Intelligent fructus aurantii sorting machine based on five-connecting-rod mechanical arm and sorting method

Abstract

The invention discloses an intelligent fructus aurantii sorting machine based on a five-connecting-rod mechanical arm and a sorting method. The sorting machine comprises a collecting mechanism, a conveying mechanism, a supporting mechanism, a sorting mechanism, a camera support, a feeding device, a lower computer, an upper computer and a vacuum adsorption control system. The sorting mechanism comprises a vacuum adsorption sorting device, a five-connecting-rod mechanical arm and a sorting mechanism supporting base, double motors are adopted to cooperatively drive and control space positioning of an end effector of the mechanical arm, and high-precision target grabbing is achieved through angle difference compensation. The upper computer converts the target center position recognized by the visual recognition device into space coordinates of the tail end of the mechanical arm through a coordinate conversion algorithm, and the five-connecting-rod mechanical arm is driven to complete precise sorting operation. Stable adsorption of the fructus aurantii is completed by means of vacuum negative pressure, the fructus aurantii naturally falls off by directly cutting off a negative pressure air source, and intelligent grading and sorting operation of the quality of the fructus aurantii is achieved. The device is suitable for automatic sorting of the traditional Chinese medicinal material fructus aurantii and is high in sorting efficiency, accurate in positioning and high in adaptability.

Description

TECHNICAL FIELD

[0001] The present application relates to an intelligent fructus aurantii sorting machine based on a five-link mechanical arm and a sorting method, which belongs to the technical field of agricultural machinery. BACKGROUND

[0002] As a traditional Chinese medicinal material, the quality of fructus aurantii directly affects the efficacy. Traditional fructus aurantii sorting mainly relies on manual screening, which has problems such as low efficiency, high labor intensity, and unstable sorting precision. In recent years, with the development of machine vision and industrial automation technology, intelligent sorting equipment based on image recognition has been gradually applied in the field of agricultural product processing. However, existing sorting equipment is mostly designed for spherical or regular-shaped materials, while the shape of fructus aurantii is irregular and the surface texture is complex, making it difficult for traditional sorting machines to accurately identify and grasp.

[0003] At present, the intelligent sorting machines on the market mainly use conveyors combined with air blowing / air suction type sorting devices, such as the walnut sorting machine based on convolutional neural network proposed in patent CN111687069A. Although such equipment can achieve a certain degree of automation, it is limited by fixed sorting mechanisms and has poor adaptability to irregular-shaped materials, and the sorting process is easily disturbed by factors such as material stacking and shadow blocking. In addition, traditional mechanical arm sorting equipment mostly uses six-axis mechanical arms, which have complex structures and high costs, making it difficult to popularize in small and medium-sized medicinal material processing plants.

[0004] Therefore, it is of important application value to develop an intelligent sorting machine with low cost, high precision, and adaptability to the shape characteristics of fructus aurantii. SUMMARY

[0005] The present application is to solve the problems existing in the prior art, and provides an intelligent fructus aurantii sorting machine based on a five-link mechanical arm and a sorting method, which combines optimized mechanical structure design with machine learning algorithms to achieve efficient and accurate sorting of fructus aurantii, meeting the urgent needs of the pharmaceutical industry for automated sorting.

[0006] The application adopts the following technical scheme: an intelligent fructus aurantii sorting machine based on a five-link mechanical arm, comprising a collecting mechanism, a conveying mechanism, a supporting mechanism, a sorting mechanism, a camera support, a feeding device, a lower computer, an upper computer and a vacuum adsorption control system, the main frame of the supporting mechanism is vertically installed on a flat ground, the conveying mechanism is installed in the middle of the main frame of the supporting mechanism, the feeding device is installed on the supporting mechanism and above one end of the conveying mechanism, the collecting mechanism is located on the ground and below the other end of the conveying mechanism, the camera support is installed on the top crossbeam of the main frame of the supporting mechanism, the visual recognition device is installed on the camera support, the visual recognition device is vertically aligned with the center line of the conveying belt on the conveying mechanism, the visual recognition device, the vacuum adsorption control system, the conveying mechanism and the sorting mechanism are all in communication connection with the upper computer, the upper computer is provided with a data processing system and can receive dynamic images collected by the visual recognition device in real time and process data, the visual recognition device is connected with the upper computer through a network cable, and the proportional relationship between the virtual coordinate system and the actual coordinate system is calibrated through a calibration board grid, and the vacuum adsorption control system controls the sorting mechanism by receiving the instruction program sent by the upper computer.

[0007] Further, the sorting mechanism comprises a vacuum adsorption sorting device, a five-link mechanical arm and a sorting mechanism support base, a bearing mechanism is installed on one side of the transmission mechanism, the sorting mechanism support base is fixed on the upper surface of the bearing mechanism, the five-link mechanical arm is accurately docked with the support base through a positioning pin, and the vacuum adsorption sorting device is connected to the end of the five-link mechanical arm away from the sorting mechanism support base.

