CN104483977A - No-speed feedback dynamic surface control method for steering engine pan-tilt - Google Patents

Abstract

The invention discloses a no-speed feedback dynamic surface control method for a steering engine pan-tilt. The method comprises camera azimuth angle and pitch angle calculation, steering engine model construction and analysis, high-gain observer design, dynamic surface control input voltage design, control parameter design and adjustment, and tracking performance examination according to a simulation experiment. The control method does not need to add an extra angular speed hardware measurement circuit, control input design is simple, control parameters are easy to adjust, and steering engine control precision is high.

Description

A kind of steering wheel The Cloud Terrace without velocity feedback dynamic surface control method

Technical field

The present invention relates to automatic control technology field, be specifically related to a kind of steering wheel The Cloud Terrace without velocity feedback dynamic surface control method.

Background technology

By by camera carrying on unmanned plane or ground robot, the supervision to specific objective can be realized, the danger signal in testing environment.Moving and rocking due to unmanned plane during flying or ground robot, must be arranged on video camera on The Cloud Terrace.Cradle head controllor can adjust the position of camera, makes it aim at monitoring objective all the time.Common two-freedom all-directional tripod head, it can either left rotation and right rotation can deflect again up and down in pitching.According to the difference of driving mechanism, The Cloud Terrace can be divided into AC type and once-through type.Steering wheel The Cloud Terrace is a kind of once-through type The Cloud Terrace in essence, and its adopts steering wheel as driving mechanism, has that structure is simple, quality is light and handy, control the advantages such as flexible.As long as know the positional information of monitoring objective point, in conjunction with the positional information of mobile apparatus human body, expectation position angle and angle of pitch instruction that camera light axial vector points to impact point all the time just can be calculated.By servo driving, change the direction of camera, follow the tracks of expected angle fast, make monitoring objective point be positioned at the central area of the plane of delineation all the time.

In recent years, many advanced control methods are used in above the control of steering wheel The Cloud Terrace, comprising PD control, feedforward compensation, method for optimally controlling etc.Dynamic surface control is a kind of control technology of novelty, and its design process is simple, is easy to physics realization.In design procedure, each step introducing low-pass filter, overcomes counter " differential blast " phenomenon pushing away control method and exist.Further, it does not have " shivering " phenomenon of sliding-mode control.

The control method of existing steering wheel The Cloud Terrace, needs the rotational angular velocity using steering wheel The Cloud Terrace mostly.The method obtaining rotational angular velocity has two kinds, and a kind of is by directly carrying out difference to the angle of rotating, there is larger error, causes following the tracks of regulating time and becomes large; Another is by turn meter, and this can introduce peripheral circuit, increases the volume and weight of The Cloud Terrace, reduces the reliability of system.

Summary of the invention

For solving the deficiency that prior art exists, the invention discloses a kind of steering wheel The Cloud Terrace without velocity feedback dynamic surface control method, for the steering wheel The Cloud Terrace in video monitoring system, the dynamic surface control method not needing rotational angular velocity to measure proposed, expected angle is followed the tracks of quickly and accurately for control flaps machine head, not needing in rotational angular velocity measurement situation, in conjunction with dynamic surface control technology and High-gain observer, realize the fast and accurately tracking of steering wheel to expected angle.High-gain observer can by the angle information rotated, and Accurate Reconstruction angular velocity information, realizes the software measurement to rotational angular velocity.

For achieving the above object, concrete scheme of the present invention is as follows:

Steering wheel The Cloud Terrace without a velocity feedback dynamic surface control method, comprising:

Obtain the monitoring objective point of camera head and carry unmanned plane or the coordinate information of ground robot in inertial system of steering wheel The Cloud Terrace, according to the coordinate information in inertial system, obtaining camera commands value;

Build actuator model, choose the relation that order transfer function represents steering wheel input voltage and output corner;

The rotational angle information exported by steering wheel, is utilized High-gain observer to reconstruct angular velocity, realizes the software measurement to steering wheel rotational angular velocity;

According to camera commands value and the angular velocity signal that utilizes High-gain observer to measure, utilize dynamic surface control method to calculate input voltage that controller controls steering wheel, then controls the velocity of rotation of steering wheel, makes Camera location command value.

