Intelligent weeding robot for vegetable planting

Abstract

The invention discloses an intelligent weeding robot for vegetable planting, which comprises a skeleton structure, a sliding table module mounted on the inner side of the skeleton structure, an electric push rod mounted through the sliding table module, a manipulator mounted at the telescopic end of the electric push rod, and an equipment supporting plate mounted at the top of the skeleton structure. A crushing device is installed at the position, close to the bottom, of the inner side of the framework structure, weeds grabbed by the mechanical arm are fed into the crushing device, driving motors capable of independently running are installed at the four corners of the bottom of the framework structure, driving wheels are installed at the output ends of the driving motors, and a control panel is installed at the front end of the framework structure; according to the equipment, independent moving devices can be arranged in three-dimensional directions to control the mechanical claws to move, the problem that the seedling damage rate is high due to irregular mechanical weeding can be solved, and intelligent and accurate weed removal is achieved.

Description

Technical Field

[0001] This invention relates to an intelligent weeding robot for vegetable cultivation. Background Technology

[0002] Agriculture occupies a vital position in my country's national industry. In 2016, agricultural land accounted for 56.21% of my country's total land area, covering approximately 5.28 million (5,277,330) square kilometers, and the gross agricultural product has been increasing year by year. However, the level of agricultural mechanization in my country is not high, reaching only about 47% by 2020. While my country's agricultural machinery market is large, the intelligence level of domestically produced agricultural machinery still needs improvement.

[0003] The most widely used method in current agricultural production is inter-row cultivation and weeding, which uses indiscriminate weeding machines. While efficient for large-area weeding, the working area and height are fixed and cannot be adjusted, making it difficult to adapt to complex terrain. It cannot remove weeds within wide-row planting areas, and even in narrow-row planting, existing weeding machines not only remove weeds between plants but also cause irregular mechanical scratching damage to the plants, resulting in a high rate of seedling damage. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent weeding robot for vegetable planting, which has independent moving devices in three dimensions to control the movement of the mechanical claw, thereby solving the problem of high seedling damage rate caused by irregular mechanical weeding and realizing intelligent and precise removal of weeds.

[0005] To solve the above problems, the present invention adopts the following technical solution:

[0006] A smart weeding robot for vegetable cultivation includes a skeleton structure, a sliding platform module installed on the inner side of the skeleton structure, an electric push rod installed on the sliding platform module, a robotic arm installed at the telescopic end of the electric push rod, an equipment tray installed on the top of the skeleton structure, a crushing device installed on the inner side of the skeleton structure near the bottom, weeds picked up by the robotic arm are fed into the crushing device, an independently operable drive motor is installed at each of the four corners of the bottom of the skeleton structure, a drive wheel is installed at the output end of the drive motor, and a control panel is installed at the front end of the skeleton structure.

[0007] Preferably, the skeleton structure includes a first rod, a second rod, a third rod, a fourth rod, a fifth rod, and a connector. The fifth rod is arranged in a rectangular array of four rods. The first rod is connected to two longitudinally adjacent fifth rods through the connector. The second rod is connected to two laterally adjacent fifth rods through the connector. The third rod is assembled into a U-shaped structure through the connector. The assembled U-shaped structure is installed between the front and rear second rods. The crushing device is installed through the third rod, which is installed between two longitudinally adjacent fifth rods. The sliding table module is also installed through the fifth rod.

[0008] Preferably, a first sunshade and a second sunshade are installed on the outside of the assembled frame structure, and multiple first sunshades and multiple second sunshades are provided.

[0009] Preferably, the slide module includes a first stepper motor and a second stepper motor. A transverse screw is mounted on the output end of the second stepper motor. A first connecting member mates with the fifth rod. The second stepper motor is fixedly mounted via the first connecting member. The transverse screw is rotatably mounted to the first connecting member. A transverse support rod is also fixedly mounted transversely between the two fifth rods. A fixing member mates with the fifth rod. Two or more transverse support rods are arranged in parallel. A transverse slide is fitted between the two or more transverse support rods. The transverse slide has a circular connector that mates with the transverse support rod. A nut seat is installed on the top of the transverse slide, and the transverse screw mates with the nut seat. The rotation of the transverse screw drives the nut seat to slide laterally. Two pulley seats are longitudinally arranged below the slide, and a belt is rotatably engaged between the two pulley seats. The first stepper motor is installed at one end of the pulley seat and drives the belt to rotate. A longitudinal support rod is installed between the two pulley seats, and a longitudinal slide is slidably engaged through the longitudinal support rod. The electric push rod is installed at the bottom of the longitudinal slide.