[0008] Further, the five-link mechanical arm comprises a motor, a first square aluminum plate, a first carbon rod, a second square aluminum plate, a second carbon rod, a third square aluminum plate, a third carbon rod, a fourth carbon rod, a fourth square aluminum plate, a first aluminum block, a second aluminum block, a third aluminum block and a fourth aluminum block, the two ends of the first carbon rod are respectively fixedly connected with a first aluminum block, the first square aluminum plate is symmetrically installed on the opposite sides of the first carbon rod and the first aluminum block, the first carbon rod, the first aluminum block and the first square aluminum plate together constitute a first-stage mechanical arm, the second square aluminum plate, the second carbon rod and the second aluminum block constituting a second-stage mechanical arm are installed in the same way as the first-stage mechanical arm, the third square aluminum plate, the third carbon rod and the third aluminum block constituting a third-stage mechanical arm are installed in the same way as the first-stage mechanical arm, and the fourth carbon rod, the fourth square aluminum plate and the fourth aluminum block constituting a fourth-stage mechanical arm are installed in the same way as the first-stage mechanical arm.

[0009] Further, the first aluminum block on the first stage mechanical arm away from the second stage mechanical arm and the fourth aluminum block on the fourth stage mechanical arm away from the third stage mechanical arm are respectively connected with a motor, the first aluminum block and the fourth aluminum block are fixedly connected with the output shaft of the motor and are directly driven by the motor, and the motor is installed on the sorting mechanism support base through a flange.

[0010] Further, the central axes of the adjacent first aluminum block and second aluminum block pass through a jam screw, the middle positions of the adjacent first aluminum block and second aluminum block are provided with a large bearing, and the ends of the first aluminum block and the second aluminum block away from the large bearing are respectively provided with a small bearing, and the adjacent first aluminum block, the second aluminum block, the large bearing and the small bearing are tightly fitted by tightening the jam screw.

[0011] Further, the vacuum adsorption sorting device comprises a cylinder, a first pneumatic joint, a suction cup sleeve, a vacuum suction cup and a second pneumatic joint (4-1-5), one end of the first pneumatic joint is connected with the gas path interface of the cylinder, the output end of the piston rod of the cylinder extends to below the suction cup sleeve after penetrating the suction cup sleeve and is connected with the suction cup sleeve, the vacuum suction cup is connected and installed on the suction cup sleeve, so that the vacuum suction cup and the suction cup sleeve can move up and down with the piston rod of the cylinder, one end of the second pneumatic joint is directly embedded in the internal passage of the vacuum suction cup, and the other end is connected with an external negative pressure gas source, so as to directly supply the actual working surface of the vacuum suction cup with the external gas source, form an independent negative pressure adsorption path, and adsorb objects.

[0012] Further, after the data processing system in the upper computer collects the image of the fructus aurantii material through the visual recognition device, sequentially performs image grayscale, Gaussian filtering, adaptive threshold binary processing, labels each fructus aurantii target area by using region connected domain segmentation theory to realize real-time positioning of the target, extracts color features, shape features, texture features and spectral features from each target area, constructs a feature vector and inputs the classification model to realize identification of the quality category of fructus aurantii.

[0013] Further, the real-time positioning of the target is realized by combining the region connected domain segmentation theory with the minimum circumscribed circle fitting, for each labeled fructus aurantii target area, the contour pixel points thereof are extracted, the least square method is used to fit the minimum circumscribed circle, the center coordinates of the minimum circumscribed circle are taken as the target positioning reference point, the set of contour pixel points of the target area is {(x k ,y k )|k=1,2,...,N}, N is the number of contour pixel points, and the center coordinates (x c ,y c ) of the minimum circumscribed circle are obtained by solving the following optimization problem:

[0014]

[0015] wherein r is the minimum circumscribed circle radius; the solved (x c , y c ) is the geometric center horizontal and vertical coordinates of the target in the industrial camera image coordinate system.

[0016] The application also adopts the following technical scheme: a sorting method of an intelligent fructus aurantii sorting machine based on a five-bar linkage mechanical arm, and the steps are as follows:

[0017] Step 1: collecting image information of the fructus aurantii on the conveying belt through a visual recognition device;

[0018] Step 2: processing the image, identifying the quality of the fructus aurantii and classifying them by using an upper computer;

[0019] Step 3: converting the identification result into a five-bar linkage mechanical arm motion coordinate;

[0020] Step 4: accurately grabbing the fructus aurantii according to the coordinate information by the five-bar linkage mechanical arm;

[0021] Step 5: placing the fructus aurantii into the corresponding collecting mechanism according to the quality classification.