Described coordinate information in inertial system is obtained by the infrared sensor of indoor or outdoor GPS.

The described process obtaining camera commands value is:

According to coordinate information, calculate the optical axis phasor coordinate P of camera points impact point _e; Then optical axis vector P is calculated by coordinate transform _ebe P at the coordinate of mobile robot's body coordinate system _b; Last according to P _bcoordinate, calculate camera orientation angle command value ψ _cwith angle of pitch command value θ _c.

Angle x is followed the tracks of in the expectation of steering wheel _1d; For the steering wheel x of horizontal direction deflection _1d=ψ _cif, pitch orientation then have x _1d=θ _c.

Described structure actuator model is the topworks's steering wheel for the steering wheel The Cloud Terrace being operated in linear zone, chooses the relation that order transfer function represents its input voltage and output corner, order transfer function is converted into state-space expression.

The described dynamic surface control method that utilizes calculates the input voltage controlling steering wheel, and the step comprised is:

Setting camera orientation angle command value and the camera orientation angle of measurement are subtracted each other and are obtained error S ₁, virtual controlling is inputted low-pass first order filter and is exported;

The angular velocity signal that High-gain observer is measured and the output of low-pass first order filter are subtracted each other and are obtained error S ₂, obtain servos control input voltage.

Described utilize dynamic surface control method calculate the input voltage controlling steering wheel after also comprise the adjustment of controling parameters, comprising:

The adjustment of High-gain observer controling parameters, coefficient matrices A in High-gain observer process of establishing is made to be Hurwitz matrix, ε characterizes the precision of observational error, the observational error of High-gain observer to angular velocity signal is reduced by reducing ε, the adjustment of controller parameter, if tracking error is excessive or regulation time is long, discontented sufficient requirement of real-time control, then increase c ₁, c ₂or reduce τ ₁, on the one hand, increase c ₁, c ₂be equivalent to the amplitude and the bandwidth that increase control inputs; On the other hand, τ is reduced ₁improve the response speed of system.

Emulation experiment is carried out after the adjustment of High-gain observer controling parameters and the adjustment of controller parameter, steering wheel is made to be stabilized to expectation value in setting-up time, and the observational error of High-gain observer is less than setting value, if control effects can not meet the demands, continue regulable control parameter, until control effects reaches requirement.

A kind of steering wheel The Cloud Terrace without velocity feedback dynamic surface control device, described without velocity feedback dynamic surface control device for obtaining the monitoring objective point of camera head and carrying unmanned plane or the coordinate information of ground robot in inertial system of steering wheel The Cloud Terrace, according to the coordinate information in inertial system, obtain camera commands value;

The rotational angle information exported by steering wheel, is utilized High-gain observer to reconstruct angular velocity, realizes the software measurement to steering wheel rotational angular velocity;

According to the actuator model of camera commands value, the angular velocity signal utilizing High-gain observer to measure and structure, utilize dynamic surface control method to calculate the input voltage controlling steering wheel, then control the velocity of rotation of steering wheel, make Camera location command value.

Beneficial effect of the present invention:

The present invention is directed to steering wheel The Cloud Terrace, devise a kind of without velocity feedback dynamic surface control method.Do not need rotational angular velocity to measure, achieve its high-precision Angle ambiguity.The control method of design, do not need to increase extra hardware circuit, make system bulk little, quality is light.Adopt dynamic surface control method, avoid derivative operation repeatedly, parameter regulates simple.After calculating camera orientation angle command value and angle of pitch command value, can quickly and smoothly trace command signal, make monitoring objective be positioned at picture centre region.

Accompanying drawing explanation

Fig. 1 the invention process steps flow chart block diagram;

Fig. 2 Control system architecture schematic diagram of the present invention;

Fig. 3 (a) angleonly tracking emulation of the present invention schematic diagram;

Fig. 3 (b) angular velocity of the present invention follows the tracks of emulation schematic diagram;

Fig. 4 control inputs voltage emulation of the present invention schematic diagram;

Fig. 5 (a) angular observation error emulation of the present invention schematic diagram;

Fig. 5 (b) angular velocity observational error emulation of the present invention schematic diagram.

