CN105123127B - Medlar-picking machine device people and its control method - Google Patents
Medlar-picking machine device people and its control method Download PDFInfo
- Publication number
- CN105123127B CN105123127B CN201510670042.0A CN201510670042A CN105123127B CN 105123127 B CN105123127 B CN 105123127B CN 201510670042 A CN201510670042 A CN 201510670042A CN 105123127 B CN105123127 B CN 105123127B
- Authority
- CN
- China
- Prior art keywords
- picking
- sensor
- medlar
- coordinate
- matrimony vine
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Harvesting Machines For Specific Crops (AREA)
Abstract
The invention discloses a kind of medlar-picking machine device people, medlar-picking machine device people system uses pin-hole imaging technology, mainly includes data collecting card and analog-to-digital conversion module.Its control mode mainly uses Fuzzy PID self-adaptive control, and mainly visual aspects are controlled.Its hardware aspect is mainly made up of mechanical arm, rotor structure, picking head and various motors.The present invention can realize the automatic harvesting of matrimony vine, and manpower is substituted completely, the problem of manually plucking protected from environmental is solved.Its high degree of automation, picking efficiency is high.
Description
【Technical field】
The present invention relates to mechanical automation design field, more particularly to a kind of medlar-picking machine device people.
【Background technology】
Matrimony vine is the medicinal material with health and disease cure function, can also regard the fruit with abundant nutrition value to eat
With.In NORTHWEST CHINA, northeast, the substantial amounts of plantation in the north.The big pericarp of matrimony vine water content is relatively thin to belong to berry, easy when winning
Sustain damage.The harvesting work of matrimony vine is always based on manpower, and matrimony vine mature period Relatively centralized, labor intensity is larger.And
Matrimony vine is general ripe in summer 6, July, and the season belongs to summer time, and temperature extremely rises.It is general in matrimony vine ground to have nature extensive use of chemical fertilizer fertilizer
Deng sharp aroma.But present matrimony vine harvesting is artificial harvesting, and general band mouth mask operation exacerbation is scorching during manpower harvesting
Hot degree.Not because of picking efficiency underground, and sense of discomfort is brought to workman in hot day, may seriously cause heatstroke situation.
For at present, although picking robot has a larger development and progress, but appoint and so there is asking for following several respects
Topic.
1st, position and recognize that work can be poor.Environment is relatively complicated in picking process, and leaf is easily repeated with fruit
Together, cause identification difficult.And weather has unstability, while the also interference information such as noise, reduces identification and fixed
The degree of accuracy of position.Identification and positioning to matrimony vine are required further study.
2nd, picking efficiency is not high.Various countries are all high without man efficiency with the picking efficiency for all kinds of picking robots researched and developed.
The strawberry picking rate that for example agriculture university professor Zhang Tiezhong develops is 6 pieces/min；The automatic picking robot of apple of South Korea, is plucked
Taking one piece of apple averagely needs 15s；The cucumber picking robot of Dutch IMAG research and development, which plucks a cucumber, needs 45s.Since it is desired that
Control system also wants image procossing etc., inefficient.Fruit and vegetable picking robot is really applied to actual production, just have to
Improve operating efficiency and the operation degree of accuracy.
3rd, cost is higher.Fruit and vegetable picking robot system and structure are more complicated, and usage cycles are short, usually substantially not
With.Safeguard, use, manufacturing cost will be higher than industrial robot.
4th, poor universality.A kind of picking robot completes single task, all kinds of machines just for a certain specified plant
Almost without versatility between people.
In order to break through the development factors of limitation fruit and vegetable picking robot, the problem of must just solving the above could so add
The development of fast fruit and vegetable picking robot, realizes commercialization as early as possible.
【The content of the invention】
It is an object of the invention to provide a kind of medlar-picking machine device people, the robot can combine the growth characteristic of matrimony vine
Require to be plucked with harvesting, manpower is substituted completely and completes harvesting work.
In order to achieve the above object, the present invention is achieved using following technical scheme：
Power supply and dynamic control device, roadblock sensing are provided with medlar-picking machine device people, including dolly, described dolly
Device, mechanical arm, picking head and collecting fruit case；Wherein, roadblock sensor is arranged on dolly headstock；The mechanical arm upper end and
Picking head is connected, and the sensor being provided with branch clip, branch clip for judging matrimony vine position is provided with picking head；Adopt
Pluck a bottom and be provided with the fruit delivery pipe connected with collecting fruit case；The electric control system of described dolly, the electricity of mechanical arm
Gas control system, the electric control system of picking head, roadblock sensor, the electric control system of branch clip and sensor with
Power supply and dynamic control device connection.
As a further improvement on the present invention, described mechanical arm include be sequentially connected lifting platform, waist rotary joint,
Large arm, forearm and prism joint, lifting platform bottom are arranged on dolly, and the rotary joint rotation of driving waist is provided with lifting platform
Waist motor and driving large arm rotation large arm motor, forearm motor and drive that driving forearm is rotated are provided with the top of large arm
The flexible prism joint motor in dynamic prism joint；Prism joint upper end and picking head connection；The electric control system of lifting platform, waist
Portion's motor, large arm motor, forearm motor and prism joint motor are connected with power supply and dynamic control device.
As a further improvement on the present invention, the adjustable height of described lifting platform is 0~0.8m；Waist rotary joint revolves
Gyration is -160 °~160 °；The anglec of rotation of large arm is -80 °~80 °, the anglec of rotation of forearm is -80 °~80 ° and prism
The collapsing length in joint is 0~0.5m.