[0022] Further, the specific steps are as follows:

[0023] Firstly, the fructus aurantii to be sorted are placed into a feeding device, the feeding device uniformly sends the fructus aurantii to be sorted into a conveying mechanism, the fructus aurantii moves to the shooting area of the visual recognition device with the conveying mechanism, the visual recognition device collects the image of the fructus aurantii according to the preset parameters, and the image is transmitted to the upper computer through a network line;

[0024] Then, the data processing system in the upper computer classifies and identifies the fructus aurantii in the image based on a target recognition algorithm, accurately determines the position of each fructus aurantii in the image by using a region connected domain segmentation theory positioning algorithm, and calculates the center coordinates;

[0025] Next, the upper computer converts the center coordinates from the image coordinate system through the world coordinate system into the end space coordinates of the five-bar linkage mechanical arm by using a coordinate conversion algorithm, after the coordinate conversion is completed, the upper computer transmits the coordinate information to a vacuum suction control system to realize the control of the sorting mechanism, after the sorting mechanism reaches the target position, the upper computer sends a motion instruction to the cylinder, the air inlet path of the cylinder is turned on, the air outlet path is closed, the piston rod of the cylinder extends downward and drives the vacuum suction disc to move downward synchronously until the vacuum suction disc adheres to the surface of the fructus aurantii;

[0026] Finally, the host computer controls the negative pressure gas source passage to be turned on, the external gas source is directly supplied to the actual working surface of the vacuum chuck through the second pneumatic connector, so that the vacuum chuck forms a negative pressure to adsorb the schisandra chinensis, after the vacuum chuck stably adsorbs the schisandra chinensis, the cylinder exhaust passage is turned on, the air inlet passage is closed, the piston rod is retracted upward, the schisandra chinensis is driven to move to the preset sorting area with the mechanical arm, after reaching the sorting area, the host computer controls the negative pressure gas source passage to be cut off, the negative pressure in the vacuum chuck disappears, the schisandra chinensis naturally falls off to the corresponding sorting station, the piston rod of the cylinder is reset again, and a single schisandra chinensis adsorption and sorting action is completed, and the next sorting instruction is waited to cycle.

[0027] The present application has the following beneficial effects:

[0028] (1) The five-bar linkage mechanical arm is adopted, the cost is reduced through optimization of the mechanical structure, and the sorting precision is ensured;

[0029] (2) The intelligent recognition of special-shaped traditional Chinese medicinal materials is realized by combining a machine learning algorithm and a machine vision technology;

[0030] (3) A special vacuum adsorption and sorting device is designed, and stable grabbing of schisandra chinensis of different sizes is ensured;

[0031] (4) The whole system adopts a modular design, and is convenient to maintain and upgrade. DETAILED DESCRIPTION OF DRAWINGS

[0032] Figure 1 It is a structure schematic view of an intelligent schisandra chinensis sorting machine based on a five-bar linkage mechanical arm.

[0033] Figure 2 It is a structure schematic view of a sorting mechanism.

[0034] Figure 3 It is a structure schematic view of a five-bar linkage mechanical arm.

[0035] Figure 4 It is a partial schematic view of Figure 3 .

[0036] Figure 5 It is a structure schematic view of a five-bar linkage mechanical arm connecting shaft.

[0037] Figure 6 It is a sectional view along line A-A of Figure 5 .

[0038] Figure 7 It is a structure schematic view of a vacuum adsorption and sorting device.

[0039] Figure 8 It is a sorting process schematic view.

[0040] In the figure: 1: collecting mechanism; 2: conveying mechanism; 3: supporting mechanism; 4: sorting mechanism; 5: industrial camera; 6: camera support; 7: feeding device; 4-1: vacuum adsorption sorting device; 4-2: five-link mechanical arm; 4-3: sorting mechanism support base; 4-1-1: air cylinder; 4-1-2: first pneumatic joint; 4-1-3: suction cup sleeve; 4-1-4: vacuum suction cup; 4-2-1: motor; 4-2-2: first square aluminum plate; 4-2-3: first carbon rod; 4-2-4: second square aluminum plate; 4-2-5: second carbon rod; 4-2-6: third square aluminum plate; 4-2-7: third carbon rod; 4-2-8: fourth carbon rod; 4-3-9: fourth square aluminum plate; 4-2-10: fourth aluminum block; 4-2-11: small bearing; 4-2-12: large bearing; 4-2-13: first aluminum block; 4-1-5: second pneumatic joint. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments.

[0042] In the description of the present application, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and therefore cannot be understood as indicating or implying that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.

[0043] The intelligent fructus aurantii sorting machine based on a five-link mechanical arm according to the present application comprises a collecting mechanism 1, a conveying mechanism 2, a supporting mechanism 3, a sorting mechanism 4, a camera support 6 and a feeding device 7. The main frame of the supporting mechanism 3 is vertically installed on a flat ground to form a stable frame structure. The conveying mechanism 2 is installed in the middle of the main frame of the supporting mechanism 3. The feeding device 7 is installed on the supporting mechanism 3 and above one end of the conveying mechanism 2. The gap between the discharge port of the feeding device 7 and the conveying mechanism 2 is 15 mm. The collecting mechanism 1 is located on the ground and below the other end of the conveying mechanism 2. The camera support 6 is installed on the top beam of the main frame of the supporting mechanism 3. The visual recognition device 5 (such as an industrial camera) is installed on the camera support 6, and the visual recognition device 5 is vertically aligned with the center line of the conveying belt of the conveying mechanism 2.