Embodiment:

Below in conjunction with accompanying drawing, the present invention is described in detail:

As shown in Figure 1, a kind of steering wheel The Cloud Terrace without velocity feedback dynamic surface control method, comprise the following steps:

Step 1: resolving of camera orientation angle and the angle of pitch

Obtain monitoring objective point by infrared sensor (indoor) or GPS (outdoor) and be respectively P with the unmanned plane or the coordinate of ground robot in inertial system carrying steering wheel The Cloud Terrace ₁=[x ₁, y ₁, z ₁] ^tand P ₂=[x ₂, y ₂, z ₂] ^t, then the optical axis phasor coordinate of camera points impact point is P _efor:

P _e＝P ₁-P ₂(1)

Then the coordinate of optical axis vector in mobile robot's body coordinate system is

P _b＝RP _e＝[x _c,y _c,z _c] ^T(2)

Here R represents Douglas Rodríguez rotation matrix, and it is defined as follows:

$R=\left[\begin{array}{ccc}{C}_p{C}_r& S_p{S}_q{C}_r-C_q{S}_r& C_q{S}_p{C}_r+S_q{S}_r\\ C_p{S}_r& S_q{S}_p{S}_r+C_q{C}_r& C_q{S}_p{S}_r-S_q{C}_r\\ -S_p& S_q{C}_p& C_q{C}_p\end{array}\right]---\left(3\right)$

Vector [p, q, r] ^trepresent carry steering wheel The Cloud Terrace unmanned plane or ground robot relative to the attitude angle of inertial coordinates system, can be obtained by sensor measurement.

C ₍₎and S ₍₎represent cosine and sine function respectively, C _pc _rs _ps _qc _qs _rrepresent corresponding lower target sine value or cosine value.

The luffing of video camera is completed by the steering wheel that execution scope is 0 degree to 90 degree, calculates angle of pitch command value θ according to optical axis phasor coordinate _c:

${\mathrm{\θ}}_c=\left\{\begin{array}{cc}\arctan \left(\frac{z_c}{\sqrt{x_c+y_c}}\right)& z_c>0\\ 0& z_c\≤0\end{array}\right\}---\left(4\right)$

Z _cx _cy _coptical axis vector P _bthree components, as shown in formula (2).

It is that the steering wheel of 0 degree to 360 degree completes that the horizontal deflection of video camera moves through execution scope, calculates position angle command value ψ according to optical axis phasor coordinate _c:

${\mathrm{\&psi;}}_c=\left\{\begin{array}{cc}\arctan (y_c/x_c)& x_c>0\\ \arctan (y_c/x_c)-\mathrm{\&pi;}& x_c<0,y_c\&le;0\\ \arctan (y_c/x_c)+\mathrm{\&pi;}& x_c<0,y_c>0\\ -\frac{\mathrm{\&pi;}}{2}& x_c=0,y_c\&le;0\\ \frac{\mathrm{\&pi;}}{2}& x_c=0,y_c>0\end{array}\right.---\left(5\right)$

Calculate the angle of pitch command value θ of video camera _cwith position angle instruction angle command value ψ _c, angle x is followed the tracks of in the expectation obtaining topworks's steering wheel _1d.For the steering wheel x of horizontal direction deflection _1d=ψ _cif, pitch orientation then have x _1d=θ _c, just can by controlling steering wheel rotational angle, actuated camera trace command value.Two steering wheels of steering wheel The Cloud Terrace are except execution scope has difference, and control program is completely the same.Below steering wheel cradle head control rule design part, only need a design steering wheel without speed feedback control scheme.