[0010] Preferably, a connecting plate is installed at the output end of the electric push rod, and a mini push rod is installed through the connecting plate. The robotic arm includes a square block, a robotic claw, a square pointed block, and a long rod. The square block is fixedly installed at the movable end of the mini push rod, the square pointed block is fixedly installed at the bottom of the mini push rod, and the long rod is rotatably connected between the square block and the square pointed block. The square block is also rotatably connected to the robotic claw. When the mini push rod extends, the square block moves downward, and the robotic claw rotates and opens along the rotational position relative to the square block. When the mini push rod retracts, the square block moves upward, and the robotic claw retracts.

[0011] When the robot reaches the designated position, it calculates the height of the weeds. The mini push rod extends outward, causing the four-sided block to move downward and the mechanical claw to open. The electric push rod then begins its work, pushing downward. After the electric push rod has extended a designated distance, the mechanical claw has penetrated the soil, enveloping the roots of the weeds and the surrounding soil. The mini push rod retracts, causing the mechanical claw to close, thus pulling out the entire weed by the roots, achieving the effect of "root and branch." After weeding is complete, the electric push rod retracts back to its initial height. During the weeding process, the principles of mechanical mechanics, specifically fixed-axis rotation and the law of energy conversion, are utilized. The axial driving force of the electric push rod is converted into the opening and closing and gripping force of the mechanical claw, and kinetic and mechanical energy are converted into the digging driving force of the claw to grasp the weeds and soil.

[0012] Preferably, the crushing device includes a housing, with a feeding port machined at the upper end of the housing. The housing is connected to the frame structure. A motor base is installed on the outside of the housing. A third stepper motor is installed on the top of the motor base. A crushing shaft is installed at the output end of the third stepper motor. The crushing shaft is inserted into the housing. Crushing blades are installed at the part of the crushing shaft that is inserted into the housing. A box wall blade body that matches the crushing blades is provided on the side wall of the feeding port.

[0013] After the actuator completes the task of picking up weeds and soil, the third stepper motor drives the crushing blades to start operating. The crushing blades on the crushing shaft rotate at high speed and impact the blades on the box wall, generating a strong deformation force. The weeds and soil that fall at the meshing point are completely crushed by the squeezing force, thus achieving the effect of crushing weeds and soil.

[0014] This project uses a YOLO v5-based weed recognition model to identify weed types and quantities and provide relative position information. A two-dimensional combination slide and a flat-bottomed push rod are used to move the weeding claw to a designated position for precise weed removal.

[0015] As the device travels along the field ridges, it uses a deep neural network based on SSD and AlexNet models to analyze and diagnose weeds. The control panel displays the detected weed types, quantities, and relative positions in real time. When the target value reaches the weeding threshold, the first and second stepper motors are controlled to move the mechanical claw directly above the target weeds. Combined with the raising and lowering of the electric push rod and the extension and retraction of the mini push rod, a weeding operation is completed. After the mechanical claw pulls out and gathers the weeds, it is moved via a slide to directly above the crushing device. The crushing device is activated, the mechanical claw opens, and the weeds and soil are thrown into the rolling crushing device for crushing. The crushed soil is then returned to the field ridges as fertilizer, preventing the weeds from taking root and growing again. This completes the entire weeding process.

[0016] By observing and analyzing the motion of the claws in a shopping mall claw machine, and combining this with the goal of weed removal, the robot was rationally designed and modified. Based on mechanical analysis and mathematical calculations, the relationship between the axial distance variation of the mechanical claw and its opening and closing was drafted, along with the corresponding claw size and design for weed removal. This robot can change different claw models according to the different weed removal needs of field crops.