[0044] The sorting mechanism 4 includes a vacuum adsorption sorting device 4-1, a five-link mechanical arm 4-2, and a sorting mechanism support base 4-3, and a bearing mechanism is installed on one side of the transmission mechanism 2. The sorting mechanism support base 4-3 is fixed to the upper surface of the bearing mechanism by high-strength bolts, ensuring that the installation flatness error is ≤0.1 mm. The five-link mechanical arm 4-2 is precisely connected to the support base 4-3 using a positioning pin, thereby ensuring the installation coaxiality. The vacuum adsorption sorting device 4-1 is connected to the distal end of the five-link mechanical arm 4-2 away from the sorting mechanism support base 4-3 by bolts, and the initial position of the vacuum adsorption sorting device 4-1 is adjusted to be parallel to the conveying belt plane by adjusting the five-link mechanical arm 4-2.

[0045] The five-link mechanical arm 4-2 includes a motor 4-2-1, a first aluminum square plate 4-2-2, a first carbon rod 4-2-3, a second aluminum square plate 4-2-4, a second carbon rod 4-2-5, a third aluminum square plate 4-2-6, a third carbon rod 4-2-7, a fourth carbon rod 4-2-8, a fourth aluminum square plate 4-3-9, and a fourth aluminum block 4-2-10. The motor 4-2-1 is installed on the sorting mechanism support base 4-3 by a flange, ensuring that the output shaft perpendicularity error is ≤0.05 mm. The two distal ends of the first carbon rod 4-2-3 are respectively fixedly connected to a first aluminum block 4-2-13, and a first aluminum square plate 4-2-2 is installed on the opposite sides of the installed first carbon rod 4-2-3 and the first aluminum block 4-2-13. The installed first carbon rod 4-2-3, the first aluminum block 4-2-13, and the first aluminum square plate 4-2-2 together form a first-stage mechanical arm, wherein the first carbon rod 4-2-3 functions as a bending moment to ensure the load stability of the first-stage mechanical arm structure, and the first aluminum square plate 4-2-2 is symmetrically installed on both sides of the two first aluminum blocks 4-2-13 to form a rigid support structure of the arm body. Similarly, the second aluminum square plate 4-2-4, the second carbon rod 4-2-5, and the second aluminum block (not shown) of the second-stage mechanical arm are installed in the same way as the first-stage mechanical arm. The third aluminum square plate 4-2-6, the third carbon rod 4-2-7, and the third aluminum block (not shown) of the third-stage mechanical arm are installed in the same way as the first-stage mechanical arm. The fourth carbon rod 4-2-8, the fourth aluminum square plate 4-3-9, and the fourth aluminum block 4-2-10 of the fourth-stage mechanical arm are installed in the same way as the first-stage mechanical arm. The first aluminum block 4-2-13 away from the second-stage mechanical arm on the first-stage mechanical arm and the fourth aluminum block 4-2-10 away from the third-stage mechanical arm on the fourth-stage mechanical arm are respectively connected to the motor 4-2-1, and the first aluminum block 4-2-13 and the fourth aluminum block 4-2-10 are fixedly connected to the output shaft of the motor 4-2-1, which is directly driven by the motor 4-2-1.

[0046] The first aluminum block 4-2-13 and the second aluminum block are connected between the first mechanical arm and the second mechanical arm, specifically, when connected, a plug screw passes through the center of the first aluminum block 4-2-13 and the second aluminum block in the axial direction, a large bearing 4-2-12 is arranged at the middle position of the first aluminum block 4-2-13 and the second aluminum block, and a small bearing 4-2-11 is arranged at the end of the first aluminum block 4-2-13 and the second aluminum block away from the large bearing 4-2-12. By tightening the plug screw, the first aluminum block 4-2-13, the second aluminum block, the large bearing 4-2-12 and the small bearing 4-2-11 are tightly fitted, which not only realizes the fixing and limiting of the adjacent mechanical arms, but also guarantees the flexibility of the movement of the mechanical arms through the rotation cooperation of the large bearing 4-2-12 and the small bearing 4-2-11. By analogy, the second mechanical arm and the third mechanical arm, and the third mechanical arm and the fourth mechanical arm are connected in the above-mentioned manner, so that each arm of the five-bar linkage mechanical arm can realize smooth rotary motion under the support and rotation of the corresponding bearing, thereby completing various complex actions and operations.