Step 2: the structure of actuator model and analysis

The steering wheel The Cloud Terrace being normally operated in linear zone can with order transfer function approximate description below:

$G\left(s\right)=\frac{{\mathrm{\&omega;}}_n^2}{s^2+2\mathrm{\&xi;}{\mathrm{\&omega;}}_{n}s+{\mathrm{\&omega;}}_n^2}---\left(6\right)$

Wherein: ω _nrepresent the undamped natural frequency of steering wheel; ξ represents ratio of damping; Parameter in model can according to the free-running frequency ω that instructions provides _ncalculate with ratio of damping ξ.Parameter 2 ξ ω _ncan according to the free-running frequency ω that instructions provides _ncalculate with ratio of damping ξ.

For the ease of design, write transport function above as state space form, defined two state variable x ₁, x ₂, formula (6) just can be write as:

$\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_1=x_2\\ {\stackrel{\&CenterDot;}{x}}_2={-\mathrm{ax}}_2-{\mathrm{bx}}_1+\mathrm{bu}\left(t\right)\\ y=x_1\end{array}---\left(7\right)$

Here a=2 ξ ω is had _n, x ₁and x ₂represent rotational angle and the rotational angular velocity of steering wheel respectively.The rotational angle that y represents steering wheel exports, and u (t) represents control inputs voltage.Therefore steering wheel is converted into system (7) to instruction angle without velocity feedback tracking and is only x in output feedack amount ₁condition under, to command signal x _1dcarry out the control problem of following the tracks of.

Step 3: the design of High-gain observer

Design High-gain observer, the rotational angle information exported by steering wheel, reconstruct angular velocity.The effect of this observer under the condition not increasing hardware circuit, can realize the software measurement to rotational angular velocity, overcomes the error of calculation that direct angular difference brings.For dynamic system (7), design High-gain observer (8), reconstruct angular velocity signal.

$\begin{array}{c}{\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_1={\hat{x}}_2+\frac{{\mathrm{\&mu;}}_1}{\mathrm{\&epsiv;}}(y-{\hat{x}}_1)\\ {\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_2=-{a\hat{x}}_2-{\mathrm{bx}}_1+\mathrm{bu}+\frac{{\mathrm{\&mu;}}_2}{{\mathrm{\&epsiv;}}^2}(y-{\hat{x}}_1)\end{array}---\left(8\right)$

In formula (8), get μ ₁and μ ₂for arithmetic number, ε < < 1.

Wherein, state variable as High-gain observer (8) represents the estimated value of steering wheel rotational angle and rotational angular velocity; The rotational angle that y represents steering wheel exports; represent High-gain observer variable along with the rule of conversion of time; U is writing a Chinese character in simplified form of u (t), represents control inputs voltage; A=2 ξ ω _n,

Order then High-gain observer can be:

$\begin{array}{c}{\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_1={\hat{x}}_2+l_1(y-{\hat{x}}_1)\\ {\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_2=-{a\hat{x}}_2-{\mathrm{bx}}_1+\mathrm{bu}+l_2(y-{\hat{x}}_1)\end{array}---\left(9\right)$

High-gain observer observational error is can obtain according to (7) and (9)

$\begin{array}{c}{\stackrel{\stackrel{\&CenterDot;}{~}}{x}}_1=-l_1{\tilde{x}}_1+{\tilde{x}}_2\\ {\stackrel{\stackrel{\&CenterDot;}{~}}{x}}_2=-l_2{\tilde{x}}_1-a{\tilde{x}}_2\\ y=x_1\end{array}---\left(10\right)$

Namely wherein matrix of coefficients $A=\left[\begin{array}{cc}{-l}_1& 1\\ {-l}_2& -a\end{array}\right],$ Error vector if choosing coefficient A is Hurwitz, namely the eigenwert of A is negative, then can obtain observational error exponential convergence is to zero.Secular equation is $|\mathrm{sI}-A|=\left|\begin{array}{cc}s+l_1& -1\\ l_2& s+a\end{array}\right|=0,$ Proper polynomial is s ²+ (l ₁+ a) s+l ₂=0, according to Routh Criterion, proper polynomial above be made to meet Hurwitz condition, then requirement

