CN104476544A - Self-adaptive dead zone inverse model generating device of visual servo mechanical arm system - Google Patents

Abstract

The invention discloses a self-adaptive dead zone inverse model generating device of a visual servo mechanical arm system. The visual servo mechanical arm system is composed of a visual servo controller, a visual module, a motion control module, a driving module, a six-degree-of-freedom mechanical arm, a torque feedback module, a speed and position collecting module and a detecting module. The self-adaptive dead zone inverse model generating device of the visual servo controller is used for eliminating dead zone non-linear constraint and comprises a dead zone inverse model module, a self-adapting module and an operating module, wherein the dead zone inverse model module is used for structuring a smooth dead zone nonlinear inverse model; the self-adapting module adjusts estimated dead zone parameters through self-adaptive laws and transmits the parameters into the dead zone inverse model module to change the parameters of the inverse model; by means of the signal communication of the modules, a closed-loop control system can be formed inside the visual servo mechanical arm system. The self-adaptive dead zone inverse model generating device of the visual servo mechanical arm system can effectively eliminate the influence of the dead zone nonlinear restraint and achieve high image tracing precision.

Description

A kind of adaptive dead zone inversion model generating means of visual servo mechanical arm system

Technical field

The present invention relates to a kind of adaptive dead zone inversion model generating means of visual servo mechanical arm system, belong to visual servo mechanical arm system renovation technique on the input signals.

Background technology

The research of robot starts from 20 middle of century, and it grows up along with the development of automation and computer technology and atomic exploitation.Robot is in from simple to complexity always, from single to various, in development process from rudimentary to senior, along with the development of electronics technologies, the computing ability of computer is greatly improved with various sensor application in robot system, most of robots working environment is pre-set, so once any change occurs working environment, just need to reset robot system.So just strongly limit the range of application of industrial robot.In order to change this situation, robot is made to adapt to the environment of various change, people by various external sensor as sense of touch, distance, power feel and vision sensor be applied in robot control system, in order to provide extensively abundant information to robot without the need to carrying out contact type measurement to environment, and then making robot have the wider scope of application and higher performance, people are robot introducing vision sensor.Among the vision servo system research of mechanical arm, the Intelligent Control Strategy between robotic vision sensor and actuator is with a wide range of applications, and has become a study hotspot.So-called visual servo, the target position information obtained by vision sensor is exactly as feedback, thus the robot control system P of construction location closed loop.。Visual servo carrys out control motion by automatic acquisition and evaluating objects image.In a word, visual servo is exactly the principle utilizing machine vision, carries out fast processing to image feedback information, and provides position control signal fast according to processing the information obtained, and forms the robot controlling of position closed loop.

Early 1970s, people start computer vision and robot servo's combine with technique.Shirai and Inoue is the scholar the earliest vision being applied to robot system.The visual servo mode that they adopt is the position that the image information collected by vision system calculates estimating target, and this is a kind of visual spatial attention mode of static state.The connection of vision and robot system is open loop, so the positioning precision of robot is relevant with vision system precision and robot precision.Wherein, vision system precision is mainly subject to the impact of vision calibration error and resolution ratio.And robot precision is main and joint position sensor accuracy, robot inverse move

Learn the factors such as model accuracy, joint control algorithm, backlash and robot flexibility relevant.According to the difference of feedback information, visual servo is mainly divided into visual servo based on image and location-based visual servo two class.Based on the visual servo of image directly with the image error of calculation, pass to the motion that vision controller carrys out planning robot.This method does not need the position calculating target, insensitive to the pose of robot, but the design difficulty of controller is larger.Location-based visual servo calculates the position of target according to the pose of image and robot self, and vision controller carrys out the motion of planning robot according to target location.The advantage of this method is the signal feeding back to vision controller is position signalling, and interior ring controller needs is also position signalling, and therefore the design of vision controller realizes all relatively easy.But the method needs the three-dimensional information of computed image, obtain the space coordinates of target, meanwhile, the location-dependent query of target is in the pose of robot and video camera, more responsive to the calibrated error of robot and video camera.