As a further improvement on the present invention, described picking head includes the space surrounded by shell body and upper lid, and top
Portion opens up harvesting entrance；The other end being provided with picking head close to upper lid one end in upper base plate and lower seat board, picking head is set
There is bedplate；Direct current generator is arranged on upper lid by motor cabinet, and the front end of direct current generator is fixed on upper base plate, and direct current
The output shaft of machine is provided with idler gear；Left-hand rotation daughter and right-hand rotation daughter are arranged side by side into combined rotor, left-hand rotation daughter and right-hand rotation
The two ends of daughter are separately positioned on lower seat board and bedplate, and left-hand rotation daughter and right-hand rotation daughter are all provided with one end of lower seat board
Rotor gear is equipped with, two rotor gears and idler gear are meshed, and left-hand rotation daughter and right-hand rotation daughter rotation direction are opposite；Harvesting
The bottom of head is provided with harvesting outlet, and harvesting outlet is connected with fruit delivery pipe.
As a further improvement on the present invention, described left-hand rotation daughter and right-hand rotation daughter include rotor block, and described turns
Daughter includes having three flexible pipes being uniformly arranged along helix on cylinder, cylinder；Its middle column body is by with the soft of mechanical strength
Property silicone tube is made, and hole array is drilled with according to the distance every 15mm in three hands of spiral of silicone tube, by hollow hoses
Worn with silicone tube along hole array with nylon wire and connect fixation, the projection of the hollow hoses formation arcuate structure between adjacent holes；Turn left
The hand of spiral of daughter and the hollow hoses in right-hand rotation daughter is opposite.
As a further improvement on the present invention, crash sensor, position sensor, pressure are provided with described branch clip
Force snesor and vision sensor；Crash sensor is arranged on the top of branch clip, position sensor and pressure sensor point
It is not arranged on two intermediate plates of branch clip, vision sensor is arranged on two intermediate plate bottoms of branch clip, collision sensing
Device, position sensor, pressure sensor and vision sensor are communicated with power supply and dynamic control device.
The control method of medlar-picking machine device people, comprises the following steps：
1) start, medlar-picking machine device people starts, dolly is walked under matrimony vine tree from collection；
2) recognize, the video camera on vision sensor carries out IMAQ using pin-hole imaging technology, feed back to power supply and
Dynamic control device；
3) position, the industrial computer in power supply and dynamic control device is handled the image of collection, carry out image recognition
Processing, identification region carries out position identification with the image of matrimony vine color-match and fixed；
4) pluck, industrial computer uses Fuzzy PID self-adaptive control, the motor of driving mechanical arm carries out position adjustment, position
When the branch that signal on sensor is had matrimony vine is blocked, start the direct current generator of picking head, matrimony vine enters at the top of picking head
The relative rotation that entrance carries out harvesting process, left-hand rotation daughter and right-hand rotation daughter is plucked, matrimony vine completes harvesting, and the matrimony vine plucked is led to
Fruit delivery pipe is crossed into collecting fruit case；After branch shelf upper position sensor signal recovers, picking head is stopped, and completes
The harvesting of single branch；Return to step, continues to pluck matrimony vine；
5) collect, when the weight of the collecting fruit case on medlar-picking machine device people exceedes preset value, medlar-picking machine device
People, which returns, collects ground, completes matrimony vine and collects process.
Further, the specific geometrical relationship of the position adjustment of mechanical arm is：Camera coordinate system (xc, yc, zc) and robot
Base coordinate system (x0, y0, z0) each reference axis be parallel to each other；Coordinate of the objective fruit in basis coordinates system is (Xm0, Ym0,
Zm0), its coordinate in camera coordinate system is (Xn, Yn, Zn), will in harvesting when prism joint is contracted in forearm
Picking robot regards only three articulated robot processing, the transformation relation formula between camera coordinates and robot coordinate as
For：
Y=X × ctg θ1
Z=L1+L2×sinθ2+L3×sin(θ2+θ3) (1)
L in formula1Represent waist, L2Represent big arm lengths, L3Represent the length of forearm；θ1、θ2、θ3Respectively mechanical arm
Three rotary freedoms in centre joint angles；
Coordinate (X of the objective fruit in basis coordinates system can be obtained by (1) formulam0,Ym0,Zm0) with ripe matrimony vine in video camera
Coordinate (the X of coordinate system0,Y0,Z0) relation：
Xn=Xmo- X, Yn=Ymo- Y, Zn=Zmo-Z (2)。
Further, in position fixing process, the specific method for building up of relation of camera coordinates and position coordinates is to set up coordinate system
U, v, each pixel can represent its line number and columns with the coordinate (u, v) corresponding to it, resettle image coordinate system
(X, Y), it has physical unit (mm), O0X-axis is parallel to u axles, O0Y-axis takes optical axis in the intersection point O of image parallel to v axles0For
Origin；Take O0Coordinate in U, V coordinate systems is (u0, v0), dx, dy is distance of the pixel on two axles, then in image
All pixels point has following coordinate relation：
During 2 points of coincidences, then have：
Further, image recognition processing is concretely comprised the following steps：
Matrimony vine color in the image of collection is compared industrial computer with the matrimony vine color in default image, meets then
Position identification is carried out to fix；
The weight preset value of collecting fruit case is 30Kg~32Kg；
Also include roadblock rate-determining steps and anticollision rate-determining steps, described roadblock rate-determining steps are：Medlar-picking machine device
Obstacle on robot travel path is identified roadblock sensor on people, when being judged as barrier, feeds back signal to
Power supply and dynamic control device are adjusted the direct of travel of dolly, are avoided；Described anticollision rate-determining steps are：Collision
Sensor carries out being identified for barrier during branch clip carries out clamping branch, prevents branch clip collision damage
Bad sensor.
