CN103344243A - Friction parameter identification method for aerial remote-sensing inertial stabilization platform - Google Patents

Abstract

The invention relates to a friction parameter identification method for an aerial remote-sensing inertial stabilization platform. The friction parameter identification method comprises the following steps of: establishing a LuGre friction model of the inertial stabilization platform based on the analysis of sinusoidal response curves of systems; and obtaining parameters of the LuGre friction model by respectively designing parameter identification schemes to three frame systems of the inertial stabilization platform and utilizing a two-step-identified and dynamic-parameter-optimized friction parameter identification method. According to the friction parameter identification method, the parameters of the LuGre friction model can be accurately obtained, the friction moment compensation can be realized, and the control precision and the stabilization precision of the systems of the inertial stabilization platform can be improved; the friction parameter identification method can be widely applied to the systems of various remote-sensing inertial stabilization platforms.

Description

A kind of airborne remote sensing inertially stabilized platform friction parameter discrimination method

Technical field

The present invention relates to a kind of airborne remote sensing inertially stabilized platform friction parameter discrimination method.

Background technology

The airborne remote sensing system maneuverability, real-time repeated measures, remote sensing equipment convertibility, obtain high-definition remote sensing data capability, financial cost and stereopsis etc. a lot of aspect, all have a special advantages.Therefore, at western developed country, be used for city planning and base map and survey and draw about high resolution space data dependence airborne remote sensing system 65% or more and guarantee.The high resolving power earth observation is the important directions of airborne remote sensing system development, but because the influence of air turbulence and carrier aircraft oneself factor, and the carrier aircraft body can't held stationary, causes the imaging load optical axis that is installed on the carrier aircraft to rock, and image quality descends.The high resolving power airborne remote sensing platform that inertially stabilized platform is formed is the effective measures that address this problem.Inertially stabilized platform is used for isolating the angular motion of carrier aircraft, follows the tracks of local geographical level, eliminates disturbance torque to the influence of imaging load, makes imaging load stablize imaging.The airborne remote sensing inertially stabilized platform is a branch of spiral shell tenacious tracking device, and its 26S Proteasome Structure and Function and traditional airborne photoelectric hang the storehouse similarity, hangs the storehouse but be different from traditional airborne photoelectric.Photoelectricity hangs the storehouse only needs to carry antenna, the first-class little load of guiding, thus photoelectricity to hang the storehouse generally lighter, load-bearing capacity is little; And airborne remote sensing need be carried the bigger imaging load of quality with inertially stabilized platform, and for the imaging load of many camera lenses, its quality can reach 100kg especially, and the weight of platform self requires generally to be no more than 40kg simultaneously.In low speed, big bearing ratio, the finite motion scope back and forth during the high-precision real under the angular motion condition control be the fundamental characteristics of inertially stabilized platform.

Lasting accuracy is one of the key technical indexes of inertially stabilized platform, has reflected the inhibition ability of stable platform to disturbance torque.The disturbance torque of platform mainly comprises unbalanced moments, moment of friction, and control system is generally by the effective inhibition of FEEDBACK CONTROL realization to various disturbance torques.Imaging load requires inertially stabilized platform to realize high control accuracy usually under low tracking angular rate, and moment of friction is the principal element that influences in the low-speed performance factor, directly influences system's transmission accuracy and robust motion.Moment of friction presents the non-linear of height, and its size is mainly relevant with structure and the factors such as lubrication circumstances, load size and angular velocity thereof of axle system, and moment of friction may be with position, angle, time and variation of temperature generation random variation.Three frameworks of inertially stabilized platform generally drive by rolling bearing supporting, torque motor and mechanical reduction gear, because the low carrying of working speed is big, antifriction-bearing raceway surface in contact lubrication circumstances is relatively poor, and the low speed jerking motion phenomenon takes place easily, causes bumpy motion and control accuracy to reduce; In addition, bigger stiction makes transmission gear generation elastic deformation, thereby causes site error, reduces systemic resolution and repetition rate., in the design of airborne remote sensing inertially stabilized platform control system, only moment of friction is suppressed as a disturbing moment before, do not considered that the nonlinear characteristic of moment of friction is to the influence of total system stability.Along with the requirement to system performance improves constantly, the harmful effect that the non-linear and dynamic perfromance of moment of friction is brought to control system more and more can not be ignored, and this just need more effectively eliminate or suppress the moment of friction in the control system.Elimination or inhibition moment of friction can be considered from two aspects the harmful effect that control system causes: be the method that adopts some reduction moment of frictions the physical construction design on the one hand.Yet the method for designing of pure machinery tends to be subject to processing the restriction of conditions such as technological level, material horizontal and reasearch funds, and is not easy to safeguard, to such an extent as to be difficult to obtain use; Then be the control law that to eliminate or to suppress the moment of friction influence from the control angle design of system on the other hand, namely control the friciton compensation method in field.Therefore, the friction model of research low-speed conditions underframe system also carries out friciton compensation, the friction link that exists in the reasonable resolution system is the model parameter problem of non-linear friction link especially, be to improve airborne remote sensing inertially stabilized platform system performance, realize the key that control requires to airborne remote sensing inertially stabilized platform high precision.