[0017] The beneficial effects of this invention are:

[0018] This equipment is suitable for farmland in various environments. This intelligent weeding robot can achieve autonomous wheel drive and autonomous weed identification. Under the calculation of the control system, it can move in a plane by sliding along the X and Y axes, and move vertically with the help of the electric push rod. Through the above-mentioned movement, the robot arm is positioned above the weeds, the mini push rod is controlled to open the mechanical claw, the electric push rod is controlled to extend the gripper into the soil, and the mechanical claw is closed to grab the target weeds and some soil at the same time. Then the weeds and soil are sent to the crushing device to be crushed and recycled.

[0019] This device can intelligently grab weeds without damaging the planted plants. It has a high recognition rate, and the grabbed weeds are directly crushed and used as nutrients to supply the plants. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the structure of the present invention;

[0022] Figure 2 A 3D view of the slide module;

[0023] Figure 3 This is a schematic diagram of the robotic arm closing.

[0024] Figure 4 A schematic diagram of the robotic arm opening;

[0025] Figure 5 A perspective view of the crushing device;

[0026] Figure 6 A 3D view of the drive wheels;

[0027] Figure 7 This is a diagram of the convolutional neural network structure.

[0028] Figure 8This is a model framework based on deep learning neural networks;

[0029] Figure 9 This is the AlexNet model. Detailed Implementation

[0030] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0031] Any feature disclosed in this specification (including any appended claims, abstract, and drawings) may be replaced by other equivalent or similar features, unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is merely one example of a series of equivalent or similar features.

[0032] In the description of this invention, it should be understood that the terms "one end", "the other end", "outer side", "upper", "inner side", "horizontal", "coaxial", "center", "end", "length", "outer end", etc., 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 this invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0033] Furthermore, in the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0034] In this invention, unless otherwise explicitly specified and limited, the terms "set," "socket," "connect," "through," and "plug-in" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise explicitly limited. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0035] See Figures 1 to 9The illustrated intelligent weeding robot for vegetable cultivation includes a skeleton structure 888, a sliding table module 3 installed inside the skeleton structure 888, an electric push rod 5 installed through the sliding table module 3, a robotic arm 9 installed at the telescopic end of the electric push rod 5, an equipment tray 2 installed on the top of the skeleton structure 888, a crushing device 10 installed on the inner side of the skeleton structure 888 near the bottom, weeds picked up by the robotic arm 9 are fed into the crushing device 10, an independently operable drive motor 11 is installed at each of the four corners of the bottom of the skeleton structure 888, a drive wheel 12 is installed at the output end of the drive motor 11, and a control panel 15 is installed at the front end of the skeleton structure 888.

[0036] The skeleton structure 888 includes a first rod 7, a second rod 14, a third rod 19, a fourth rod 20, a fifth rod 21, and a connector 6. The fifth rod 21 is arranged in a rectangular array of four rods. The first rod 7 is connected to two longitudinally adjacent fifth rods 21 via the connector 6. The second rod 14 is connected to two laterally adjacent fifth rods 21 via the connector 6. The third rod 19 is assembled into a U-shaped structure via the connector 6. The assembled U-shaped structure is installed between the front and rear second rods 14. The crushing device 10 is installed via the third rod 19, which is installed between two longitudinally adjacent fifth rods 21. The sliding table module 3 is also installed via the fifth rod 21.

[0037] A first sunshade 1 and a second sunshade 13 are installed on the outside of the assembled frame structure, and multiple pieces of both the first sunshade 1 and the second sunshade 13 are provided.

[0038] The slide module 3 includes a first stepper motor 613 and a second stepper motor 621. A transverse screw 62 is mounted on the output end of the second stepper motor 621. A first connecting member 61 is connected between the transverse screw 62 and the fifth rod 21. The second stepper motor 621 is fixedly installed via the first connecting member 61. The transverse screw 62 is rotatably installed with the first connecting member 61. A transverse support rod 620 is also fixedly installed transversely between the two fifth rods 21. A fixing member 619 is connected between the transverse support rod 620 and the fifth rod 21. Two or more transverse support rods 620 are arranged in parallel. A transverse slide 66 is connected between the two or more transverse support rods 620. A circular connector 616 is provided to cooperate with the transverse support rod 620. A nut seat 65 is installed on the top of the transverse slide 66, and the transverse screw 62 cooperates with the nut seat 65. The rotation of the transverse screw 62 drives the nut seat 62 to slide laterally. Two pulley seats 618 are arranged longitudinally below the slide 66, and a belt 612 is rotatably engaged between the two pulley seats 618. The first stepper motor 613 is installed at one end of the pulley seat 618 and drives the belt 612 to rotate. A longitudinal support rod 617 is installed between the two pulley seats 618, and a longitudinal slide 615 is slidably engaged through the longitudinal support rod 617. The electric push rod 5 is installed at the bottom of the longitudinal slide 615.