Compared with prior art, the present invention has advantages below：
Hardware aspect of the present invention is mainly made up of mechanical arm, rotor structure, picking head and various motors, passes through mechanical electric
The combination of gas, realizes robot instead of artificial labor type, has liberated productivity.And robot has image recognition, Chinese holly
Qi is plucked, branch positioning, and the function that keeps in obscurity, it is possible to achieve the operating process for voluntarily plucking matrimony vine of robot, according to matrimony vine
The rotation dynamics of copper plate body in the size design picking head of the weight capacity of ripening fruits, in guarantee fruit shapes and compared with less breakage
The harvesting fruit of rate.And the sensor such as visual servo is designed beside picking head, it is ensured that the normal harvesting of picking robot and
Correct work.Good response characteristic, each motion link is intended to the progress of continuous effective.Strong antijamming capability, it is light and
Small volume.Maintenance is easy, and without the need for maintenance.Stable, characteristic is identical during rotating.By to medlar-picking machine device
People needs the understanding for the function of realizing.Design the theoretical model of medlar-picking machine device people.For the work(to be realized of robot
Can, initial setting, and theory analysis are carried out to the control of its software.The problem of for being likely to occur in picking process, carry out just
Step is estimated to avoid and controls the research in adjustable extent as far as possible in terms of hardware design.
Further, mechanical arm is made up of five frees degree, realizes the rotation or flexible in five free degree directions, position is adjusted
Whole quick and positioning is more accurate.
Further, the particular design of picking head is key modules.The characteristics of having considered matrimony vine, is devised screw type
The mechanism of flexible pipe and colloidal silica, is rotated against, and the kneading of helical structure carries out the harvesting of matrimony vine, can either realize that harvesting again will not
Chinese wolfberry fruit is damaged, and harvesting works well.
Further, branch clip is provided with four different sensors of function, it is ensured that the automatic harvesting of robot can
Implementation.The module that keeps in obscurity must be set up when being limited in harvesting due to harvesting environment.The intelligence of the robot can be embodied with this
Change direction.The position of the not only possible decision tree branch of position sensor, moreover it is possible to feed back to control system and carry out starting picking head
Motor, realizes the continuous realization that can automatically control, it is to avoid the idle running of picking head, has saved the energy and realizes and automates.
In the method for the present invention, medlar-picking machine device people system uses pin-hole imaging technology, mainly including data collecting card
And analog-to-digital conversion module.Its control mode mainly uses Fuzzy PID self-adaptive control, and mainly visual aspects are controlled.Depending on
It is nucleus module to feel control and color acquisition, and control mode mainly uses Self Adaptive Control.Matrimony vine is identified, positioned,
Pluck, the flow such as collection, Robotics Autonomous's system is accurate, high degree of automation.
Further, the corresponding relation of camera coordinates and position coordinates, as successful key can be plucked.Can be accurate
The position for navigating to Chinese wolfberry fruit.
Further, roadblock rate-determining steps and anticollision rate-determining steps are effectively prevented outer during robot self-operating
Disturb on boundary, it is ensured that the normal picking process of robot.
【Brief description of the drawings】
Fig. 1 is the main body schematic diagram of medlar-picking machine device people in the present invention；
Fig. 2 is picking head structural representation in the present invention；
Fig. 3 is rotor structure schematic diagram in the present invention；
Fig. 4 is combined rotor structure schematic diagram in the present invention；
Fig. 5 is branch clip sensor distribution map in the present invention；
Fig. 6 is picking robot structural relation schematic diagram；
Fig. 7 is picking robot joint geometrical relationship；
Fig. 8 is camera coordinates and the relation schematic diagram of position coordinates；
Fig. 9 is medlar-picking machine device people's Model Reference Adaptive Control System structure in the present invention；
Figure 10 is self-tuning adaptive control system construction drawing；
Figure 11 is that medlar-picking machine device people plucks flow chart；
Figure 12 is that medlar-picking machine device people plucks main program control flow chart；
Figure 13 is harvesting control flow chart；
Figure 14 is fruit reclaimer control flow chart.
In figure, 1. dollies；2. power supply and dynamic control device；3. lifting platform；4. waist motor；5. large arm motor；6. it is big
Arm；7. forearm；8. branch clip；9. picking head；10. rotary body；11. direct current generator；12. forearm motor；13. fruit is conveyed
Pipe；14. collecting fruit case；15. roadblock sensor；10-1. cylinders；10-2. flexible pipes；16. motor cabinet；17. on cover；18. the seat of honour
Plate；19. lower seat board；20. shell body；21. left-hand rotation daughter；22. right-hand rotation daughter；23. bedplate；24. idler gear；25. rotor
Gear；8-1. crash sensors；8-2. position sensors；8-3 pressure sensors；8-4 vision sensors.
【Embodiment】
The present invention is described in further detail below in conjunction with the accompanying drawings, and described is explanation of the invention rather than limit
It is fixed.
As shown in figure 1, a kind of medlar-picking machine device people of the invention, including dolly 1；Be provided with the one end of dolly 1 power supply and
Dynamic control device 2, the other end is provided with roadblock sensor 15, and its top is additionally provided with mechanical arm.Mechanical arm includes connecting successively
Lifting platform 3, waist rotary joint, large arm 6, forearm 7 and the prism joint connect, lifting platform 3 is provided with waist motor 4, the and of large arm 6
Large arm electricity motor 5；The upper end of large arm 6 and forearm 7 are connected, and the upper end of forearm 7 is connected by prism joint and picking head 9, the bottom of forearm 7
Portion is provided with forearm motor 12；The driving waist rotary joint rotation of waist motor 4, large arm electricity 5 drives large arm 6 to rotate, forearm electricity
The driving forearm 7 of machine 12 is rotated, and prism joint motor driving prism joint is stretched；The one end of picking head 9 is provided with branch clip 8, tree
There is sensor branch clip 8 upper end.The bottom of picking head 9 is provided with the fruit delivery pipe 13 connected with collecting fruit case 14, collecting fruit
Case 14 is placed on dolly 1.The inside of picking head 9 includes travelling gear and the copper plate body 10 being connected with travelling gear；Travelling gear
Drive copper plate body 10 to rotate and carry out matrimony vine harvesting.
