrating a structure of the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention; and

[0016] FIG. 4 is a diagram illustrating a flight path of a wing-fixed unmanned aerial vehicle that is equipped with the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention.

DETAILED DESCRIPTION OF THE INVENTION

[0017] The above-described aspects of the present invention and additional aspects thereof will be apparent from a preferable embodiment that will be described with reference to the accompanying drawings. Descriptions will be provided below in sufficient detail so that a person of ordinary skill in the art clearly can understand and implement the embodiment of the present invention.

[0018] According to the present invention, there is provided a path-following guidance technique based on a nonlinear disturbance observer (NDO) for a small fixed-wing unmanned aerial vehicle that moves under the influence of a disturbance, such as wind. There is provided a control method based on a nonlinear disturbance observer that compensates for an influence of a disturbance in order that a small fixed-wing unmanned aerial to vehicle employing a Lyapunov Guidance Vector Field (LGVF) guidance technique follows a path more precisely in a situation where a disturbance, such as wind, occurs. The LGVF guidance technique is more robust against a disturbance than many other guidance techniques, and is advantageously capable of tracking a target object moving on the ground using single or multiple unmanned aerial vehicles. The DOBC control technique is a general path-following guidance technique. Unlike an existing technique that is applied to an inner loop, the DOBC control technique is applied to an outer loop. The nonlinear disturbance observer applied to the outer loop computes a disturbance to a path. The computed disturbance is input into an LGVF path-following controller to compensate for the disturbance.

[0019] FIG. 1 illustrates operation of the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention. In FIG. 1, the LGVF path-following controller based on the nonlinear disturbance observer (NDO) generates a heading angle command and guidance commands, such as an airspeed and an altitude, and sends the generated commands to the outer loop. The outer loop that is a type of proportional integral controller generates posture commands, such as a roll and a pitch, using PIXHAWK that is an automatic control device, and sends the generated posture commands. PIXHAWK receives the posture commands and stabilizes the inner loop.

[0020] With reference to FIG. 1, the LGVF path-following controller based on the nonlinear disturbance observer (NDO) generates the guidance commands, such as the airspeed and the altitude, and the heading angle command and provides the generated guidance commands, such as the airspeed and the altitude, and the generated heading angle command to an outer loop controller.

[0021] The outer loop controller generates the posture commands, such as the roll and the pitch, and provides the generated gesture commands to PIXHAWK (a flight controller). PIXHAWK generates a servo command and controls an unmanned aerial vehicle using the generated servo command. A servo compares a state of a certain device with a reference and provides feedback in the direction of stabilizing the device. Thus, the device is automatically controlled with the most suitable value or in a manner that satisfies an arbitrary target value. For this reason, the servo finds application in increasing the flight stability of the unmanned aerial vehicle.

[0022] In a case where a fixed-wing unmanned aerial vehicle employs a low-level automatic flight control system for functions of maintaining a direction, a speed, and an altitude, according to the present invention, a guidance command is input into the low-level automatic flight control system in order that the fixed-wing unmanned aerial vehicle follows a path. According to the separation principle, when it is assumed that a bandwidth of the inner loop is 5 to 10 times broader than a bandwidth of the outer loop, the inner loop and the outer loop may be individually designed into the low-level flight automatic control system. According to the present invention, the following simple two-dimensional motion equation for an unmanned aerial vehicle is applied.

{dot over (x)}=V.sub.α cos ψ+W.sub.x

{dot over (y)}=V.sub.α sin ψ+W.sub.y

ψ=u  Equation 1

[0023] where V.sub.a, ψ, and μ denotes input commands, such as a flight speed, heading angle, and turn rate, respectively, of the unmanned aerial vehicle, W denotes wind speed, W.sub.x denotes wind speed in the x-axis direction, W.sub.y denotes wind speed in the y-axis direction, {dot over (x)} denotes a speed in the x-axis direction of the unmanned aerial vehicle, and {dot over (y)} denotes a speed in the Y-direction of the unmanned aerial vehicle.

[0024] In order to facilitate application of the nonlinear disturbance observer, Equation 1 is rewritten as in the form of the following Equation 2. That is, Equation 1 is rewritten using functions f(x), g.sub.1(x), and g.sub.2(x) as in Equation 2.

{dot over (x)}=f(x)+q.sub.1(x)u+g.sub.2(x)d  Equation 2

[0025] The functions f(x), g.sub.1(x), and g.sub.2(x) are computed from Equation 1. When it is assumed that a disturbance changes over time ({dot over (d)}≈0), the nonlinear disturbance observer (NDO) is derived as follows.