[0047] The vacuum adsorption sorting device 4-1 includes a cylinder 4-1-1, a first pneumatic connector 4-1-2, a suction cup sleeve 4-1-3, a vacuum suction cup 4-1-4 and a second pneumatic connector 4-1-5. The cylinder 4-1-1 serves as a power execution element and provides driving force for the entire vacuum adsorption sorting action. The first pneumatic connector 4-1-2 serves as an airflow guide, one end of which is connected to the gas path interface of the cylinder 4-1-1 to ensure the sealing and reliability of the gas path connection. The piston rod output end of the cylinder 4-1-1 extends to below the suction cup sleeve 4-1-3 after penetrating through the suction cup sleeve 4-1-3 and is connected to the suction cup sleeve 4-1-3. The vacuum suction cup 4-1-4 is installed on the suction cup sleeve 4-1-3 by bolt connection, so that the vacuum suction cup 4-1-4 and the suction cup sleeve 4-1-3 can move up and down with the piston rod of the cylinder 4-1-1. One end of the second pneumatic connector 4-1-5 is directly embedded in the internal passage of the vacuum suction cup 4-1-4, and the other end is connected to an external negative pressure gas source, which can directly supply the actual working surface of the vacuum suction cup with an external gas source, form an independent negative pressure adsorption path, and thereby adsorb objects and realize the sorting function.

[0048] The sorting process is as follows: the five-bar linkage mechanical arm 4-2 moves to the upper side of the fructus aurantii, the upper computer controls the cylinder 4-1-1 to intake air to make the piston rod move downward, the vacuum suction cup 4-1-4 is attached to the fructus aurantii, the negative pressure gas source is turned on to adsorb the fructus aurantii, then the cylinder is exhausted to make the piston rod retract, the five-bar linkage mechanical arm 4-2 drives the fructus aurantii to the sorting area, the negative pressure is cut off, the fructus aurantii falls off, and the cylinder is reset to wait for the next instruction.

[0049] The intelligent fructus aurantii sorting machine based on the five-link mechanical arm further comprises a lower computer, an upper computer and a vacuum adsorption control system. The visual recognition device 5 is located above the sorting mechanism 4 and is used for real-time acquisition of fructus aurantii material images. The upper computer carries a data processing program and can receive dynamic images collected by the visual recognition device 5 in real time and process data. The visual recognition device 5 is connected to the upper computer through a network cable, and the proportional relationship between the virtual coordinate system and the actual coordinate system is calibrated through a calibration board grid to realize accurate mapping of the material position.

[0050] The data processing system in the upper computer is constructed by a target recognition algorithm and a regional connected component segmentation theory positioning algorithm. The vacuum adsorption control system receives the instruction program sent by the upper computer through the lower computer to control the sorting mechanism 4. The visual recognition device 5, the vacuum adsorption control system, the conveying mechanism 2 and the sorting mechanism 4 are all connected to the upper computer to realize data interaction and collaborative control through a preset communication protocol, forming a closed-loop integrated workflow. The lower computer includes a servo driver and a relay IO module integrated in an integral joint servo motor. The servo driver receives the motion control instruction sent by the upper computer and drives the five-link mechanical arm 4-2 to complete the accurate action, and the relay IO module receives the pneumatic control signal of the upper computer and drives the vacuum chuck and other pneumatic actuators to complete the corresponding work. The two ensure the synchronization and accuracy of the movement of the five-link mechanical arm 4-2 and the pneumatic execution action through real-time response to the control instruction.

[0051] After the data processing system collects the image of the fructus aurantii material through the visual recognition device 5, it sequentially performs image grayscale, Gaussian filtering, adaptive threshold binary processing, labels each fructus aurantii target area using the regional connected component segmentation theory to realize real-time positioning of the target, extracts color features (such as RGB mean, saturation, etc.), shape features (such as contour perimeter, area, etc.), texture features (such as gray level co-occurrence matrix features, etc.), and spectral features from each target area, constructs a feature vector and inputs it into a classification model to realize identification of the quality category of fructus aurantii, including image preprocessing, region segmentation, feature extraction and category determination.

[0052] Real-time target positioning is realized by the regional connected component segmentation theory combined with the minimum circumscribed circle fitting. For the fructus aurantii image that has been preprocessed and labeled with regional connected components, for each labeled fructus aurantii target area, the contour pixel points are extracted, the least square method is used to fit the minimum circumscribed circle, and the center coordinates of the minimum circumscribed circle are taken as the target positioning reference point. Let the set of contour pixel points of the target area be {(x k , y k )|k=1,2,...,N} (N is the number of contour pixel points), and the center coordinates (x c , y c ) of the minimum circumscribed circle are obtained by solving the following optimization problem:

[0053]

[0054] Where r is the minimum circumcircle radius; the solution (x c ,y c ) represents the horizontal and vertical coordinates of the geometric center of the target in the industrial camera image coordinate system.