Specifically, the hardware composition of the present invention is as follows：
As shown in Figure 1.Mechanical arm is PRRRP structures, is made up of five frees degree, the effect of each free degree of mechanical arm and class
Type is as shown in table 1：
The type of each free degree of table 1 and effect
The major parameter of its mechanical arm of medlar-picking machine device people is as shown in table 2：
The major parameter in each joint of table 2：
The wherein flexible Feng Ding in prism joint selects U.S. DANAF] ER electric pushrods, its body model：EC2-X-10-
04A-400-ME2-MTM-23X, its flexible stroke is 400mm.Maximum speed Vmax=0.3m/s.Normal thrust：F=120N.
As shown in Figure 2；Picking head 9 includes the space surrounded by shell body 20 and upper lid 17, and top starts picking mouth；Adopt
Pluck the other end being provided with first 9 close to upper one end of lid 17 in upper base plate 18 and lower seat board 19, picking head 9 and be provided with bedplate
23；Motor cabinet 16 is arranged on lid 17, and direct current generator 11 is arranged in motor cabinet 16, and the front end of direct current generator 11 is fixed on
On seat board 18, and the output shaft of direct current generator 11 is connected with idler gear 24, the two ends point of left-hand rotation daughter 21 and right-hand rotation daughter 22
It is not arranged on lower seat board 19 and bedplate 23, and is provided with rotor gear 25, gear 25 and idler gear 24 is meshed.When
When branch clip 8 fixes branch, the power supply of picking head 9 is turned on automatically, and direct current generator 11 drives central gear 24, central gear 24
The rotor gear 25 on left and right rotor block (21,22) is driven again, and left and right rotor block (21,22), which is rotated against, to be started to pluck matrimony vine.
Shown in Fig. 3 and Fig. 4.The structure of rotor block 10 includes：Cylinder 10-1, specification：20 millimeters of diameter, 3 millimeters of wall thickness, length
140 millimeters, material：There is the flexible silicone tube of certain mechanical strength.10-1 has been uniformly arranged three folders on the cylinder of silicone tube
Angle is the flexible pipe 10-2 of 120 ° of spiral line type.Flexible pipe 10-2 is 3 millimeters of diameter, and 1 millimeter of wall thickness, length is 50 millimeters of sky
Heart cylindrical structure.Each hollow cylindrical flexible pipe 10-2 spacing is 15 millimeters, and the flexible pipe 10-2 of 3 spirals is along silicone tube vertical table
Surface helix is set.In the hand of spiral of silicone tube according to every 15mm apart from the hole array that drill diameter is 1 millimeter.Again by sky
Soft-hearted pipe 10-2 and silicone tube are connected with 1mm nylon wire, so silica gel cylinder 10-1 mechanical strength and nylon wire
Flexibility combine, it is possible to achieve matrimony vine pluck.Combined rotor structure, including the two left-hand rotation daughters 21 and the right side being arranged side by side
Rotor block 22 is constituted, and left-hand rotation daughter 21 is opposite with the hand of spiral of the flexible pipe in right-hand rotation daughter 22.
As shown in figure 5, sensor distribution map on branch clip 8 in the present invention, in installed different sensors, have position
Put, in terms of there are visual and some aerodynamic pressures.In these sensors, position sensor is to utilize infrared ray
Positioning principle realize, visually sensor can be realized using the camera of very high pixel.The addition of sensor causes machine
The perception of device people to external world is more powerful, and during work, efficiency is significantly improved.Above
Various kinds of sensors used in these, the more intelligent base for having laid identification of robot of harvesting fruit can be made for us
Plinth.Crash sensor 8-1, position sensor 8-2, pressure sensor 8-3, visual sensing are provided with specific branch clip 8
Device 8-4.Crash sensor 8-1 is arranged on the top of branch clip 8, prevents damaged in collision branch clip 8 and each sensor；Position
Put sensor 8-2 and pressure sensor 8-3 to be separately positioned on two intermediate plates of branch clip 8, position sensor 8-2 monitoring trees
Branch jam, and feed back to control system play direct current generator 11 carry out harvesting process；Vision sensor 8-4 is arranged on tree
Two intermediate plate bottoms of branch clip 8, the position for judging matrimony vine feeds back to the adjustment that control system carries out robot location.
Medlar-picking machine device people's positive kinematics are as follows：
As shown in Figure 6 and Figure 7, in picking process, robot must judge that the position of target branch and mechanical arm are each first
The relation having between joint, first determines the position of matrimony vine branch bar, determines its relation between each coordinate system.So
The coordinate system relation between video camera and robot is determined afterwards.
Pluck the structural relation of matrimony vine robot as shown in figs. 2-9, the vision sensor 8-4 of picking robot video camera
Mounting means uses eye-in-hand.Camera coordinate system (xc, yc, zc) and robot base coordinate sys-tem (x0, y0, z0)
Each reference axis is parallel to each other.Assuming that coordinate of the objective fruit in basis coordinates system is (Xm0, Ym0, Zm0), it is in camera coordinate system
Interior coordinate is (Xn, Yn, Zn).3D coordinate (X of the target in camera coordinate system is obtained by video camera imaging general principlen,
Yn,Zn) with its 2D coordinates (x in video camera imaging plane1,y1) between relation can draw.
In harvesting when prism joint is contracted in forearm, regard picking robot as only three joint machines
People is handled.Conversion between camera coordinates and robot coordinate can be obtained with method shown in Fig. 7 to solve.Physical relationship formula
For：
Y=X × ctg θ1
Z=L1+L2×sinθ2+L3×sin(θ2+θ3) (1)
L in formula1Represent waist, L2Represent big arm lengths, L3Represent the length of forearm；θ1、θ2、θ3Respectively mechanical arm
Three rotary freedoms in centre joint angles.