Summary of the invention

Technology of the present invention is dealt with problems and is: at the LuGre friction model difficult parameters of three inertially stabilized platform systems of the airborne remote sensing high precision control difficult problem with accurate identification, designed the parameter identification method that two step identifications combine with dynamic parameter optimization, architectural characteristic and disturbance factor according to three each frame systems of inertially stabilized platform, design different parameter identification schemes respectively, obtain static parameter and dynamic parameter initial value by two step discrimination methods, select suitable objective function that the dynamic parameter initial value is optimized again, obtain the dynamic parameter optimal value, obtain the more accurate LuGre model parameter of each frame system.This invention identification scheme is effective, can access each framework friction model comparatively accurately, for the development of high precision, high reliability, low-cost three inertially stabilized platforms provides technical support.

Technical solution of the present invention is: a kind of airborne remote sensing inertially stabilized platform friction parameter discrimination method, at first carry out the identification of friction model static parameter, structure airborne remote sensing inertially stabilized platform three framework servo-drive system PI closed loops, make three framework servomotors with different rotating speeds rotating equal angular slow running, obtain Stribeck negative slope section curve, record the electric current loop electric current under the corresponding rotating speed, match Stribeck curve carries out the static parameter identification then, simultaneously, stable platform three framework servo-drive systems are imported suitable electric current, obtain the steady state angle displacement; Secondly, carry out Identifying Dynamical Parameters, calculate the dynamic parameter initial value according to kinetic model, utilizing the dynamic parameter initial value is that objective function carries out dynamic parameter optimization with actual measurement moment of friction under the steady state conditions with the similarity degree of rate variation curve and friction model simulation curve.Concrete steps are as follows:

(1) adopts K _p+ K _i/ S controller is to airborne remote sensing inertially stabilized platform three framework servo electrical machinery systems structure PI control speed, current double closed-loop system, K _p+ K _i/ S is the controller architecture under Laplce's representation, K _pBe proportional component enlargement factor, K _i/ S represents integral element, K _iBe the integral element enlargement factor;

(2) under quiet pedestal, three framework servomotors are distinguished given different roll frame motor speed value W under load and speed closed-loop case _{L_R}, pitching frame motor speed W _{L_P}, orientation frame motor speed W _{L_A}, under the rotating speed of ± 0.5～± 6 °/s, move, make three framework servomotors forward or backwards constant speed rotate same angle, run on Stribeck negative slope section curve, measure roll frame electric current loop electric current I corresponding under the different rotating speeds constant-speed operation _R, pitching frame electric current loop electric current I _R, orientation frame electric current loop electric current I _AWherein, the Stribeck curve here is the framework rotating speed and the curve of corresponding moment of friction;

(3) adopt non-linear least square method, speed and current value that integrating step (2) records are to the stable state LuGre friction model of three frameworks $T_{\mathrm{fs}}=[T_c+(T_s-T_c)e^{-{\left(\frac{\mathrm{wl}}{\mathrm{ws}}\right)}^2}]\mathrm{sgn}\left(w_l\right)+{\mathrm{\σ}}_2{w}_{l.}$ Carry out match, obtain the static parameter T of the stable state LuGre friction model of three frameworks _s, T _c, ω _sAnd σ ₂

(4) static parameter that match obtains according to step (3), pitch axis, roll spindle motor input current make its servomotor output torque between T _s+ T _iAnd T _c+ T _iBetween, wherein Ti is disturbance torque, records the corresponding stable state angular displacement in roll frame, pitching frame static friction zone _{S_R}And θ _{S_P}, azimuth axis motor input current makes the servomotor output torque between T _sAnd T _cBetween, record corresponding stable state angular displacement _{S_A}

(5) Identifying Dynamical Parameters is at first according to formula σ ₀≈ T _cSgn (ω _l)/θ _s, ω _lBe load end angular velocity, by step (3), (4) gained static parameter, obtain the dynamic parameter initial value σ of motion three frame models when stopping ₀, basis then

Try to achieve dynamic parameter initial value σ ₁

(6) dynamic parameter optimization is to survey moment of friction with the similarity degree objective function of rate variation curve and friction model simulation curve under the steady state conditions For the foundation that dynamic parameter is optimized, utilize Matlab optimization tool case, obtain dynamic parameter σ ₀With σ ₁Optimal value.