But current key issue is: first, general band visual servo mechanical arm system all needs to demarcate camera, but camera calibration is a loaded down with trivial details and job for inefficiency, although there have been many methods to be used to the demarcation of vision system, the cost of camera calibration is still very high.The process asked for external parameter and the inner parameter of video camera is the demarcation of video camera.The operation principle of vision system is exactly obtain two-dimensional image information from video camera, calculates the geological informations such as the shape of object in three-dimensional environment, position, and rebuilds three-dimensional body thus.On two dimensional image, the position of every bit is relevant to the geometric position of this body surface respective point in three dimensions.This correlation is decided by video camera imaging geometrical model, and the parameter of geometrical model is also called camera parameters, mainly comprises outer parameter and intrinsic parameter.Its China and foreign countries' parameter refers to that camera coordinates ties up to the expression in reference frame, and intrinsic parameter then mainly comprises the amplification coefficient of imaging plane coordinate to image coordinate, the image coordinate, distortion coefficients of camera lens etc. of optical axis center point.Camera calibration provides contacting between professional camera and non-measured video camera.And so-called non-measured video camera refers to that its inner parameter is completely unknown, part the unknown or in principle uncertain such class video camera.Camera calibration is exactly the inside and outside parameter being obtained video camera by calibration experiment; Second, because gear is difficult to the loss with service time and number of times of complete compact occlusion and machinery, various non-linear factor can be there is in the moment input of mechanical arm system, the effect that not only greatly can affect control more can cause servo-drive system unstable, and wherein more common is exactly the constraint of input dead-time voltage.When driver provides input torque for motor, the dead-time voltage existed shows as engaged gears and occurs overlapping, when certain limit moment of resistance can be applied on motor accurately, but when moment exceeds dead band threshold values, show as the defeated moment be applied on motor and present threshold values characteristic, keep the input maximum in dead band, control performance is acutely declined and even can cause instability.

Summary of the invention

The object of the invention is to the adaptive dead zone inversion model generating means considering a kind of visual servo mechanical arm system that above-mentioned visual servo mechanical arm control system proposes by dead-time voltage effect of constraint value on the input signals.

Technical scheme of the present invention is: the adaptive dead zone inversion model generating means of visual servo mechanical arm system, includes Visual servoing control device (1), vision module (2), motion-control module (3), driver module (4), sixdegree-of-freedom simulation (5), torque-feedback module (6), speed (7) forms with station acquisition module (8) and detection module (9), Visual servoing control device (1) is by computer control unit (15), control signal generating unit (11), adaptive dead zone inversion model generating means (12), self adaptation camera calibration device (13), and communication unit (14) forms, it is characterized in that, the error signal of the image path formation of the real image track that Visual servoing control device (1) reception graphics processing unit (23) obtains and expectation, the position signalling gathered by station acquisition module (7), the rate signal gathered by speed acquisition module (8), the torque signals gathered by torque-feedback module (6), by Computing control unit (15) computing, information exchange is carried out by self adaptation camera calibration device (13) on-line proving camera by communication unit (14) (21) between servo vision controller (1) and vision module (2), dead band is built inverse and act on control signal by adaptive dead zone inversion model generating means (12), transmitted control signal to control module (3) by control signal generating unit (11), motion-control module (3) modulation (PWM) ripple moves in driver module (4) drive motors driver mechanical arm (5), detect current of electric, speed and the positional information in driver module (4) by detection module (9), and feedback and motion-control module (3) realize closed-loop control, the image coordinate of vision module (2) harvester mechanical arm (5) End features point the input of feeding back in controller (1), form the closed-loop control of the Visual servoing control system of band dead-time voltage constraint.

Above-mentioned adaptive dead zone inversion model generating means (12), comprises adaptation module (121), dead band inversion model module (122), operation control module (123); Adaptation module (121) comprises adaptive law memory (1211), deadzone parameter adjusted value memory (1212), deadzone parameter initial value storing value (1213); Dead band inversion model module (122) comprises parameter storing value (1221), power amplifier (1222), Piezoelectric Driving (1223), threshold values circuit (1224), circuit for generating (1225); Operation control module (123) comprises operand store (1231) (1234), integration module (1232), multiplication and division computing module (1233), plus and minus calculation module (1234).