ż=−l(x)g.sub.2(x)z−l(x)[g.sub.2(x)p(x)+f(x)+g.sub.1(x)u]

{circumflex over (d)}=z+p(x)  Equation 3

[0026] where {circumflex over (d)}=[Ŵ.sub.x Ŵ.sub.y].sup.T denotes an estimated speed of the disturbance and includes a modeling error, uncertainty, sensor noise, and the like, ż denotes an amount of change in an inner state of an observer, and Ŵ.sub.x is a disturbance in the x-axis direction, which is estimated by the nonlinear disturbance observer. At this point, disturbances that are estimated by the nonlinear disturbance observer include wind in the x-axis direction, systematic uncertainty, sensor noise, and the like.

[0027] z denotes an inner state of a nonlinear observer, and p(x) denotes a designed nonlinear function. l(x) denotes a gain value of the nonlinear disturbance observer, and is expressed as follows.

[00001]$\begin{array}{cc}l\left(x\right)=\frac{\partial p\left(x\right)}{\partial x}e=d-\hat{d}={\left[e_x{e}_y\right]}^T& \mathrm{Equation}4\end{array}$

expresses an estimation error of the NOD described above. When it is assumed that the disturbance has a fixed trend by comparison with an observer dynamic and changes slowly, Equation 2, Equation 3, and Equation 4 are combined, and thus the following estimation error dynamics can be derived. d denotes a disturbance speed reflecting wind speed.

[00002]$\begin{array}{cc}\stackrel{.}{e}=\stackrel{.}{d}=\hat{d}=-\stackrel{.}{z}-\frac{\partial p\left(x\right)}{\partial x}\stackrel{.}{x}=-l\left(x\right)g_2\left(x\right)e& \mathrm{Equation}5\end{array}$

[0028] Therefore, a problem of designing the disturbance observer leads to a problem of selecting a suitable gain value for achieving exponential stabilization regardless of a state x. According to the present invention, a g.sub.2(x) function is a constant matrix, and thus an observer gain may be set as follows.

[00003]$\begin{array}{cc}l\left(x\right)=L=\left[\begin{array}{cc}{l}_x& 0\\ 0& l_y\end{array}\right]& \mathrm{Equation}6\end{array}$

[0029] where l.sub.x, l.sub.y denotes a positive gain value that is adjustable and determines a convergence ratio for an estimation error. Therefore, a nonlinear function p(x) can be obtained by integrating l(x) with respect to a state x using Equation 4.

[0030] An LGVF uses an input speed that appears in the following Equation 7.

[00004]$\begin{array}{cc}\left[\begin{array}{c}{\stackrel{.}{x}}_d\\ {\stackrel{.}{y}}_d\end{array}\right]=\frac{-v_d}{k_{l}r\left(r^2+r_d^2\right)}\left[\frac{\delta x\left(r^2-r_d^2\right)+\delta y\left(2{\mathrm{rr}}_d\right)}{\delta y\left(r^2-r_d^2\right)+\delta x\left(2{\mathrm{rr}}_d\right)}\right]& \mathrm{Equation}7\end{array}$

[0031] where v.sub.d and r.sub.d denote an input speed and input radius of the unmanned aerial vehicle. r=√{square root over (δx.sup.2+δy.sup.2)} expresses a distance between the unmanned aerial vehicle and the origin, as illustrated in FIG. 2. k.sub.l is a gain value that determines a speed at which the unmanned aerial vehicle converges on a circular path. A vehicle angle command to be input for the unmanned aerial vehicle is determined as follows. δ denotes a displacement between the origin and a position of the unmanned aerial vehicle, and {dot over (x)}.sub.d denotes an input speed in the x-direction.

[00005]$\begin{array}{cc}{\psi }_d={\tan }^{-1}\left(\frac{{\stackrel{.}{y}}_d}{{\stackrel{.}{x}}_d}\right)& \mathrm{Equation}8\end{array}$

[0032] The heading angle command is obtained from a two-dimensional speed that is given by Equation 7. A guidance command (u.sub.w) for the turn rate of the unmanned aerial vehicle is expressed, as the sum of proportional feedback and feedforward terms, as follows.

u.sub.w=−k.sub.w(ψ−ψ.sub.d)+ψ.sub.d  Equation 9

[0033] where k.sub.w denotes a gain value for the turn rate and is generally set by tuning.

[00006]$\begin{array}{cc}{\stackrel{.}{\psi }}_d=4v_d\frac{r_d{r}^2}{{\left(r^2+r_d^2\right)}^2}& \mathrm{Equation}10\end{array}$

[0034] where Ψ.sub.d denotes an input turn-rate command obtainable by differentiating Equation 8.

[0035] A disturbance, such as wind, is estimated by the nonlinear disturbance observer (NDO) as in Equation 3. To compensate for this, a new input speed for the LGVF in Equation 7 can be computed as follows.