[0055] The transformation between the image coordinate system of the visual recognition device 5 and the base coordinate system of the five-bar linkage robotic arm 4-2 adopts an "eye outside the hand" calibration method (visual recognition device 5 is fixed, and the five-bar linkage robotic arm 4-2 moves independently), thereby establishing the transformation relationship between the two. The coordinates of the center of the smallest circumscribed circle of the target in the image coordinate system of the visual recognition device 5 (x...) are then used to establish the transformation relationship between the two. c y c When converting the coordinates of the five-bar linkage robot to (X, Y) coordinates in the working plane, the coordinate transformation is performed using the transformation matrix M (a 3×3 homogeneous transformation matrix containing in-plane rotation and translation parameters) obtained through hand-eye calibration. The transformation formula is as follows:

[0056]

[0057] Since the five-bar linkage 4-2 moves at a fixed height with no displacement in the Z direction, the resulting (X, Y) coordinates are used as the planar position parameters for the five-bar linkage 4-2 to grasp the target. The host computer combines these planar coordinates with the kinematic model of the five-bar linkage 4-2 to convert them into motion parameter commands such as the angles of each joint of the five-bar linkage 4-2, which are then sent to the slave computer to drive the five-bar linkage 4-2 to complete precise grasping and achieve efficient sorting of trifoliate orange peels.

[0058] The present invention relates to a sorting method for an intelligent trifoliate orange (Citrus aurantium) sorting machine based on a five-bar linkage robotic arm, as detailed below:

[0059] Step 1: Collect image information of the trifoliate orange peel on the conveyor belt using the visual recognition device 5;

[0060] Step 2: Use a host computer to process the images, identify the quality of the trifoliate orange peel, and classify them.

[0061] Step 3: Convert the recognition results into 4-2 motion coordinates for the five-bar linkage robotic arm;

[0062] Step 4: The five-bar linkage robotic arm 4-2 precisely grasps the trifoliate orange peel according to the coordinate information;

[0063] Step 5: Place the bitter orange peels into the corresponding collection devices according to their quality classification.

[0064] The specific steps described above are explained below:

[0065] First, the orange peels to be sorted are placed into the feeding device 7, which evenly feeds them into the conveying mechanism 2. The orange peels move with the conveying mechanism 2 to the imaging area of ​​the visual recognition device 5. At this time, the visual recognition device 5 acquires images of the orange peels according to preset parameters and transmits them to the host computer via a network cable. Then, the data processing system in the host computer classifies and identifies the orange peels in the image based on a target recognition algorithm, judging their variety, size, color, and other characteristics. At the same time, using the region connected component segmentation theory positioning algorithm, the position of each orange peel in the image is accurately determined, and the center coordinates are calculated. Next, the host computer uses a coordinate transformation algorithm to convert the center coordinates from the image coordinate system to the world coordinate system and then to the end space coordinates of the five-bar linkage robotic arm 4-2. This process takes into account factors such as the installation of the visual recognition device 5 and the initial position of the five-bar linkage robotic arm 4-2. After the coordinate transformation is completed, the host computer transmits the coordinate information to the vacuum adsorption control system to realize the control of the sorting mechanism 4. The vacuum adsorption control system employs a dual-motor coordinated drive for a five-bar linkage robotic arm 4-2. By precisely controlling the motor speed and direction, combined with an angle difference compensation algorithm, the joints of the five-bar linkage robotic arm 4-2 move in tandem, ensuring the end effector accurately reaches the target location of the bitter orange peel. Upon reaching the target location, the host computer sends an action command to the relay I / O module controlling the cylinder, opening the cylinder's air intake path and closing the exhaust path. The piston rod extends downwards, simultaneously moving the vacuum suction cup downwards until it adheres to the bitter orange peel surface. Next, the host computer controls the negative pressure air source path to open via the I / O module. An external air source is directly supplied to the actual working surface of the vacuum suction cup through the second pneumatic connector 4-1-5, creating negative pressure to adsorb the bitter orange peel. Once the vacuum suction cup has stably adsorbed the bitter orange peel, the cylinder's exhaust path is opened and the air intake path is closed. The piston rod retracts upwards, moving the bitter orange peel with the robotic arm to the preset sorting area. Upon reaching the sorting area, the host computer controls the negative pressure air supply to be cut off, the negative pressure in the vacuum suction cup disappears, and the bitter orange peel naturally falls to the corresponding sorting station. The cylinder piston rod resets again, completing a single bitter orange peel adsorption and sorting action, and then waits for the next sorting instruction to carry out the cycle.

[0066] This invention relates to an intelligent trifoliate orange peel sorting machine based on a five-bar linkage robotic arm. By combining visual recognition with high-precision robotic arm control, it achieves intelligent and automatic sorting of trifoliate orange peel, featuring accurate positioning, high sorting efficiency, and strong adaptability.

[0067] The above description is only a preferred embodiment of the present invention. It should be noted that those skilled in the art can make several improvements without departing from the principle of the present invention, and these improvements should also be considered within the scope of protection of the present invention.