Coordinate (X of the objective fruit in basis coordinates system can be obtained by formula above formulam0,Ym0,Zm0) with ripe matrimony vine in video camera
Coordinate (the X of coordinate system0,Y0,Z0) relation：
Xn=Xmo- X, Yn=Ymo- Y, Zn=Zmo-Z (2)。
Medlar-picking machine device people's vision sensor 8-4 of the present invention uses pin-hole imaging technology, mainly including data collecting card
And analog-to-digital conversion module, its control mode mainly use Fuzzy PID self-adaptive control, mainly visual aspects are controlled.Depending on
Feel that control and color acquisition are nucleus modules.Control mode mainly uses Self Adaptive Control.Vision sensor 8-4 video camera is sat
Mark and the corresponding relation of position coordinates, as successful key can be plucked.
The vision mode composition of the present invention is as follows：
Pin-hole imaging, throws on photo-sensitive cell through the light that camera lens is collected into, optical signal is converted to by photo-sensitive cell
Electric signal；Electric signal is converted into data signal；Data signal is transferred to after being compressed on monitor or recording arrangement, so that
Show or record into it can be seen that image.
The foundation of plane of delineation coordinate system, the image information being collected into is inputted into calculator memory storage in electrical signal form is
M row N columns groups.The pixel of each array represents the contrast between image in array.
The relation of camera coordinates and position coordinates, as shown in figure 8, setting up coordinate system u, v, each pixel can use it
Corresponding coordinate (u, v) represents its line number and columns.Because (u, v) only represents the array of pixel, figure is not represented
Position as in where pixel.So, it would be desirable to image coordinate system (X, Y) is resettled, it has physical unit (mm).O0X
Axle is parallel to u axles, O0Y-axis takes optical axis in the intersection point O of image parallel to v axles0For origin.The point heart in the picture in general
Place, but a little deviate because some reasons of video camera can also be produced.Take O0Coordinate in U, V coordinate systems is (u0, v0),
Dx, dy are distance of the pixel on two axles, then all pixels point in image has following coordinate relation：
During 2 points of coincidences, then have：
Now calculate and more facilitate.
1st, control system of the invention composition is as follows：
1st, data collecting card, medlar-picking machine device people work realizes that sensor is in whole process by sensor
In can gather various signals and be controlled.Sensor spread out of some simulation signal, analog signal applied to computer it
It is preceding all to carry out digital-to-analogue conversion.Data collecting card connects external environment and control system well, data acquisition
Card suffers from simulation output, simulation input, numeral output input port.Simulation input is itself for a capture card
Function compared with based on.It is to utilize sampling hold circuit, amplifier, and variable connector and critically important analog-digital converter have come
Into.
2nd, Self Adaptive Control, as shown in Figure 9 and Figure 10.The kinetic model of robot exist it is non-linear and it is uncertain because
Element, containing unknown coefficient factor (such as frictional force) and non-linear dynamic characteristic (gravity, coriolis force, centripetal force it is non-linear), with
And robot in the course of the work the property and changing features of environment and target when, it is to run that one of solution, which is,
The feature of controll plant is constantly measured in journey, control system is realized the optimal control of closed loop according to new characteristic according to the information of measurement
System, i.e. Self Adaptive Control.
When the kinetic model of robot has non-linear and uncertain factor, containing unknown coefficient factor (such as frictional force)
With non-linear dynamic characteristic (gravity, coriolis force, centripetal force it is non-linear), and robot environment and work in the course of the work
When Properties of Objects and changing features, it is the feature for constantly measuring controll plant in the process of running, root that one of solution, which is,
Control system is set to realize closed loop optimum control, i.e. Self Adaptive Control according to new characteristic according to the information of measurement.Self Adaptive Control point
For：Model Reference Adaptive Control System is as shown in Figure 9；Self-correcting Adaption Control System is as indicated by 10：
When controlled system parameter changes, adaptive control system can be advised by identification, study to adjust control
Rule.But the suitable complexity of its implementation process, it is desirable to it is too harsh.When Parameter uncertainties, system is unable to reach stabilization.Robust
Control is to be directed to the probabilistic another control strategy of robot, can make up the deficiency of Self Adaptive Control, be adapted to not true
Determine the situation that factor changes in certain scope, it is ensured that the stability and the certain performance indications of maintenance of system.If by Shandong
Rod is combined with H ∞ control theories, gained controller can the effective attenuation of unknown disturbances to external world in fact, while ensure system with
The asymptotic Convergence Property of track error
Attached most importance in the control aspect of medlar-picking machine device people with the control of visual aspects.For medlar-picking machine device people's
Movement characteristic, selection image based visual servo control.According to the difference between identification image and default image, computer
System control machine device people is constantly moved close to realization of goal motion control.By PID control method and fuzzy logic control methodology phase
With reference in each joint for applying to robot.Visual Servoing System is adjusted using fuzzy control strategy online adaptive
Pid parameter, can strengthen the adaptive ability of fuzzy controller, improve the dynamic and static performance of control system.
Robot of the present invention is operated as is illustrated by figs. 11 and 12, medlar-picking machine device people's control flow：Video camera is fixed on
The lower section of branch clip 8, and branch clip 8 is immediately below matrimony vine branch, it can be moved with mechanical arm, remote apart from ground,
Obtain image small, high resolution.Correctness and rapidity to the processing of collection image determine Visual Servoing System
Can normal work.Accurate positional information can be improved for robot localization.The operation principle of the system is：1, pass through shooting
Head collection image information；2, image is handled through image processing software,；3, the positional information of objective fruit is obtained,；4, meter
Calculate the deviation of present image and desired image；5, it is mapped as control signal (such as speed or joint of joint of robot controller
Angular displacement increment), control signal is sent to each joint with industrial computer；6, each joint of servo driver drives picking robot is done
Correspondingly move,.The harvesting function of robot is realized with this.As shown in figure 12, harvesting control flow also includes being arranged on dolly 1
Barrier in the roadblock sensor 15 of front end, the robot of detection in real time advance route, is effectively avoided.