In above-mentioned adaptive dead zone inversion model generating means (12), the parameter storage (1221) in dead band inversion model module (122) accepts the data transmitted by communication unit (14), operand store (1231), obtains the precompensation parameter vector of dead band inversion model and then regulation and control power amplification circuit (1222) put on the magnitude of voltage that electrical verification drives (1223), threshold values circuit (1224) utilizes voltage signal to carry out switch control rule to circuit for generating (1225), and the control signal for deadband eliminating non-linear effects that dead band inversion model module (122) produces is sent in control signal generating unit (11) by last circuit for generating; Deadzone parameter adjusted value memory (1212) and parameter storage (1221) carry out camera precompensation parameter and dead band information exchange simultaneously.

In above-mentioned adaptive dead zone inversion model generating means (12), the adaptive law memory (1211) in adaptation module (121) stores the programming code of adaptive law, can be expressed as by mathematical form:

${\stackrel{\·}{\widehat{\mathrm{\β}}}}_j=-{\mathrm{\Λ}}_{{\mathrm{\β}}_j}^{-1}{\widehat{\mathrm{\ω}}}_j{s}_{q_j}$

Wherein for the precompensation parameter vector of dead band inversion model derivative, for positive definite symmetric matrices, for dead space arrangements parameter prediction value, for joint velocity error, for the territory reference angular velocities relevant to vision system, for angular speed; Aforesaid parameter is for be delivered in adaptive law memory (1211) by parameter storage (1221) by deadzone parameter adjusted value memory (1212); At the system cloud gray model initial stage, by deadzone parameter initial value memory (1213) by initial information transfer in operand store (1234); After system cloud gray model, the data will carrying out calculating by adaptive law memory (1211) and deadzone parameter adjusted value memory (1212) are delivered in operand store (1234).

In above-mentioned adaptive dead zone inversion model generating means (12), the data received are carried out calculus, multiplication and division plus and minus calculation by integration module (1232), differential module (1232), multiplication and division computing module (1233) and plus-minus module (1233) by the operand store (1234) in operation control module simultaneously, then by calculate after data stored in operand store (1231), then transmit and parameter storage (1221) in.

Above-mentioned adaptive dead zone inversion model generating means (12) needs to be connected with self adaptation camera calibration device (13) with communication unit (14) by bus, parameter variable values in adaptive law memory (1211) is that vision module (2) is delivered to the position signalling and rate signal that gather in the data of controller and station acquisition module (8), speed acquisition module (7), and the parameter storage (1211) in dead band inversion model module (122) also stores the signal that torque-feedback module (6) is transmitted; The moment generation signal of deadband eliminating non-linear effects is produced by dead band inversion model (122).

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, Visual servoing control device (1) is communicated by communication unit (14) (21) with vision module (2), the model parameter of self adaptation camera calibration device (13) online Prediction vision module (2), set up the independent deep vision model of a non-demarcation, and the image that camera unit (24) is taken is undertaken processing the real image track obtaining characteristic point in real time by graphics processing unit (23) and Operations Analysis (22).

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, Visual servoing control device (1) receives the image error formed by input picture trajectory signal and the real image trajectory signal that obtained after image procossing by vision system, the joint of mechanical arm angle that receiving position acquisition module (8) and speed acquisition module (7) obtain, joint velocity, terminal position, receive the moment variations of moment after dead-time voltage module gathered by torque-feedback module (6), realize the collection to the positional information of mechanical arm, the movement locus of quantification machine mechanical arm, and the mechanical arm positional information expected directly is passed to motion-control module.

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, motion-control module (3) adopts dsp controller to realize Three-loop control and PWM controls; The most outer shroud of described Three-loop control is the position control ring realized by position control (31), a middle ring is the speeds control ring realized by speeds control (32), the current regulator of innermost ring for being realized by Current Control (33), described dsp controller communicates with control signal generating unit.

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, driver module (4) accepts the PWM modulation signal that PWM control (34) sends, motor (42) (43) of driver (41) rotating band dead band constraint, Electric Traction transmission device (44) also drags sixdegree-of-freedom simulation (5) motion thus.

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, what torque-feedback module (6) gathered driver (41) to put on motor without the moment of dead band constraint and the motor speed after the constraint of dead band, realizes the feedback of dead band constraint to Visual servoing control device (1).