$\left\{\begin{array}{c}{l}_1+a>0\\ l_2>0\end{array}\right.---\left(11\right)$

According to observer dynamic system formula (4), definition then

$\begin{array}{c}{\stackrel{\stackrel{\&CenterDot;}{~}}{x}}_2={\stackrel{\&CenterDot;}{x}}_2-{\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_2\\ =-a{\tilde{x}}_2-l_2{\tilde{x}}_1\end{array}---\left(12\right)$

Order the dynamic system (7) then comprising evaluated error can be write as following form:

$\begin{array}{c}{\stackrel{\&CenterDot;}{x}}_1=x_2={\hat{x}}_2+{\tilde{x}}_2={\hat{x}}_2+o_1\\ {\stackrel{\stackrel{\&CenterDot;}{^}}{x}}_2=-a{\hat{x}}_2-{\mathrm{bx}}_1+\mathrm{bu}\left(t\right)+o_2\end{array}---\left(13\right)$

O _i(i=1,2) represent High-gain observer observational error, and exponential convergence is to zero in time, are the dimensionlesss about the time.

Step 4: dynamic surface control In-put design

For the steering wheel of horizontal deflection, design con-trol input voltage.The design of this control inputs is the process of progressively going forward one by one, and one is divided into two small steps.First small step: setting camera orientation angle command value x _1d=π/3, with the angle x measured ₁subtract each other and obtain error S ₁=x ₁-x _1d, controller parameter c ₁value is 5, the input of design virtual controlling it is input to timeconstantτ ₁value is the low-pass first order filter of 0.1 in obtain export x _2d.Second small step: low-pass first order filter exports with the angular velocity signal that High-gain observer is measured define second error signal choose controller parameter c ₂value is 10, calculates servos control input voltage can find out that control inputs voltage u employs the Attitude rate estimator value of High-gain observer output replace the actual value x of angular velocity ₂.

Concrete design of control law step is as follows:

Step 1: tracking error is S ₁=x ₁-x _1d, then tracking error to the derivative of time is

${\stackrel{\&CenterDot;}{S}}_1=x_2-{\stackrel{\&CenterDot;}{x}}_{1d}={\hat{x}}_2-{\stackrel{\&CenterDot;}{x}}_{1d}+o_1---\left(14\right)$

Note, in error derivative above, not comprising angular velocity component, but the estimator of angular velocity choose Lyapunov function $V_1=\frac{1}{2}{S}_1^2,$ In order to make ${\stackrel{\&CenterDot;}{V}}_1=\frac{1}{2}{S}_1{\stackrel{\&CenterDot;}{S}}_1=\frac{1}{2}{S}_1({\hat{x}}_2-{\hat{x}}_{1d}+o_1)\&le;0,$ Get virtual controlling be

${\stackrel{\&OverBar;}{x}}_2=-c_1{S}_1+{\stackrel{\&CenterDot;}{x}}_{1d}---\left(15\right)$

Wherein c ₁for arithmetic number.In order to avoid right differentiate, adopts low-pass filter to export x _2d, approach respectively

$\begin{array}{c}{\mathrm{\&tau;}}_1{\stackrel{\&CenterDot;}{x}}_{2d}+x_{2d}={\stackrel{\&OverBar;}{x}}_2\\ x_{2d}\left(0\right)={\stackrel{\&OverBar;}{x}}_2\left(0\right)\end{array}---\left(16\right)$

Step 2: definition then

${\stackrel{\&CenterDot;}{S}}_2=-a{\hat{x}}_2-b{x}_1+\mathrm{bu}\left(t\right)+o_2-{\stackrel{\&CenterDot;}{x}}_{2d}---\left(17\right)$

Get Lyapunov function actual control inputs voltage design is as follows

$u=\frac{1}{b}(a{\hat{x}}_2+{\mathrm{bx}}_1+{\stackrel{\&CenterDot;}{x}}_{2d}-c_2{S}_2)---\left(18\right)$

Wherein c ₂for arithmetic number.Then ${\stackrel{\&CenterDot;}{S}}_2=-c_2{S}_2+o_2,$ Therefore ${\stackrel{\&CenterDot;}{V}}_2=S_2{\stackrel{\&CenterDot;}{S}}_2=-c_2{S}_2^2+S_2{o}_2\&le;0,$ Whole system is stablized.