As shown in figure 13, it is harvesting control flow chart, including beginning vision sensor 8-4 is taken pictures, and is carried out color and is sentenced
It is disconnected, determine whether matrimony vine；Such as no, return restarts；It is judged as Chinese wolfberry fruit, then carries out positioning clamping, mainly position
Sensor 8-2 is operated, and starts the associated motor of mechanical arm, position adjustment is carried out, until position sensor 8-2 signal
Blocked by branch, be then judged as having clamped, feed back to control system, the direct current generator 11 for starting picking head 9 is operated, and is adopted
Pluck head and carry out matrimony vine harvesting, until the signal on position sensor 8-2 is reconnected due to harvesting matrimony vine, be then judged as the tree
Branch harvesting is completed, and is stopped, and carries out the harvesting of next branch.Complete a picking process.
As shown in figure 14, it is fruit reclaimer control flow chart；When in picking process, the sensing in collecting fruit case 14
Device judges that its weight is more than 30Kg, during less than 32Kg, and the motor on the mechanical arm of robot stops, and robot returns to matrimony vine and collected
Ground carries out the collection of harvesting matrimony vine.Process is once collected in completion.
The present invention operation principle be：When harvesting ripe FRUCTUS LYCII, vision sensor 8-4 (colour recognition device) is identified as
Ripe matrimony vine, the fixation branch of branch clip 8, which is entangled, to be stuck in clip.Now the position of picking head 9 is matrimony vine location,
Electric motor starting simultaneously drives the rotation left and right rotor of idler gear 25 to start to relatively rotate in succession, and silicagel column 10-1 starts rotation harvesting Chinese holly
Qi.When matrimony vine constantly reaches collecting box 14 by picking head harvesting from matrimony vine delivery pipe 13.When on branch shelf 8 sensor from tree
The harvesting of a branch is completed when departing from branch.When the supporting power of fruit case reaches 30Kg, it is past specified that robot starts return
Stack Chinese wolfberry fruit in place.
Claims (7)
1. it is provided with power supply and power control on medlar-picking machine device people, it is characterised in that including dolly (1), described dolly (1)
Control equipment (2), roadblock sensor (15), mechanical arm, picking head (9) and collecting fruit case (14)；Wherein, roadblock sensor (15)
It is arranged on dolly (1) headstock；The mechanical arm upper end and picking head (9) connection, picking head are provided with branch clip on (9)
(8) sensor for judging matrimony vine position, is provided with branch clip (8)；Picking head (9) bottom is provided with and collecting fruit case
(14) fruit delivery pipe (13) of connection；The electric control system of described dolly (1), the electric control system of mechanical arm, adopt
Pluck electric control system, roadblock sensor (15), the electric control system of branch clip (8) and the sensor of head (9) with electricity
Source and dynamic control device (2) connection；
Lifting platform (3), waist rotary joint, large arm (6), forearm (7) and the prism that described mechanical arm includes being sequentially connected are closed
Section, lifting platform (3) bottom is arranged on dolly (1), and the waist electricity of driving waist rotary joint rotation is provided with lifting platform (3)
The forearm that driving forearm (7) is rotated is provided with the top of machine (4) and the large arm motor (5) of driving large arm (6) rotation, large arm (6)
The flexible prism joint motor of motor (12) and driving prism joint；Prism joint upper end and picking head (9) connection；Lifting platform
(3) electric control system, waist motor (4), large arm motor (5), forearm motor (12) and prism joint motor is and power supply
And dynamic control device (2) connection；
The adjustable height of described lifting platform (3) is 0~0.8m；The waist rotary joint anglec of rotation is -160 °~160 °；Large arm
(6) collapsing length that the anglec of rotation is -80 °~80 °, the anglec of rotation of forearm (7) is -80 °~80 ° and prism joint is 0
~0.5m；
Described picking head (9) includes the space surrounded by shell body (20) and upper lid (17), and top opens up harvesting entrance；Adopt
The other end being provided with upper base plate (18) and lower seat board (19), picking head (9) close to upper lid (17) one end in head (9) is plucked to set
There is bedplate (23)；Direct current generator (11) is arranged on upper lid (17) by motor cabinet (16), and the front end of direct current generator (11) is consolidated
It is scheduled on upper base plate (18), and the output shaft of direct current generator (11) is provided with idler gear (24)；Left-hand rotation daughter (21) and right-hand rotation
Daughter (22) is arranged side by side into combined rotor, and the two ends of left-hand rotation daughter (21) and right-hand rotation daughter (22) are separately positioned on lower seat board
(19) and on bedplate (23), left-hand rotation daughter (21) and right-hand rotation daughter (22) are provided with one end of lower seat board (19) and turned
Sub- gear (25), left-hand rotation daughter (21) and right-hand rotation daughter (22) are provided with rotor gear on one end of lower seat board (19)
(25), two rotor gears (25) and idler gear (24) are meshed, left-hand rotation daughter (21) and right-hand rotation daughter (22) rotation direction
Conversely；The bottom of picking head (9) is provided with harvesting outlet, and harvesting outlet is connected with fruit delivery pipe (13).
2. medlar-picking machine device people according to claim 1, it is characterised in that described left-hand rotation daughter (21) and right rotor
Body (22) includes rotor block (10), and described rotor block (10) includes having three along spiral shell on cylinder (10-1), cylinder (10-1)
The hollow hoses (10-2) that spin line is uniformly arranged；Its middle column body (10-1) is made up of the flexible silicone tube with mechanical strength,
Hole array is drilled with according to the distance every 15mm in three hands of spiral of silicone tube, by hollow hoses (10-2) and silicone tube
Worn along hole array with nylon wire and connect fixation, the projection of hollow hoses (10-2) the formation arcuate structure between adjacent holes；Left rotor
The hand of spiral of hollow hoses (10-2) on body (21) and right-hand rotation daughter (22) is opposite.