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, detection module (9) realizes detecting and provides the closed loop feedback signal of Three-loop control, comprises QEP circuit (91) and frequency measurement circuit (92), photoelectric encoder (93), A/D converter (94), current sensor (95), the pulse signal transmission that photoelectric encoder (93) on machine shaft exports is to QEP circuit (91) and frequency measurement circuit (92), pulse signal obtains position feed back signal through QEP circuit (91) process, and the position control ring (31) sent in motion-control module (3), pulse signal is through frequency measurement circuit process, obtain feedback speed signal, and the rate control module (32) sent in motion-control module (3), current sensor (95) detects machine winding current, and obtain its digital current signal by A/D converter (94), sent to the current regulator (33) in motion-control module (3) again.

In the visual servo mechanical arm system of above-mentioned band dead-time voltage constraint, the characteristic point that sixdegree-of-freedom simulation (5) multiplely can be taken by camera unit (24) at end mark, graphics processing unit (23) detects, the image coordinate of this characteristic point is obtained by vision module (2).

The visual servo mechanical arm structure that the present invention adopts trick to be separated, namely camera is arranged on the fixed position being convenient to observe mechanical arm tail end characteristic point, by image path image transfer being extracted characteristic point to graphics processing unit of taking pictures.Fully take into account the problem of camera calibration in addition, by self adaptation camera calibration device online Prediction vision mode, decrease the complicated workload that calibration for cameras produces.Simultaneously the present invention also fully take into account mechanical arm system when face lose nonlinear restriction input, utilize adaptive dead zone inversion model generating means deadband eliminating nonlinear restriction, the adaptive law taked effectively can set up corresponding dead band inversion model.Experiment proves that this method reaches good effect, and the present invention is a kind of function admirable, be easy to build on the computer systems adaptive dead zone inversion model generating means.

Accompanying drawing explanation

The visual servo mechanical arm system the general frame of Fig. 1 band dead-time voltage constraint

Fig. 2 adaptive dead zone inversion model generating means theory diagram

Fig. 3 trick is separated visual servo physical arrangement schematic diagram

The inverse dead-zone model that Fig. 4 builds and the dead-time voltage schematic diagram that input torque passes through

Fig. 5 does not adopt adaptive dead zone inversion model generating means pursuit path schematic diagram

Fig. 6 adopts adaptive dead zone inversion model generating means pursuit path schematic diagram

The control block diagram of Fig. 7 visual servo mechanical arm control system

Detailed description of the invention

The present invention relates to a kind of dead band inversion model generating means of visual servo mechanical arm control system, the adaptive law of design is utilized to carry out the parameter of online Prediction dead-time voltage restricted model, build corresponding dead band inversion model again, can effectively eliminate input signal be tied lower Visual servoing control device controller mechanical arm make its End features point on the image plane progressive tracking expect image path, reach higher image trace precision.Below in conjunction with accompanying drawing and instantiation, the dead band inversion model generating means to the visual servo mechanical arm control system designed by the present invention is described in detail.

Fig. 1 is the visual servo mechanical arm system the general frame of band dead-time voltage constraint.The object designing Visual servoing control device is in FIG: when camera is not demarcated and mechanical arm input torque is subject to like to go nonlinear restriction, the motion of controller mechanical arm enables the projection on the image plane of the characteristic point on mechanical arm tail end follow the tracks of the desired image track of tracing preset.The error signal of the image path formation of the real image track that controller Visual servoing control device reception graphics processing unit obtains and expectation, by the position signalling of station acquisition module acquires, by the rate signal of speed acquisition module acquires, by the torque signals of torque-feedback module acquires, by the computing of Computing control unit, information exchange is carried out by self adaptation camera calibration device on-line proving camera by the communication unit between servo vision controller and vision module, dead band is built inverse and act on control signal by adaptive dead zone inversion model generating means, transmitted control signal to control module by control signal generating unit, motion-control module modulation (PWM) ripple moves in driver module drive motors driver mechanical arm, detect current of electric, speed and the positional information in driver module by detection module, and feedback and motion-control module realize closed-loop control, the image coordinate of vision module harvester mechanical arm End features point the input of feeding back in controller, form the closed-loop control of the Visual servoing control system of band dead-time voltage constraint, controller can according to image feedback, and velocity feedback adjusts controller in time and exports the image trace performance providing the best.