Step 5: the design of controling parameters and adjustment

In High-gain observer, l _i(i=1,2) affect the speed of convergence measured angle and the angular velocity true value of steering wheel The Cloud Terrace.High-gain observer parameter l _idesign make A be Hurwitz.Being Hurwitz to meet A, needing the eigenwert real part ensureing A to be negative, namely | sI-A|=s ²+ (l ₁+ a) s+l ₂the characteristic root real part of=0 is negative.Choose ε=0.1, μ ₁=1, μ ₂=3.Being respectively-6.5+16.96i and-6.5-16.96i through calculating two characteristic roots that can obtain matrix A, satisfying condition.

Controller parameter c ₁, c ₂, τ ₁for regulating parameter.If tracking error is excessive or regulation time is long, discontented sufficient requirement of real-time control, then can increase c ₁, c ₂or reduce τ ₁.On the one hand, c is increased ₁, c ₂be equivalent to the amplitude and the bandwidth that increase control inputs; On the other hand, τ is reduced ₁the response speed of system can be improved.Therefore these two kinds of ways all contribute to the tracking performance of improvement system.

Step 6: according to emulation experiment inspection tracking performance

Utilize matlab7.0 simulation software, carry out emulation experiment.Control system requires that steering wheel is stabilized to expectation value in 1s, and the observational error of High-gain observer is less than 1 × 10 ^-3.If control effects can not meet the demands, return step 5 and continue adjustment control parameter, until control effects reaches requirement.If control effects meets the demands, then design end.

During emulation, according to parameter ξ=0.72 on the steering wheel The Cloud Terrace instructions of reality, ω _n=25, calculate model parameter a=36, b=625.Choose horizontal azimuth as an example, if clearing position angle instruction angle command value ψ out _c=π/3 (60 degree), namely expectation value x _1d=π/3.Control system requires that steering wheel is stabilized to expectation value in 1s, and the observational error of High-gain observer is less than 1 × 10 ^-3.If control effects can not meet the demands, return step 5 and continue adjustment control parameter, until control effects reaches requirement.If control effects meets the demands, then design end.Fig. 2 is shown in by control system schematic diagram, simulated effect, angle of the present invention and angular velocity follow the tracks of emulation schematic diagram as Fig. 3 (a)-3 (b), control inputs voltage emulation schematic diagram of the present invention, as shown in Figure 4, angle of the present invention and angular velocity observational error emulation schematic diagram Fig. 5 (a)-5 (b).

Whole design process, starts with from the Angle ambiguity problem of the steering wheel The Cloud Terrace of reality.First according to monitoring requirement, position angle expectation value and the angle of pitch expectation value of video camera is calculated.Under the condition that angular velocity signal can not be measured by hardware circuit, devise High-gain observer, utilize the angle signal reconstruct angular velocity that can measure, realize angular velocity software measurement.Adopt dynamic surface control method afterwards, calculate steering wheel input voltage, reach the quick accurate tracking of angle expectation value.Controller parameter is simple, can be regulated, reach high-precision system responses quality by simulation result.Realize the fast and accurately tracking of steering wheel to expected angle, make monitoring objective be positioned at the central area of video camera.

Claims (9)

1. steering wheel The Cloud Terrace without a velocity feedback dynamic surface control method, it is characterized in that, comprising:

Obtain the monitoring objective point of camera head and carry unmanned plane or the coordinate information of ground robot in inertial system of steering wheel The Cloud Terrace, according to the coordinate information in inertial system, obtaining camera commands value;

Build actuator model, choose the relation that order transfer function represents steering wheel input voltage and output corner;

The rotational angle information exported by steering wheel, is utilized High-gain observer to reconstruct angular velocity, realizes the software measurement to steering wheel rotational angular velocity;

According to camera commands value and the angular velocity signal that utilizes High-gain observer to measure, utilize dynamic surface control method to calculate input voltage that controller controls steering wheel, then controls the velocity of rotation of steering wheel, makes Camera location command value.