The present invention's advantage compared with prior art is:

(1) high resolving power of airborne remote sensing system, high precision imaging requirements inertially stabilized platform have high precision of real time control and lasting accuracy.Wherein, realize that high precision control is research emphasis under the low tracking angular rate of inertially stabilized platform.And moment of friction is the principal element that influences in the low-speed performance factor, directly influences system's transmission accuracy and robust motion.The invention solves three inertially stabilized platform low-speed conditions of airborne remote sensing underframe servo-drive system axle is the foundation of non-linear friction model and the high precision identification of non-linear friction model parameter, for realizing that by friciton compensation the high stable precision provides the basis, satisfies the demand of high resolving power airborne remote sensing.

(2) relative empirical method, the present invention utilizes the friction parameter discrimination methods of two step identifications and dynamic parameter optimization to obtain LuGre friction model parameter more accurately, can realize the compensation of high-precision friction square;

(3) the present invention is directed to the parameter identification that the LuGre model rubs, structure and the disturbance factor characteristics different according to stable platform three frameworks, to three Frame Design parameter identification schemes, it is more accurate to obtain three frameworks rubbing characteristics separately respectively;

(4) discrimination method of the present invention's proposition can effectively obtain quiet, the dynamic parameter of LuGre model, and therefore the friction model of setting up is more accurate, can realize the control of airborne remote sensing inertial platform high precision.

Description of drawings

Fig. 1 is the invention process step synoptic diagram;

Fig. 2 for airborne remote sensing with three inertially stabilized platform structural representations;

Fig. 3 is for using airborne remote sensing inertially stabilized platform three framework servo electrical machinery systems structure PI control speed closed-loop system function block diagram of the present invention.

Embodiment

Be illustrated in figure 2 as three inertially stabilized platform structural representations of airborne remote sensing, platform is respectively roll frame, pitching frame and orientation frame from outside to inside.The revolving shaft of roll frame is along heading, and the revolving shaft of pitching frame is along the flight spanwise, the revolving shaft of orientation frame vertically downward, M _x, M _y, M _zBe three torque motors, drive the rotation of roll frame, the rotation of pitching frame, the rotation of orientation frame respectively; G _x, G _y, G _zFor being installed in the rate gyro on each framework, responsive three frameworks are with respect to the rotational angular velocity of inertial space, A _x, A _yFor being installed in the accelerometer on the pitching frame, wherein A _xSensitive axes overlap A with the turning axle of roll frame _ySensitive axes overlap with the turning axle of pitching frame; R _x, R _y, R _zBe three rotary transformers, attitude information is provided.R _xBe used for measuring the roll frame with respect to the rotational angle of support, R _yBe used for measuring pitching with respect to the rotational angle of roll frame, R _zBe used for measuring the orientation frame with respect to the rotational angle of roll frame.Servo controller according to the rate gyro sensitivity to frame corners rate information and the attitude measured of accelerometer or POS and positional information produce control signal and give torque motor, torque motor output drive strength square rotates to offset disturbance torque and to drive framework, reaches purpose stable and that follow the tracks of.

As shown in figures 1 and 3, specific implementation method of the present invention is as follows:

(1) adopts K _p+ K _i/ S controller is to airborne remote sensing inertially stabilized platform three framework servo electrical machinery systems structure PI control speed, current double closed-loop system, stable platform three framework servo electrical machinery system PI control closed-loop system structured flowchart such as Fig. 3.

(2) under quiet pedestal, the pitch axis servomotor under load and speed closed-loop case, given different speed value W _{L_R}, make its forward or backwards constant speed rotate same angle, obtain Stribeck negative slope section curve, measure electric current loop electric current I corresponding under the different slow-speed of revolution constant-speed operations _RThe pitch axis friction model as shown in the formula:

$T_{\mathrm{fs}}=[T_c+(T_s-T_c)e^{-{\left(\frac{\mathrm{wl}}{w_s}\right)}^2}]\mathrm{sgn}\left(w_l\right)+{\mathrm{\σ}}_2{w}_l$

In the formula, T _FsMoment of friction when representing quiet pedestal stable state, T _s, T _c, ω _sAnd σ ₂Be four static parameters of friction model, T _sExpression maximum static friction moment, T _cExpression coulomb moment of friction, ω _sRepresent critical Stribeck angular velocity, σ ₂Expression viscous friction coefficient, ω _lExpression load end angular velocity.