Fig. 7 is the control block diagram of visual servo mechanical arm control system, and this control block diagram is exactly the embodiment of principle installation drawing on controlling of figure mono-.

Fig. 2 is adaptive dead zone inversion model generating means theory diagram.Can obtain the part that Fig. 2 is Visual servoing control device from Fig. 1, the effect of this device is that structure one the dead band inversion model of on-line tuning can eliminate the impact of unbalanced input, to greatest extent also intrinsic input torque.Dead band inversion model module is according to the precompensation parameter vector obtained build dead band inversion model, the adaptive law designed is delivered in operation control module and calculates, according to adaptive law by adaptation module:

${\stackrel{\·}{\widehat{\mathrm{\β}}}}_j=-{\mathrm{\Λ}}_{{\mathrm{\β}}_j}^{-1}{\widehat{\mathrm{\ω}}}_j{s}_{q_j}$

Obtain parameter vector continuous corrected parameter vector in the running of system value, i.e. the slope k in dead band _r, k _land breakpoint h _r, h _ldiscreet value, the disturbance making the input torque of design can offset non-linear input to greatest extent to bring.

Fig. 3 is that the trick that the present invention adopts is separated visual servo physical arrangement schematic diagram, and camera is installed on the fixed position that is convenient to observe mechanical arm tail end characteristic point, and mechanical arm and camera are connected the exchange of the information of carrying out with computer by bus.Relative to structure camera being arranged on mechanical arm tail end, the structure adopting trick to be separated effectively can reduce the shake because manipulator motion causes camera to take pictures, and the global motion of mechanical arm can be observed clearly, obtain the global information of characteristic point, moving control module for controlling manipulator motion can be sent instructions to by the motion control card of computer.

Fig. 4 be analog input moment by dead band after dead-zone model again by schematic diagram that dead band retrains.Dead-time voltage can be described as:

$u_d\left(t\right)=\left\{\begin{array}{cc}{k}_r(v_d\left(t\right)-h_r)& v_d\left(t\right)\≥h_r\\ 0& h_l\≤v_d\left(t\right)\≤h_r\\ k_l(v_d\left(t\right)-h_l)& v_d\left(t\right)\≤h_l\end{array}\right.$

Wherein u _d, v _dfor dead band exports and input, k _r, k _l, h _r, h _lfor constant.Parametrization is carried out to above-mentioned model, and definition β _d=[k _r, k _rh _r, k _l, k _lh _l] ^t, can obtain the dead-zone model after parametrization is: $u_d\left(t\right)=-{\mathrm{\β}}_d^T{\mathrm{\ω}}_d.$ Wherein

Build a level and smooth dead band inversion model as follows:

$v_d=\frac{u_d+k_r{h}_r}{k_r}{\mathrm{\δ}}_r\left(u_d\right)+\frac{u_d+k_l{h}_l}{k_l}{\mathrm{\δ}}_l\left(u_d\right)$

${\mathrm{\δ}}_r\left(u_d\right)=\frac{e^{u_d/e_0}}{e^{u_d/e_0}+e^{-u_d/e_0}}$

${\mathrm{\δ}}_l\left(u_d\right)=\frac{e^{-u_d/e_0}}{e^{u_d/e_0}+e^{-u_d/e_0}}$

Wherein v _d, u _dfor output and the input of inversion model, δ _r(u _d) and δ _l(u _d) be continuous print indicator function.Parameter due to dead-zone model is unknown, therefore is the discreet value that can only adopt parameter in design, is also go to build level and smooth dead band inversion model:

$v_o=\frac{u_o+{\hat{k}}_r{\hat{h}}_r}{{\hat{k}}_r}{\mathrm{\δ}}_r\left(u_o\right)+\frac{u_o+{\hat{k}}_l{\hat{h}}_l}{{\hat{k}}_l}{\mathrm{\δ}}_l\left(u_o\right)$

Wherein v _oand u _odo not estimate output and the input of inversion model.Can find out that the deadband eliminating the effect of constraint value whether inversion model of structure can be maximum depends on parameter from formula above discreet value be enough accurate.Explain the discreet value how application self-adapting rule asks for parameter vector below in detail:

1) according to initiation parameter and the follow-up Input and output measurements of dead band inversion model, defining the dead band inversion model Product-factor estimated is:

Wherein v _ofor the output of dead band inversion model, with definition as noted above.