2. a kind of steering wheel The Cloud Terrace as claimed in claim 1 without velocity feedback dynamic surface control method, it is characterized in that, described coordinate information in inertial system is obtained by the infrared sensor of indoor or outdoor GPS.

3. a kind of steering wheel The Cloud Terrace as claimed in claim 1 without velocity feedback dynamic surface control method, it is characterized in that, described in obtain camera commands value process be:

According to coordinate information, calculate the optical axis phasor coordinate P of camera points impact point _e; Then optical axis vector P is calculated by coordinate transform _ebe P at the coordinate of mobile robot's body coordinate system _b; Last according to P _bcoordinate, calculate camera orientation angle command value ψ _cwith angle of pitch command value θ _c.

4. a kind of steering wheel The Cloud Terrace as claimed in claim 3 without velocity feedback dynamic surface control method, it is characterized in that, angle x is followed the tracks of in the expectation of steering wheel _1d; For the steering wheel x of horizontal direction deflection _1d=ψ _cif, pitch orientation then have x _1d=θ _c.

5. a kind of steering wheel The Cloud Terrace as claimed in claim 1 without velocity feedback dynamic surface control method, it is characterized in that, describedly utilize dynamic surface control method to calculate to control the input voltage of steering wheel, the step comprised is:

Setting camera orientation angle command value and the camera orientation angle of measurement are subtracted each other and are obtained error S ₁, virtual controlling is inputted low-pass first order filter and is exported;

The angular velocity signal that High-gain observer is measured and the output of low-pass first order filter are subtracted each other and are obtained error S ₂, obtain servos control input voltage.

6. a kind of steering wheel The Cloud Terrace as claimed in claim 1 without velocity feedback dynamic surface control method, it is characterized in that, described structure actuator model is the topworks's steering wheel for the steering wheel The Cloud Terrace being operated in linear zone, choose the relation that order transfer function represents its input voltage and output corner, order transfer function is converted into state-space expression.

7. a kind of steering wheel The Cloud Terrace as claimed in claim 1 without velocity feedback dynamic surface control method, it is characterized in that, described utilize dynamic surface control method calculate the input voltage controlling steering wheel after also comprise the adjustment of controling parameters, comprising:

The adjustment of High-gain observer controling parameters, coefficient matrices A in High-gain observer process of establishing is made to be Hurwitz matrix, ε characterizes the precision of observational error, the observational error of High-gain observer to angular velocity signal is reduced by reducing ε, the adjustment of controller parameter, if tracking error is excessive or regulation time is long, discontented sufficient requirement of real-time control, then increase c ₁, c ₂or reduce τ ₁, on the one hand, increase c ₁, c ₂be equivalent to the amplitude and the bandwidth that increase control inputs; On the other hand, τ is reduced ₁improve the response speed of system.

8. a kind of steering wheel The Cloud Terrace as claimed in claim 7 without velocity feedback dynamic surface control method, it is characterized in that, emulation experiment is carried out after the adjustment of High-gain observer controling parameters and the adjustment of controller parameter, steering wheel is made to be stabilized to expectation value in setting-up time, and the observational error of High-gain observer is less than setting value, if control effects can not meet the demands, continue regulable control parameter, until control effects reaches requirement.

9. a steering wheel The Cloud Terrace without velocity feedback dynamic surface control device, it is characterized in that, described without velocity feedback dynamic surface control device for obtaining the monitoring objective point of camera head and carrying unmanned plane or the coordinate information of ground robot in inertial system of steering wheel The Cloud Terrace, according to the coordinate information in inertial system, obtain camera commands value;

The rotational angle information exported by steering wheel, is utilized High-gain observer to reconstruct angular velocity, realizes the software measurement to steering wheel rotational angular velocity;

According to the actuator model of camera commands value, the angular velocity signal utilizing High-gain observer to measure and structure, utilize dynamic surface control method to calculate the input voltage controlling steering wheel, then control the velocity of rotation of steering wheel, make Camera location command value.