[0039] A connecting plate 36 is installed at the output end of the electric push rod 5, and a mini push rod 38 is installed through the connecting plate 36. The robotic arm 9 includes a square block 311, a robotic claw 312, a square pointed block 315, and a long rod 316. The square block 311 is fixedly installed at the movable end of the mini push rod 38, the square pointed block 315 is fixedly installed at the bottom of the mini push rod 38, and the long rod 316 is rotatably connected between the square block 311 and the square pointed block 315. The square block 311 is also rotatably connected to the robotic claw 312. When the mini push rod 38 extends, the square block 311 moves downward, and the robotic claw 312 rotates and opens along the rotational position of the square block 311. When the mini push rod 38 retracts, the square block 311 moves upward, and the robotic claw 312 retracts.

[0040] When the robot reaches the designated position, it calculates the height of the weeds. The mini push rod extends outward, causing the four-sided block to move downward and the mechanical claw to open. The electric push rod then begins its work, pushing downward. After the electric push rod has extended a designated distance, the mechanical claw has penetrated the soil, enveloping the roots of the weeds and the surrounding soil. The mini push rod retracts, causing the mechanical claw to close, thus pulling out the entire weed by the roots, achieving the effect of "root and branch." After weeding is complete, the electric push rod retracts back to its initial height. During the weeding process, the principles of mechanical mechanics, specifically fixed-axis rotation and the law of energy conversion, are utilized. The axial driving force of the electric push rod is converted into the opening and closing and gripping force of the mechanical claw, and kinetic and mechanical energy are converted into the digging driving force of the claw to grasp the weeds and soil.

[0041] The crushing device 10 includes a housing 56, with a feeding port 18 machined at the upper end of the housing 56. The housing 56 is connected to the frame structure 888. A motor base 512 is installed on the outside of the housing 56. A third stepper motor 510 is installed on the top of the motor base 512. A crushing shaft 55 is installed at the output end of the third stepper motor 510. The crushing shaft 55 is inserted into the housing 56. A crushing blade 57 is installed at the part of the crushing shaft 55 that is inserted into the housing 56. A box wall blade body 515 that cooperates with the crushing blade 57 is provided on the side wall of the feeding port 18.

[0042] After the actuator completes the task of picking up weeds and soil, the third stepper motor drives the crushing blades to start operating. The crushing blades on the crushing shaft rotate at high speed and impact the blades on the box wall, generating a strong deformation force. The weeds and soil that fall at the meshing point are completely crushed by the squeezing force, thus achieving the effect of crushing weeds and soil.

[0043] This project uses a YOLO v5-based weed recognition model to identify weed types and quantities and provide relative position information. A two-dimensional combination slide and a flat-bottomed push rod are used to move the weeding claw to a designated position for precise weed removal.

[0044] As the device travels along the field ridges, it uses a deep neural network based on SSD and AlexNet models to analyze and diagnose weeds. The control panel displays the detected weed types, quantities, and relative positions in real time. When the target value reaches the weeding threshold, the first and second stepper motors are controlled to move the mechanical claw directly above the target weeds. Combined with the raising and lowering of the electric push rod and the extension and retraction of the mini push rod, a weeding operation is completed. After the mechanical claw pulls out and gathers the weeds, it is moved via a slide to directly above the crushing device. The crushing device is activated, the mechanical claw opens, and the weeds and soil are thrown into the rolling crushing device for crushing. The crushed soil is then returned to the field ridges as fertilizer, preventing the weeds from taking root and growing again. This completes the entire weeding process.

[0045] By observing and analyzing the motion of the claws in a shopping mall claw machine, and combining this with the goal of weed removal, the robot was rationally designed and modified. Based on mechanical analysis and mathematical calculations, the relationship between the axial distance variation of the mechanical claw and its opening and closing was drafted, along with the corresponding claw size and design for weed removal. This robot can change different claw models according to the different weed removal needs of field crops.