2) angle q, the angular speed of mechanical arm can be collected by acceleration module position module angular acceleration and position x, the tip speed of end if the disarthrial bar of mechanical arm the one is long is l ₁, l ₂.The Jacobin matrix obtaining mechanical arm is J (q (t)):

$J\left(q\left(t\right)\right)=\left[\begin{array}{ccc}-l_1*\sin \left(q\left(1\right)\right)& -l_2/\sin \left(q\left(2\right)\right)& 0\\ l_1*\cos \left(q\left(1\right)\right)& l_2*\sin \left(q\left(2\right)\right)& 0\\ 0& 1& 1\\ 0& 0& 1\end{array}\right]$

Wherein q (1), q (2), q (3) are the 1st in q, the 2nd, the 3rd element, in like manner according to the image coordinate y=[u, v] collected ^t, desired image track y _d=[u _d, v _d] ^t, and the precompensation parameter matrix that self adaptation camera calibration device transmits picture depth independence interaction matrix can be constructed for once form:

$\hat{A}\left(t\right)=\left(\begin{array}{c}{\hat{m}}_1^T-u*{\hat{m}}_3^T\\ {\hat{m}}_2^T-v*{\hat{m}}_3^T\end{array}\right)$

Wherein for matrix the one the second the third lines.And then following parameter matrix can be obtained

$Q=\hat{A}\left(t\right)*J\left(q\left(t\right)\right)$

According to the terminal position information collected, can obtain camera relative to the degree of depth of estimating of projection plane is:

$\hat{z}_c\left(t\right)={\hat{m}}_3^T*x$

Define the territory reference picture speed of the plane of delineation below, according to image error Δ y=y-y _d, have territory reference picture speed to be:

${\stackrel{\·}{y}}_r={\stackrel{\·}{y}}_d\left(t\right)-\mathrm{\λ}*\mathrm{\Δy}$

And then following parameter matrix can be obtained be:

${\hat{F}}_{\mathrm{yr}}=_c\hat{z}\left(t\right)*{\stackrel{\·}{y}}_r$

3) from above-mentioned parameter matrix, we can further the domain of definition with reference to joint velocity for:

${\stackrel{\·}{q}}_r={\hat{R}}^{+}*{\hat{F}}_{\mathrm{yr}}$

Wherein for generalized inverse, therefore joint velocity error vector can be obtained

4) according to the definition of appealing and the variate-value collected, the discreet value adaptive law about deadzone parameter vector can be constructed:

$\stackrel{\·}{\widehat{\mathrm{\β}}}=-{\mathrm{\Λ}}_{\mathrm{\β}}^{-1}\widehat{\mathrm{\ω}}{s}_q$

Wherein Λ _βfor positive definite symmetric matrices, can be passed data in operation control module by the adaptive law of appealing and carry out computing, obtain value, then have dead band inversion model module by constructing dead band inversion model.

So far, we have completed the structure of dead band inversion model, it should be noted that several steps of appeal complete in the inversion model generating means of adaptive dead zone, by the information exchange between each module, effectively can estimate deadzone parameter, guarantee that the control moment designed can be applied in mechanical arm accurately, the non-linear impact brought in deadband eliminating.We verify the performance of the adaptive dead zone inversion model generating means of design by experiment below.The image path of following the tracks of is chosen as follows:

$y_d\left(t\right)=\left(\begin{array}{c}40*\sin (0.2*t)+30\\ -40*\cos (0.2*t)-17\end{array}\right)$

It is β that the deadzone parameter chosen estimates initial value _d0=[1 ,-4,1 ,-4] ^t.

Fig. 5 is not for adopt adaptive dead zone inversion model generating means pursuit path schematic diagram.Wherein red curve is real image track, the blue image path curve for expecting.Can find out that actual track is difficult to follow the tracks of desired trajectory, there is larger image error, embody the performance impact that dead-time voltage constraint produces visual servo mechanical arm system, be especially necessary so a design adaptive dead zone inversion model generating means seems.

Fig. 6 is for have employed adaptive dead zone inversion model generating means pursuit path schematic diagram.Clearly can see that actual path just follows the tracks of desired trajectory rapidly in the incipient stage, at follow-up phase progressive tracking desired trajectory especially, embody the superperformance of the adaptive dead zone inversion model generating means except design.