Claims

1. A five-bar linkage based intelligent Shicheng sorting machine, characterized in that: The application relates to a sorting device for sorting objects, which comprises a collecting mechanism (1), a conveying mechanism (2), a supporting mechanism (3), a sorting mechanism (4), a camera support (6), a feeding device (7), a lower computer, an upper computer and a vacuum adsorption control system, the main frame of the supporting mechanism (3) is vertically installed on a flat ground, the conveying mechanism (2) is installed in the middle of the main frame of the supporting mechanism (3), the feeding device (7) is installed on the supporting mechanism (3) and above one end of the conveying mechanism (2), the collecting mechanism (1) is located on the ground and below the other end of the conveying mechanism (2), the camera support (6) is installed on the top beam of the main frame of the supporting mechanism (3), the visual identification device (5) is installed on the camera support (6), the visual identification device (5) is vertically aligned with the center line of the conveying belt of the conveying mechanism (2), the visual identification device (5), the vacuum adsorption control system, the conveying mechanism (2) and the sorting mechanism (4) are in communication connection with the upper computer, the upper computer is provided with a data processing system and can receive dynamic images collected by the visual identification device (5) in real time and process data, the visual identification device (5) is connected with the upper computer through a network cable, the proportional relationship between a virtual coordinate system and an actual coordinate system is calibrated through a calibration plate grid, and the vacuum adsorption control system receives an instruction program sent by the upper computer through the lower computer, so as to control the sorting mechanism (4).

2. The five-bar linkage based intelligent Shicheng sorting machine according to claim 1, characterized in that: The sorting mechanism (4) comprises a vacuum adsorption sorting device (4-1), a five-link mechanical arm (4-2) and a sorting mechanism supporting base (4-3), a bearing mechanism is installed on one side of the conveying mechanism (2), the sorting mechanism supporting base (4-3) is fixed on the upper surface of the bearing mechanism, the five-link mechanical arm (4-2) is accurately connected with the supporting base (4-3) through a positioning pin, and the vacuum adsorption sorting device (4-1) is connected to the end of the five-link mechanical arm (4-2) away from the sorting mechanism supporting base (4-3).

3. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 2, characterized in that: The five-link mechanical arm (4-2) comprises a motor (4-2-1), a first aluminum plate (4-2-2), a first carbon rod (4-2-3), a second aluminum plate (4-2-4), a second carbon rod (4-2-5), a third aluminum plate (4-2-6), a third carbon rod (4-2-7), a fourth carbon rod (4-2-8), a fourth aluminum plate (4-3-9), a first aluminum block (4-2-13), a second aluminum block, a third aluminum block and a fourth aluminum block (4-2-10), two ends of the first carbon rod (4-2-3) are fixedly connected with a first aluminum block (4-2-13) respectively, first aluminum plates (4-2-2) are symmetrically installed on the opposite sides of the installed first carbon rod (4-2-3) and first aluminum block (4-2-13) respectively, the installed first carbon rod (4-2-3), first aluminum block (4-2-13) and first aluminum plate (4-2-2) jointly constitute a first-stage mechanical arm, the installation mode of the second aluminum plate (4-2-4), second carbon rod (4-2-5) and second aluminum block constituting a second-stage mechanical arm is the same as that of the first-stage mechanical arm, the installation mode of the third aluminum plate (4-2-6), third carbon rod (4-2-7) and third aluminum block constituting a third-stage mechanical arm is the same as that of the first-stage mechanical arm, and the installation mode of the fourth carbon rod (4-2-8), fourth aluminum plate (4-3-9) and fourth aluminum block (4-2-10) constituting a fourth-stage mechanical arm is the same as that of the first-stage mechanical arm.

4. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 3, characterized in that: The first aluminum block (4-2-13) away from the second-stage mechanical arm on the first-stage mechanical arm and the fourth aluminum block (4-2-10) away from the third-stage mechanical arm on the fourth-stage mechanical arm are respectively connected with a motor (4-2-1), the first aluminum block (4-2-13) and fourth aluminum block (4-2-10) are fixedly connected with the output shaft of the motor (4-2-1) and are directly driven by the motor (4-2-1), and the motor (4-2-1) is installed on the sorting mechanism support base (4-3) through a flange.

5. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 4, characterized in that: The central axes of the adjacent first aluminum block (4-2-13) and second aluminum block pass through a jam screw, a large bearing (4-2-12) is arranged at the middle position of the adjacent first aluminum block (4-2-13) and second aluminum block, and a small bearing (4-2-11) is arranged at the end of the first aluminum block (4-2-13) and second aluminum block away from the large bearing (4-2-12), and the adjacent first aluminum block (4-2-13), second aluminum block, large bearing (4-2-12) and small bearing (4-2-11) are tightly attached by tightening the jam screw.

6. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 1, characterized in that: The vacuum adsorption sorting device (4-1) comprises a cylinder (4-1-1), a first pneumatic joint (4-1-2), a suction cup sleeve (4-1-3), a vacuum suction cup (4-1-4) and a second pneumatic joint (4-1-5), one end of the first pneumatic joint (4-1-2) is connected with the gas path interface of the cylinder (4-1-1), the piston rod output end of the cylinder (4-1-1) extends to below the suction cup sleeve (4-1-3) after penetrating through the suction cup sleeve (4-1-3) and is connected with the suction cup sleeve (4-1-3), the vacuum suction cup (4-1-4) is connected and installed on the suction cup sleeve (4-1-3), so that the vacuum suction cup (4-1-4) and the suction cup sleeve (4-1-3) can move up and down with the piston rod of the cylinder (4-1-1), one end of the second pneumatic joint (4-1-5) is directly embedded into the internal passage of the vacuum suction cup (4-1-4), and the other end accesses an external negative pressure gas source, so as to directly supply the actual working surface of the vacuum suction cup with the external gas source, form an independent negative pressure adsorption path, and adsorb objects.

7. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 1, characterized in that: After the data processing system in the upper computer collects the image of the fructus aurantii material through the visual recognition device (5), it sequentially performs image grayscale, Gaussian filtering, adaptive threshold binary processing, marks each fructus aurantii target area to realize real-time positioning of the target by using region connected domain segmentation theory, extracts color features, shape features, texture features and spectral features from each target area, constructs a feature vector and inputs a classification model to realize identification of the quality category of fructus aurantii.

8. The five-bar linkage based intelligent Shanzhuya sorting machine according to claim 7, characterized in that: The target real-time positioning is realized by combining the region connected domain segmentation theory with the minimum circumscribed circle fitting. For the pre-processed and region connected domain labeled citron fruit image, the contour pixel points of each labeled target region are extracted, and the minimum circumscribed circle is fitted by using the least square method. The center coordinates of the minimum circumscribed circle are taken as the target positioning reference point. The set of contour pixel points of the target region is {(x k ,y k )|k=1,2,...,N}, N is the number of contour pixel points, and the center coordinates (x c ,y c ) of the minimum circumscribed circle are obtained by solving the following optimization problem: wherein r is the minimum circumscribed circle radius; the solved (x c ,y c ) is the geometric center horizontal and vertical coordinates of the target in the industrial camera image coordinate system.

9. A sorting method of an intelligent fructus aurantii sorting machine based on a five-bar linkage arm, characterized in that: The steps are as follows: Step 1: Collect the image information of fructus aurantii on the conveying belt through the visual recognition device (5); Step 2: Process the image using the upper computer to identify the quality of fructus aurantii and classify it; Step 3: Convert the recognition result into the motion coordinates of the five-link mechanical arm (4-2); Step 4: The five-link mechanical arm (4-2) accurately grasps the fructus aurantii according to the coordinate information; Step 5: Place the fructus aurantii into the corresponding collection mechanism according to the quality classification.

10. The method for sorting of the intelligent fructus aurantii separating machine based on five-bar linkage according to claim 9, characterized in that: The specific steps are as follows: Firstly, the fructus aurantii to be sorted is placed in the feeding device (7), the feeding device (7) uniformly feeds the fructus aurantii to be sorted into the conveying mechanism (2), the fructus aurantii moves to the shooting area of the visual recognition device (5) along with the conveying mechanism (2), the visual recognition device (5) collects the image of the fructus aurantii according to the preset parameters and transmits it to the upper computer through the network line; Then, the data processing system in the upper computer identifies and classifies the fructus aurantii in the image based on a target recognition algorithm, accurately determines the position of each fructus aurantii in the image by using a region connected domain segmentation positioning algorithm, and calculates the center coordinates; Next, the upper computer converts the center coordinates from the image coordinate system through the world coordinate system into the spatial coordinates of the end of the five-link mechanical arm (4-2) by using a coordinate conversion algorithm, after the coordinate conversion is completed, the upper computer transmits the coordinate information to the vacuum adsorption control system to realize the control of the sorting mechanism (4), after the sorting mechanism (4) reaches the target position, the upper computer sends a motion instruction to the cylinder, the cylinder inlet path is turned on and the cylinder exhaust path is closed, the piston rod of the cylinder extends downward and drives the vacuum suction cup to move downward synchronously until the vacuum suction cup adheres to the surface of the fructus aurantii. Finally, the host computer controls the negative pressure air source passage to be turned on, and the external air source is directly supplied to the actual working surface of the vacuum chuck through the second pneumatic joint (4-1-5), so that the vacuum chuck forms a negative pressure to adsorb the shaddock. After the vacuum chuck stably adsorbs the shaddock, the air cylinder exhaust path is turned on and the air inlet path is closed, the piston rod is retracted upward, driving the shaddock to move to the preset sorting area with the mechanical arm. When reaching the sorting area, the host computer controls the negative pressure air source passage to be cut off, the negative pressure in the vacuum chuck disappears, and the shaddock naturally falls off to the corresponding sorting station. The piston rod of the air cylinder is reset again, and a single shaddock adsorption and sorting action is completed. The next sorting instruction is waited for to circulate the work.

Images