[00007]$\begin{array}{cc}\left[\begin{array}{c}{\stackrel{.}{x}}_{\mathrm{dn}}\\ {\stackrel{.}{y}}_{\mathrm{dn}}\end{array}\right]=\left[\begin{array}{c}{\hat{W}}_x+{\alpha }_s{\stackrel{.}{x}}_d\\ {\hat{W}}_y+{\alpha }_s{\stackrel{.}{y}}_d\end{array}\right]& \mathrm{Equation}11\end{array}$

[0036] where , denotes a disturbance estimated using Equation 3, and α.sub.s denotes a scale factor. A final input heading angle command for compensating for the disturbance is as follows.

[0037] {dot over (x)}.sub.dn denotes a new input speed in the x-direction, which results from the disturbance computed from the disturbance observer being configured for the LGVF. {dot over (y)}.sub.dn denotes a new input speed in the y-axis direction, which results from the disturbance computed from the disturbance observer being considered for the LGVF.

[00008]$\begin{array}{cc}{\psi }_d={\tan }^{-1}\left(\frac{{\stackrel{.}{y}}_{\mathrm{dn}}}{{\stackrel{.}{x}}_{\mathrm{dn}}}\right)& \mathrm{Equation}12\end{array}$

[0038] FIG. 3 illustrates a structure of the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention. The structure of the LGVF path-following controller based on the nonlinear disturbance observer according to an embodiment of the present invention will be described in detail below with reference to FIG. 3.

[0039] With reference to FIG. 3, an LGVF path-following controller 300 is configured with an inner loop and an outer loop. The outer loop includes an LGVF control unit 310 and a nonlinear disturbance control unit 330. The loop includes a heading angle computation control unit 320.

[0040] Information on a disturbance, such as wind or systematic uncertainty, which is estimated by the nonlinear disturbance observer (NDO), is input into the LGVF path-following controller to compensate for an influence of the disturbance. When a heading angle input command and guidance commands, such as a speed and an altitude, are determined by the LGVF control unit 310, the posture command is generated in the heading angle computation control unit 320 that is the outer loop which includes an anti-windup augmented system and a proportional feedback controller.

[0041] A configuration of the LGVF path-following controller according to the present invention will be described in detail below with reference to FIG. 3.

[0042] The LGVF control unit 310 receives the heading angle command and the guidance commands, such as the airspeed and the altitude, from the outside. In addition, the LGVF control unit 310 receives pieces of sensor data, such as a position, posture, and speed of a wing-fixed unmanned aerial vehicle. In addition, the LGVF control unit 310 is provided with an estimation disturbance speed computed by the nonlinear disturbance control unit 330.

[0043] The LGVF control unit 310 computes the heading angle of the wing-fixed unmanned aerial vehicle using the provided pieces of information. The computed heading angle of the unmanned aerial vehicle is transferred to the heading angle computation control unit 320 that is an inner loop and the nonlinear disturbance control unit 330 that is an outer loop.

[0044] The heading angle computation control unit 320 computes the heading angle of the wing-fixed unmanned aerial vehicle that results from considering a disturbance using the heading angle of the wing-fixed unmanned aerial vehicle provided from the LGVF control unit 310 and the motion equation (Equation 1) for the wing-fixed unmanned aerial vehicle in an ideal environment where the disturbance is not present. The heading angle of the wing-fixed unmanned aerial vehicle, which results from considering the disturbance computed by the heading angle computation control unit 320 and the pieces of sensor data, such as the position, posture, and speed of the wing-fixed unmanned aerial vehicle, which are measured by the sensor, are provided to the nonlinear disturbance control unit 330.

[0045] The nonlinear disturbance control unit 330 computes a disturbance that is estimated using the heading angle of the wing-fixed unmanned aerial vehicle that results from considering the disturbance, which is provided from the heading angle computation control unit 320, and the pieces of sensor data, such as the position, posture, and speed of the wing-fixed unmanned aerial vehicle, which are measured by the sensor. The estimated disturbance is computed using Equation 12.

[0046] As described above, the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention includes the outer loop that includes the nonlinear disturbance control unit that estimates a disturbance, such as wind, and the LGVF control unit for following a path.

[0047] FIG. 4 illustrates a flight path of the wing-fixed unmanned aerial vehicle that is equipped with the LGVF path-following controller based on the nonlinear disturbance observer according to the present invention. From FIG. 4, it can be understood that whereas the wing-fixed unmanned aerial vehicle that was equipped with the LGVF path-following controller based on the nonlinear disturbance observer flied along a path that was set, the wing-fixed unmanned aerial vehicle that was not equipped with the LGVF path-following controller based on the nonlinear disturbance observer did not fly along the path that was set.

[0048] The embodiment of the present invention is described only in an exemplary manner referring to the drawings. It will be apparent to a person of ordinary skill in the art to which the present invention pertains that various other modifications and equivalents are possible from this description.