[0046] This project uses a YOLO v5-based weed recognition model to identify weed species and their locations. Data acquisition for computer deep learning involves taking 1000 high-resolution photos (including mixed and individual photos of weeds and plants) outdoors in the planting area using a camera or mobile phone. Different growth stages of the plants were considered during data collection, with seedlings at different stages photographed at weekly intervals. The acquired images underwent grayscale processing, replacing true-color images with a color space model. The most effective color space model was found and used for grayscale conversion of the color images to extract the color components of interest. Image segmentation and grayscale processing were performed while reducing image size. To reduce noise and obtain clearer images, preprocessing of the grayscale images was performed first to obtain images that are acceptable to the human eye and ensure recognition performance.

[0047] Convolutional Neural Networks (CNNs) are the most commonly used deep learning methods in computer vision. Their unique convolutional operations excel in image-related fields such as image classification, semantic segmentation, image retrieval, and image detection. Their structure is as follows: Figure 7 As shown. Taking an image as an example, the input information first enters the original image through the input layer, then the convolutional layer extracts and retains valuable features, and a non-linear layer is introduced to make it more closely resemble various non-curvilinear models. Finally, the information is fed into the fully connected layer to calculate the probability of each classification.

[0048] This project, based on SSD, designs a model framework based on deep learning neural networks, such as... Figure 8 As shown, the model consists of a VGG16 network and a DenseBlock part. The network extracts six object features, which are then input into the DenseBlock part. Receptive field structures are added to each layer to reduce computational steps, increase information extraction capabilities, and achieve lightweight design. A two-stage regression algorithm is introduced to regress candidate boxes, filtering out prior boxes identified as background during the first regression. A Feature Pyramid Network (FPN) fusion operation is also used for network detection, and finally, a Non-Maximum Suppression (NMS) structure is applied to obtain the final result.

[0049] This project uses the AlexNet model, whose structure is as follows: Figure 9 As shown, AlexNet uses the ReLU linear rectified activation function, which significantly reduces computation and improves convergence speed. It also augments the training data through methods such as horizontal image flipping, random cropping from the original image, translation transformations, and color and lighting modifications. AlexNet employs overlapping pooling, local normalization, and a dropout layer to prevent overfitting. Compared to traditional CNN convolutional neural networks, AlexNet uses full max pooling, avoiding the blurring effects of average pooling. Furthermore, AlexNet uses data augmentation, greatly reducing mild overfitting and improving generalization ability. AlexNet's step size is smaller than the pooling kernel size, enhancing the richness of image features.

[0050] This system uses a Raspberry Pi 4B embedded development board. The Raspberry Pi 4B's processor is a Broadcom BCM2711, a quad-core Cortex-A72 (ARM v8) 64-bit SoC with a clock speed of 1.5GHz. A physical image is shown below. Figure 1 As shown. This device is small in size, low in power consumption, and can run various parallel neural network models. It supports multiple deep learning frameworks such as TensorFlow and PyTorch. The system will use the prepared dataset and deep learning framework to train an object recognition model on a Raspberry Pi 4B for weed detection and recognition.

[0051] The primary purpose of this weed identification system based on embedded neural networks is to identify weeds in food crops. After implementation, the system can provide prediction results and corresponding performance metrics. This system design involves installing dependencies on a Raspberry Pi and running YOLOv5. To train YOLOv5 for weed identification, a dataset containing weed images has already been prepared. PyTorch7 is used as the training framework, which is concise, flexible, efficient, fast, and easy to learn, making it suitable for building deep learning networks.

[0052] For farmland where vegetables are grown, the natural environment varies greatly. Considering the angle, lighting, and occlusion of photos taken when identifying weeds, it is imperative to design a new weed identification system. Target detection was performed on the collected photos. Research on accuracy, speed, and counting accuracy revealed that this system has relatively stable speed and high recognition accuracy, meeting the design requirements of this project.

[0053] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent weeding robot for vegetable cultivation, characterized in that: The system includes a frame structure (888), a slide module (3) installed inside the frame structure (888), an electric push rod (5) installed through the slide module (3), a robot arm (9) installed at the telescopic end of the electric push rod (5), an equipment tray (2) installed on the top of the frame structure (888), a crushing device (10) installed on the inner side of the frame structure (888) near the bottom, weeds grabbed by the robot arm (9) are fed into the crushing device (10), an independently operable drive motor (11) is installed at each of the four corners of the bottom of the frame structure (888), a drive wheel (12) is installed at the output end of the drive motor (11), and a control panel (15) is installed at the front end of the frame structure (888).