3. medlar-picking machine device people according to claim 1, it is characterised in that be provided with described branch clip (8)
Crash sensor (8-1), position sensor (8-2), pressure sensor (8-3) and vision sensor (8-4)；Crash sensor
(8-1) is arranged on the top of branch clip (8), and position sensor (8-2) and pressure sensor (8-3) are separately positioned on branch folder
On two intermediate plates of sub (8), vision sensor (8-4) is arranged on two intermediate plate bottoms of branch clip (8), crash sensor
(8-1), position sensor (8-2), pressure sensor (8-3) and vision sensor (8-4) with power supply and dynamic control device
(2) communicate.
4. the control method of medlar-picking machine device people, it is characterised in that comprise the following steps：
1) start, medlar-picking machine device people starts, dolly (1) is walked under matrimony vine tree from collection；
2) recognize, the video camera on vision sensor (8-4) carries out IMAQ using pin-hole imaging technology, feed back to power supply and
Dynamic control device (2)；
3) position, the industrial computer in power supply and dynamic control device (2) is handled the image of collection, carry out at image recognition
Reason, identification region carries out position identification with the image of matrimony vine color-match and fixed；
4) pluck, industrial computer uses Fuzzy PID self-adaptive control, the motor of driving mechanical arm carries out position adjustment, position sensing
When the branch that signal on device (8-2) is had matrimony vine is blocked, start the direct current generator (11) of picking head (9), matrimony vine enters harvesting
Harvesting entrance at the top of head (9) carries out the relative rotation of harvesting process, left-hand rotation daughter (21) and right-hand rotation daughter (22), and matrimony vine completes
Harvesting, the matrimony vine plucked enters collecting fruit case (14) by fruit delivery pipe (13)；Branch shelf (8) upper position sensor
After (8-2) signal recovers, picking head (9) is stopped, and completes the harvesting of single branch；Return to step (2), continues to pluck Chinese holly
Qi；
5) collect, when the weight of the collecting fruit case (14) on medlar-picking machine device people exceedes preset value, medlar-picking machine device
People, which returns, collects ground, completes matrimony vine and collects process.
5. the control method of medlar-picking machine device people according to claim 4, it is characterised in that the position adjustment of mechanical arm
Specific geometrical relationship be：Camera coordinate system (xc, yc, zc) and robot base coordinate sys-tem (x0, y0, z0) each reference axis phase
It is mutually parallel；Coordinate of the objective fruit in basis coordinates system is (Xm0, Ym0, Zm0), its coordinate in camera coordinate system is (Xn,
Yn, Zn), in harvesting when prism joint is contracted in forearm, regard picking robot as only three articulated robots
Handle, the transformation relation formula between camera coordinates and robot coordinate is：
Y=X × ctg θ1
Z=L1+L2×sinθ2+L3×sin(θ2+θ3) (1)
L in formula1Represent waist, L2Represent big arm lengths, L3Represent the length of forearm；θ1、θ2、θ3Respectively in mechanical arm
Between three rotary freedoms joint angles；
Coordinate (X of the objective fruit in basis coordinates system can be obtained by (1) formulam0,Ym0,Zm0) with ripe matrimony vine in camera coordinates
Coordinate (the X of system0,Y0,Z0) relation：
Xn=Xmo- X, Yn=Ymo- Y, Zn=Zmo-Z (2)。
6. the control method of medlar-picking machine device people according to claim 4, it is characterised in that in position fixing process, shooting
The specific method for building up of relation of machine coordinate and position coordinates is to set up coordinate system u, v, and each pixel can be with corresponding to it
Coordinate (u, v) represents its line number and columns, resettles image coordinate system (X, Y), it has physical unit (mm), O0X-axis is put down
Row is in u axles, O0Y-axis takes optical axis in the intersection point O of image parallel to v axles0For origin；Take O0Coordinate in U, V coordinate systems is
(u0, v0), dx, dy is distance of the pixel on two axles, then all pixels point in image has following coordinate relation：
During 2 points of coincidences, then have：
7. the control method of medlar-picking machine device people according to claim 4, it is characterised in that the tool of image recognition processing
Body step is：
Matrimony vine color in the image of collection is compared industrial computer with the matrimony vine color in default image, and satisfaction is then carried out
Position identification is fixed；
The weight preset value of collecting fruit case (14) is 30Kg~32Kg；
Also include roadblock rate-determining steps and anticollision rate-determining steps, described roadblock rate-determining steps are：On medlar-picking machine device people
Roadblock sensor (15) obstacle on robot travel path is identified, when being judged as barrier, feed back signal to
Power supply and dynamic control device (2) are adjusted the direct of travel of dolly (1), are avoided；Described anticollision rate-determining steps
For：Crash sensor (8-1) carries out being identified for barrier, prevented during branch clip (8) carries out clamping branch
Branch clip (8) damaged in collision sensor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510670042.0A CN105123127B (en) | 2015-10-16 | 2015-10-16 | Medlar-picking machine device people and its control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510670042.0A CN105123127B (en) | 2015-10-16 | 2015-10-16 | Medlar-picking machine device people and its control method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105123127A CN105123127A (en) | 2015-12-09 |
| CN105123127B true CN105123127B (en) | 2017-07-11 |
Family
ID=54709025
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510670042.0A Expired - Fee Related CN105123127B (en) | 2015-10-16 | 2015-10-16 | Medlar-picking machine device people and its control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105123127B (en) |