Claims (13)

1. an adaptive dead zone inversion model generating means for visual servo mechanical arm system, includes Visual servoing control device (1), vision module (2), motion-control module (3), driver module (4), sixdegree-of-freedom simulation (5), torque-feedback module (6), speed (7) forms with station acquisition module (8) and detection module (9), Visual servoing control device (1) is by computer control unit (15), control signal generating unit (11), adaptive dead zone inversion model generating means (12), self adaptation camera calibration device (13), and communication unit (14) forms, it is characterized in that, the error signal of the image path formation of the real image track that Visual servoing control device (1) reception graphics processing unit (23) obtains and expectation, the position signalling gathered by station acquisition module (7), the rate signal gathered by speed acquisition module (8), the torque signals gathered by torque-feedback module (6), by Computing control unit (15) computing, information exchange is carried out by self adaptation camera calibration device (13) on-line proving camera by communication unit (14) (21) between servo vision controller (1) and vision module (2), dead band is built inverse and act on control signal by adaptive dead zone inversion model generating means (12), transmitted control signal to control module (3) by control signal generating unit (11), motion-control module (3) modulation (PWM) ripple moves in driver module (4) drive motors driver mechanical arm (5), detect current of electric, speed and the positional information in driver module (4) by detection module (9), and feedback and motion-control module (3) realize closed-loop control, the image coordinate of vision module (2) harvester mechanical arm (5) End features point the input of feeding back in controller (1), form the closed-loop control of the Visual servoing control system of band dead-time voltage constraint.

2. adaptive dead zone inversion model generating means (12) according to claim 1, it is characterized in that, device comprises adaptation module (121), dead band inversion model module (122), operation control module (123); Adaptation module (121) comprises adaptive law memory (1211), deadzone parameter adjusted value memory (1212), deadzone parameter initial value storing value (1213); Dead band inversion model module (122) comprises parameter storing value (1221), power amplifier (1222), Piezoelectric Driving (1223), threshold values circuit (1224), circuit for generating (1225); Operation control module (123) comprises operand store (1231) (1234), integration module (1232), multiplication and division computing module (1233), plus and minus calculation module (1234).

3. adaptive dead zone inversion model generating means (12) according to claim 1, it is characterized in that the parameter storage (1221) in dead band inversion model module (122) accepts the data transmitted by communication unit (14), operand store (1231), obtain the precompensation parameter vector of dead band inversion model and then regulation and control power amplification circuit (1222) put on the magnitude of voltage that electrical verification drives (1223), threshold values circuit (1224) utilizes voltage signal to carry out switch control rule to circuit for generating (1225), and the control signal for deadband eliminating non-linear effects that dead band inversion model module (122) produces is sent in control signal generating unit (11) by last circuit for generating; Deadzone parameter adjusted value memory (1212) and parameter storage (1221) carry out camera precompensation parameter and dead band information exchange simultaneously.

4. adaptive dead zone inversion model generating means (12) according to claim 1, it is characterized in that the adaptive law memory (1211) in adaptation module (121) stores the programming code of adaptive law, can be expressed as by mathematical form:

${\stackrel{.}{\widehat{\mathrm{\β}}}}_j=-{\mathrm{\Λ}}_{{\mathrm{\β}}_j}^{-1}{\widehat{\mathrm{\ω}}}_j{s}_{q_j}$

Wherein for the precompensation parameter vector of dead band inversion model derivative, for positive definite symmetric matrices, for dead space arrangements parameter prediction value, for joint velocity error, for the territory reference angular velocities relevant to vision system, for angular speed; Aforesaid parameter is for be delivered in adaptive law memory (1211) by parameter storage (1221) by deadzone parameter adjusted value memory (1212); At the system cloud gray model initial stage, by deadzone parameter initial value memory (1213) by initial information transfer in operand store (1234); After system cloud gray model, the data will carrying out calculating by adaptive law memory (1211) and deadzone parameter adjusted value memory (1212) are delivered in operand store (1234).