T _FsComputing method are as follows: the positive and negative rotation of roll axle constant speed θ angle, and system's stress balance equation is:

$\left\{\begin{array}{c}{N}_r{k}_T{I}_{\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HDA00003453425700022.png"\; state="\{\{state\}\}">r</figure-callout>}1}=-\mathrm{mg}\&CenterDot;\sin \mathrm{\&theta;}\&CenterDot;l_z-\mathrm{mg}\&CenterDot;\cos \mathrm{\&theta;}\&CenterDot;l_y+T_{\mathrm{rfsz}}\\ N_r{k}_T{I}_{r2}=-\mathrm{mg}\&CenterDot;\sin \mathrm{\&theta;}\&CenterDot;l_z+\mathrm{mg}\&CenterDot;\cos \mathrm{\&theta;}\&CenterDot;l_y+T_{\mathrm{rfsf}}\end{array}\right.$

Electric current and moment of friction are scalar in the formula, I _R1, I _R2Represent Electric Machine Control electric current when roll axle constant speed forward and backward respectively, T _Rfsz, T _RfsfRepresent the moment of friction when roll axle constant speed forward and backward respectively, k _TBe the motor torque coefficient.The roll motion scope is low-angle, and then sin θ approximate value is 0, can carry out the low-angle linearization to following formula, roll system friction moment T when obtaining stable state _RfsFor:

T _rfs≈N _rk _T(I _r1+I _r2)/2

N _rExpression roll system ratio of gear, k _TBe the motor torque coefficient, can be obtained by the motor instructions, obtain T behind each parameter value of substitution _RfsValue.Try to achieve T _RfsAfter the value, again according to the speed and the current value that record, adopt non-linear least square method match Stribeck curve, try to achieve static parameter T _s, T _c, ω _sAnd σ ₂

(3) identical with (2) step to the static parameter identification of roll axle framework, azimuth axis framework servo-drive system.

(4) Identifying Dynamical Parameters is at first according to formula σ ₀≈ T _cSgn (ω _l)/θ _s(ω _lExpression load end angular velocity), according to (2), (3) gained static parameter, the dynamic parameter σ of three frame models when obtaining motion and stopping ₀, basis then

Try to achieve dynamic parameter initial value σ ₁

(5) dynamic parameter optimization is to survey moment of friction with the similarity degree objective function of rate variation curve and friction model simulation curve under the steady state conditions

As the foundation that dynamic parameter is optimized, utilize Matlab optimization tool case, obtain dynamic parameter σ ₀With σ ₁Optimal value.

The content that is not described in detail in the instructions of the present invention belongs to this area professional and technical personnel's known prior art.

Claims (2)

1. airborne remote sensing inertially stabilized platform friction parameter discrimination method, it is characterized in that: at first carry out the identification of friction model static parameter, structure airborne remote sensing inertially stabilized platform three framework servo-drive system PI closed loops, make three framework servomotors with different rotating speeds rotating equal angular, slow running, obtain Stribeck negative slope section curve, record the electric current loop electric current under the corresponding rotating speed, match Stribeck curve then, carry out the static parameter identification, simultaneously, stable platform three framework servo-drive systems are imported suitable electric current, obtain the steady state angle displacement; Secondly, carry out Identifying Dynamical Parameters, calculate the dynamic parameter initial value according to kinetic model, utilizing the dynamic parameter initial value is that objective function carries out dynamic parameter optimization with actual measurement moment of friction under the steady state conditions with the similarity degree of rate variation curve and friction model simulation curve, specifically may further comprise the steps:

(1) adopts K _p+ K _i/ S controller is to airborne remote sensing inertially stabilized platform three framework servo electrical machinery systems structure PI control speed, current double closed-loop system, K _p+ K _i/ S is the controller architecture under Laplce's representation, K _pBe proportional component enlargement factor, K _i/ S represents integral element, K _iBe the integral element enlargement factor;