Families Citing this family (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105729469B (en) * | 2016-02-17 | 2017-12-22 | 中国科学院自动化研究所 | A kind of peg-in-hole assembly method and system based on environment domain of attraction |
| CN107018754A (en) * | 2017-04-26 | 2017-08-08 | 广西职业技术学院 | Automatic classification orange picking machine |
| CN107637278A (en) * | 2017-04-27 | 2018-01-30 | 解晗 | Automatic picking robot |
| CN107018761A (en) * | 2017-06-20 | 2017-08-08 | 山西省农业科学院旱地农业研究中心 | Tomato automatically plucks dress basket all-in-one |
| CN107564209A (en) * | 2017-08-21 | 2018-01-09 | 朱明� | A kind of unattended closing facility office system and its processing method |
| CN107787681A (en) * | 2017-10-17 | 2018-03-13 | 江苏大学 | Robot system and implementation method of a kind of grape cluster fruit quickly without the harvesting that comes off |
| CN107862253B (en) * | 2017-10-20 | 2021-07-16 | 武汉科技大学 | Wolfberry picking effect evaluation method |
| CN108200809A (en) * | 2017-11-21 | 2018-06-26 | 铜陵市棵松农业科技有限公司 | A kind of picking fruit robot and control system |
| CN107836208B (en) * | 2017-12-15 | 2023-05-16 | 武汉轻工大学 | Fruit picking robot |
| CN108227496A (en) * | 2018-01-08 | 2018-06-29 | 石河子大学 | A kind of improvement fuzzy PID control method for the accurate operation of mechanical arm of connecting |
| CN108848929B (en) * | 2018-04-20 | 2020-12-08 | 宁夏大学 | Multi-point rotating clamping type picking device for picking medlar |
| CN108834579B (en) * | 2018-08-10 | 2021-02-02 | 遵义华邦农业科技开发有限公司 | Pepper picking device |
| CN109220225A (en) * | 2018-10-22 | 2019-01-18 | 湖北理工学院 | A kind of full-automatic fruit picker |
| CN109579702A (en) * | 2018-12-19 | 2019-04-05 | 合肥工业大学 | A kind of fructus lycii branch pull-up point position intelligent locating method |
| CN110865534A (en) * | 2019-11-15 | 2020-03-06 | 合肥工业大学 | Intelligent following system with improved Kalman filtering for navigation positioning |
| CN111492803A (en) * | 2020-05-18 | 2020-08-07 | 刘吉安 | Spiral fruit and vegetable picking device and picking method |
| CN111482976A (en) * | 2020-05-18 | 2020-08-04 | 刘吉安 | Bidirectional spiral fruit and vegetable picking device and picking method |
| CN112556606A (en) * | 2020-12-24 | 2021-03-26 | 宁夏农林科学院农业经济与信息技术研究所（宁夏农业科技图书馆） | Self-propelled wolfberry fruit actual measurement method and device based on binocular vision |
| CN112715159A (en) * | 2020-12-30 | 2021-04-30 | 武汉晴川学院 | Automatic citrus picking method and device based on mechanical arm |
| CN114946408B (en) * | 2022-06-09 | 2023-11-28 | 华南农业大学 | Four-wheel type multifunctional orchard mechanical device for facility cultivation melon |
| CN116267236B (en) * | 2023-05-23 | 2023-08-25 | 四川省机械研究设计院(集团)有限公司 | Cluster Fruit Picking Robot |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103283397A (en) * | 2012-03-02 | 2013-09-11 | 王宁州 | Portable wolfberry picking machine |
| CN203646089U (en) * | 2013-12-16 | 2014-06-18 | 西北农林科技大学 | Kiwi fruit picking robot |
| CN103947380B (en) * | 2014-04-18 | 2015-11-11 | 金陵科技学院 | A kind of both arms fruit picking robot and fruit picking method |
| CN205030148U (en) * | 2015-10-16 | 2016-02-17 | 北方民族大学 | Machine people is plucked to matrimony vine |
-
2015
- 2015-10-16 CN CN201510670042.0A patent/CN105123127B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN105123127A (en) | 2015-12-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105123127B (en) | Medlar-picking machine device people and its control method | |
| US11397174B2 (en) | Suspension slide rail platform-based greenhouse information automatic monitoring method | |
| WO2019134454A1 (en) | Suspension rail type greenhouse comprehensive information automatic cruise monitoring device | |
| CN205030148U (en) | Machine people is plucked to matrimony vine | |
| CN109725233B (en) | Intelligent substation inspection system and inspection method thereof | |
| CN109699300B (en) | Control system and control method of intelligent citrus picking machine | |
| CN103959973B (en) | One becomes more meticulous crop fertilization system and nitrogen fertilizing method | |
| CN103293156B (en) | System and method for estimating growth conditions of field crops | |
| CN109392462A (en) | Pick the robot and method of day lily | |
| CN104834313B (en) | A kind of greenhouse intelligence spray robot and method based on RFID | |
| CN104834312B (en) | A kind of farmland intelligence spray robot based on RFID | |
| CN103699095A (en) | Greenhouse plant growth posture monitoring system based on binocular stereo vision and greenhouse plant growth posture monitoring method based on binocular stereo vision | |
| WO2008136804A1 (en) | Agricultural automation system with field robot | |
| CN109197160A (en) | A kind of guava picking robot and its implementation | |
| CN209656661U (en) | A kind of wheat high throughput phenotypic information acquisition device | |
| CN109572842B (en) | Pole-climbing mechanism, pole-climbing intelligent inspection robot and pole-climbing method of transformer substation | |
| CN113057154A (en) | Greenhouse liquid medicine spraying robot | |
| CN107873510A (en) | A kind of automatic pollination machine of apple tree based on binocular vision technology | |
| CN102652510A (en) | Comprehensive spray control test stand | |
| CN109964905A (en) | Robot and its control method are administered to target based on walking certainly for fruit tree identification positioning | |
| CN206627117U (en) | A kind of space mapping system of combination space orientation scanner uni intelligent terminal | |
| CN207164045U (en) | Intelligent irrigation inspection car | |
| CN216982681U (en) | Strip matrix groove arbor cuttage robot | |
| CN106996770A (en) | The space mapping system and method for a kind of combination space orientation scanner uni intelligent terminal | |
| CN206459622U (en) | Gardens town road traffic survey tool |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170711 Termination date: 20201016 |