2. The intelligent weeding robot for vegetable cultivation according to claim 1, characterized in that: The skeleton structure (888) includes a first rod (7), a second rod (14), a third rod (19), a fourth rod (20), a fifth rod (21), and a connector (6). The fifth rod (21) is arranged in a rectangular array of four rods. The first rod (7) is connected to two adjacent fifth rods (21) in the longitudinal direction through the connector (6). The second rod (14) is connected to two adjacent fifth rods (21) in the transverse direction through the connector (6). The third rod (19) is assembled into a U-shaped structure through the connector (6). The assembled U-shaped structure is installed between the second rods (14) on the front and rear sides. The crushing device (10) is installed through the third rod (19). The third rod (19) is installed between two adjacent fifth rods (21) in the longitudinal direction. The sliding module (3) is installed through the fifth rod (21).

3. The intelligent weeding robot for vegetable cultivation according to claim 2, characterized in that: A first sunshade (1) and a second sunshade (13) are installed on the outside of the assembled frame structure. Each of the first sunshade (1) and the second sunshade (13) has multiple pieces.

4. The intelligent weeding robot for vegetable cultivation according to claim 1, characterized in that: The slide module (3) includes a first stepper motor (613) and a second stepper motor (621). A transverse screw (62) is installed at the output end of the second stepper motor (621). A first connector (61) is fitted between the transverse screw (62) and the fifth rod (21). The second stepper motor (621) is fixedly installed through the first connector (61). The transverse screw (62) is rotatably installed with the first connector (61). A transverse support rod (620) is also fixedly installed transversely between the two fifth rods (21). A fixing member (619) is fitted between the transverse support rod (620) and the fifth rod (21). Two or more transverse support rods (620) are arranged in parallel. A transverse slide (66) is fitted between the two or more transverse support rods (620). A circular connector (616) is provided to cooperate with the transverse support rod (620). A nut seat (65) is installed on the top of the transverse slide (66). The transverse screw (62) cooperates with the nut seat (65). The rotation of the transverse screw (62) drives the nut seat (62) to slide laterally. Two pulley seats (618) are arranged longitudinally below the slide (66). A belt (612) is rotatably engaged between the two pulley seats (618). The first stepper motor (613) is installed at one end of the pulley seat (618) and drives the belt (612) to rotate. A longitudinal support rod (617) is installed between the two pulley seats (618). A longitudinal slide (615) is slidably engaged through the longitudinal support rod (617). The electric push rod (5) is installed at the bottom of the longitudinal slide (615).

5. The intelligent weeding robot for vegetable cultivation according to claim 4, characterized in that: A connecting plate (36) is installed at the output end of the electric push rod (5), and a mini push rod (38) is installed through the connecting plate (36). The robotic arm (9) includes a square block (311), a robotic gripper (312), a square pointed block (315), and a long rod (316). The square block (311) is fixedly installed at the movable end of the mini push rod (38), and the square pointed block (315) is fixedly installed at the bottom of the mini push rod (38). The rod (316) is rotatably connected between the square block (311) and the square pointed block (315). The square block (311) is also rotatably connected to the mechanical claw (312). When the mini push rod (38) extends, the square block (311) moves down. At this time, the mechanical claw (312) rotates and opens along the rotational position with the square block (311). When the mini push rod (38) retracts, the square block (311) moves up, and the mechanical claw (312) closes.

6. The intelligent weeding robot for vegetable cultivation according to claim 1, characterized in that: The crushing device (10) includes a housing (56), with a feeding port (18) machined at the upper end of the housing (56). The housing (56) is connected to the frame structure (888). A motor base (512) is installed on the outside of the housing (56). A third stepper motor (510) is installed on the top of the motor base (512). A crushing shaft (55) is installed at the output end of the third stepper motor (510). The crushing shaft (55) is inserted into the housing (56). A crushing blade (57) is installed at the part of the crushing shaft (55) that is inserted into the housing (56). A box wall blade body (515) that matches the crushing blade (57) is provided on the side wall of the feeding port (18).

Images