(2) under quiet pedestal, three framework servomotors are distinguished given different roll frame motor speed W under load and speed closed-loop case _{L_R}, pitching frame motor speed W _{L_P}, orientation frame motor speed W _{L_A}, under the rotating speed of ± 0.5～± 6 °/s, move, make three framework servomotors forward or backwards constant speed rotate same angle, run on Stribeck negative slope section curve, measure roll frame electric current loop electric current I corresponding under the different rotating speeds constant-speed operation _R, pitching frame electric current loop electric current I _P, orientation frame electric current loop electric current I _AWherein, the Stribeck curve here is the framework rotating speed and the curve of corresponding moment of friction;

(3) adopt non-linear least square method, speed and current value that integrating step (2) records are to the stable state LuGre friction model of three frameworks $T_{\mathrm{fs}}=[T_c+(T_s-T_c)e^{-{\left(\frac{\mathrm{wl}}{\mathrm{ws}}\right)}^2}]\mathrm{sgn}\left(w_l\right)+{\mathrm{\&sigma;}}_2{w}_l$ Carry out match, obtain the static parameter T of the stable state LuGre friction model of three frameworks _s, T _c, ω _sAnd σ ₂

(4) static parameter that match obtains according to step (3), pitch axis, roll spindle motor input current make its servomotor output torque between T _s+ T _iAnd T _c+ T _iBetween, T wherein _iBe disturbance torque, record the corresponding stable state angular displacement in roll frame, pitching frame static friction zone _{S_R}And θ _{S_P}, azimuth axis motor input current makes the servomotor output torque between Ts and Tc, records corresponding stable state angular displacement _{S_A}

(5) Identifying Dynamical Parameters is at first according to formula σ ₀≈ T _cSgn (ω _l)/θ _s, ω _lBe load end angular velocity, by step (3), (4) gained static parameter, obtain the dynamic parameter initial value σ 0 of motion three frame models when stopping, then basis

Try to achieve dynamic parameter initial value σ ₁

(6) dynamic parameter optimization is to survey moment of friction with the similarity degree objective function of rate variation curve and friction model simulation curve under the steady state conditions

For the foundation that dynamic parameter is optimized, utilize Matlab optimization tool case, obtain dynamic parameter σ ₀With σ ₁Optimal value.

2. according to the described a kind of airborne remote sensing inertially stabilized platform friction parameter discrimination method of claim 1, it is characterized in that: in the described step (3):

1. LuGre friction model friction model equation is:

$T_{\mathrm{fs}}=[T_c+(T_s-T_c)e^{-{\left(\frac{\mathrm{wl}}{\mathrm{ws}}\right)}^2}]\mathrm{sgn}\left(w_l\right)+{\mathrm{\&sigma;}}_2{w}_l$

In the formula, T _FsMoment of friction when representing quiet pedestal stable state, T _s, T _c, ω _sAnd σ ₂Be four static parameters of friction model, T _sExpression maximum static friction moment, T _cExpression coulomb moment of friction, ω _sRepresent critical Stribeck speed, σ ₂Expression viscous friction coefficient, ω _lExpression load end angular velocity;

2. T _FsComputing method are as follows: at first calculate roll axle moment of friction, roll frame constant speed is around the positive and negative rotation of roll axle θ angle, and system's stress balance equation is:

$\left\{\begin{array}{c}{N}_r{k}_T{I}_{r1}=-\mathrm{mg}\&CenterDot;\sin \mathrm{\&theta;}\&CenterDot;l_z-\mathrm{mg}\&CenterDot;\cos \mathrm{\&theta;}\&CenterDot;l_y+T_{\mathrm{rfsz}}\\ N_r{k}_T{I}_{r2}=-\mathrm{mg}\&CenterDot;\sin \mathrm{\&theta;}\&CenterDot;l_z+\mathrm{mg}\&CenterDot;\cos \mathrm{\&theta;}\&CenterDot;l_y+T_{\mathrm{rfsf}}\end{array}\right.$

Electric current and moment of friction are scalar in the formula, I _R1, I _R2Represent Electric Machine Control electric current when roll axle constant speed forward and backward respectively, T _Rfsz, T _RfsfRepresent the moment of friction when roll axle constant speed forward and backward respectively, N _rBe known roll system ratio of gear, k _TBe known motor moment coefficient, l _zExpression arm of force size, m is the platform quality, and g is acceleration of gravity, and the roll motion scope is low-angle, and then sin θ approximate value is 0, and following formula is carried out the low-angle linearization, roll system friction moment T when obtaining stable state _RfsFor:

T _rfs≈N _rk _T(I _r1+I _r2)/ ²。

Images