5. adaptive dead zone inversion model generating means (12) according to claim 1, it is characterized in that the data received are carried out calculus, multiplication and division plus and minus calculation by integration module (1232), differential module (1232), multiplication and division computing module (1233) and plus-minus module (1233) by the operand store (1234) in operation control module simultaneously, then by calculate after data stored in operand store (1231), then transmit and parameter storage (1221) in.

6. adaptive dead zone inversion model generating means (12) according to claim 1, it is characterized in that in Visual servoing control device (1), adaptive dead zone inversion model generating means (12) needs to be connected with self adaptation camera calibration device (13) with communication unit (14) by bus, only can obtain from claim 4, parameter variable values in adaptive law memory (1211) is data and the station acquisition module (8) that vision module (2) is delivered to controller, the position signalling gathered in speed acquisition module (7) and rate signal, parameter storage (1211) in dead band inversion model module (122) also stores the signal that torque-feedback module (6) is transmitted, the moment generation signal of deadband eliminating non-linear effects is produced by dead band inversion model (122).

7. the visual servo mechanical arm system of band dead-time voltage constraint according to claim 1, it is characterized in that above-mentioned Visual servoing control device (1) is communicated by communication unit (14) (21) with vision module (2), the model parameter of self adaptation camera calibration device (13) online Prediction vision module (2), set up the independent deep vision model of a non-demarcation, and the image that camera unit (24) is taken is undertaken processing the real image track obtaining characteristic point in real time by graphics processing unit (23) and Operations Analysis (22).

8. the visual servo mechanical arm system of band dead-time voltage constraint according to claim 1, it is characterized in that above-mentioned Visual servoing control device (1) receives the image error formed by input picture trajectory signal and the real image trajectory signal that obtained after image procossing by vision system, the joint of mechanical arm angle that receiving position acquisition module (8) and speed acquisition module (7) obtain, joint velocity, terminal position, receive the moment variations of moment after dead-time voltage module gathered by torque-feedback module (6), realize the collection to the positional information of mechanical arm, the movement locus of quantification machine mechanical arm, and the mechanical arm positional information expected directly is passed to motion-control module.

9. the visual servo mechanical arm system of band dead-time voltage constraint according to claim 1, is characterized in that above-mentioned motion-control module (3) adopts dsp controller to realize Three-loop control and PWM controls; The most outer shroud of described Three-loop control is the position control ring realized by position control (31), a middle ring is the speeds control ring realized by speeds control (32), the current regulator of innermost ring for being realized by Current Control (33), described dsp controller communicates with control signal generating unit.

10. the visual servo mechanical arm system of band dead-time voltage constraint according to claim 1, it is characterized in that above-mentioned driver module (4) accepts PWM and controls (34) PWM modulation signal of sending, motor (42) (43) of driver (41) rotating band dead band constraint, Electric Traction transmission device (44) also drags sixdegree-of-freedom simulation (5) motion thus.

The visual servo mechanical arm system of 11. band dead-time voltage constraints according to claim 1, what it is characterized in that above-mentioned torque-feedback module (6) gathers driver (41) to put on motor without the moment of dead band constraint and the motor speed after the constraint of dead band, realizes the feedback of dead band constraint to Visual servoing control device (1).

The visual servo mechanical arm system of 12. band dead-time voltage constraints according to claim 1, it is characterized in that above-mentioned detection module (9) realizes detecting and provides the closed loop feedback signal of Three-loop control, comprise QEP circuit (91) and frequency measurement circuit (92), photoelectric encoder (93), A/D converter (94), current sensor (95), the pulse signal transmission that photoelectric encoder (93) on machine shaft exports is to QEP circuit (91) and frequency measurement circuit (92), pulse signal obtains position feed back signal through QEP circuit (91) process, and the position control ring (31) sent in motion-control module (3), pulse signal is through frequency measurement circuit process, obtain feedback speed signal, and the rate control module (32) sent in motion-control module (3), current sensor (95) detects machine winding current, and obtain its digital current signal by A/D converter (94), sent to the current regulator (33) in motion-control module (3) again.

The visual servo mechanical arm system of 13. band dead-time voltage constraints according to claim 1, it is characterized in that the characteristic point that above-mentioned sixdegree-of-freedom simulation (5) multiplely can be taken by camera unit (24) at end mark, graphics processing unit (23) detects, the image coordinate of this characteristic point is obtained by vision module (2).