Multi-layer three-dimensional cultivation automatic transplanting robot for plant factory

Abstract

The invention relates to the technical field of agricultural automation, in particular to a multi-layer three-dimensional cultivation automatic transplanting robot for a plant factory, the multi-layer three-dimensional cultivation automatic transplanting robot comprises a cultivation frame, a seedling raising plate is arranged on the side face of the cultivation frame, and the robot comprises a control module, a visual module, a positioning module, a moving module and a clamping module. Comprising a clamping unit and a clamping seat, the clamping seat is fixedly connected with a moving seat, a plurality of lifting units are arranged below the clamping seat, the clamping unit comprises a clamping block detachably connected with the lifting units, a clamping motor is arranged in the clamping block, a plurality of clamping grooves are formed in the lower end face of the clamping block, and clamping jaws are movably arranged in the clamping grooves. A crank is further arranged in the clamping block, the two ends of the crank are connected with the output end of the clamping motor and the clamping jaw respectively, opening and closing actions of the clamping jaw are achieved, by means of the design, the robot can accurately clamp and release the field planting basket, and the transplanting accuracy and consistency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of agricultural automation, and in particular to a multi-layer stereoscopic cultivation automated transplanting robot used in a plant factory. Background Art

[0002] In plant factories, the multi-layer three-dimensional cultivation mode can effectively improve space utilization and yield. Because the cultivation layer is high, the transplanting process in this mode, if done manually, will have greater safety hazards, low efficiency, high labor intensity, and it is difficult to ensure the accuracy and consistency of transplanting.

[0003] The existing automated transplanting equipment is not suitable for multi-layer three-dimensional cultivation mode. It is difficult to accurately locate the planting holes during the transplanting process, which is inefficient. Therefore, an automated transplanting robot is needed to improve production efficiency and quality. Summary of the invention

[0004] The present invention aims to provide a multi-layer stereoscopic cultivation automated transplanting robot for a plant factory, which is used to improve the transplanting speed and the accuracy and consistency of transplanting.

[0005] The present invention is achieved through the following technical solutions: A multi-layer stereoscopic cultivation automated transplanting robot for a plant factory comprises a cultivation frame, a seedling raising plate is arranged on the side of the cultivation frame, the robot comprises a control module, a visual module, a positioning module, a moving module and a clamping module, the moving module comprises a moving seat, a vertical frame and a horizontal frame, the horizontal frame is slidably arranged on the vertical frame and is connected to the control module signal, the moving seat is slidably arranged on the horizontal frame and is connected to the control module signal, the visual module is connected to the control module signal and is used for position information collection, the positioning module is connected to the control module signal and is used for position information determination, the clamping module is arranged on the moving seat, and clamps and fixes a planting basket in the seedling raising plate through the control module; wherein the process of determining the position information is as follows: the seedling position is identified by the visual module and scanned to obtain 3D point cloud data, the 3D point cloud data is processed to extract the position information and shape information of the planting basket and the seedling, and a coordinate system is established, and finally the boundary position information of the planting basket is transmitted to the control module.

[0006] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. The present invention integrates a clamping module, including a clamping unit and a clamping seat, the clamping seat is fixedly connected to the moving seat, a plurality of lifting units are arranged below the clamping seat, the clamping unit includes a clamping block detachably connected to the lifting unit, a clamping motor is arranged in the clamping block, a plurality of clamping grooves are opened on the lower end surface of the clamping block, a clamping claw is movably arranged in the clamping groove, a crank is also arranged in the clamping block, and the two ends of the crank are respectively connected to the output end and the clamping claw of the clamping motor, so as to realize the opening and closing action of the clamping claw. This design enables the robot to accurately clamp and release the planting basket, thereby improving the accuracy and consistency of transplanting; 2. A crank is also provided in the clamping block of the present invention, and the two ends of the crank are respectively connected to the output end of the clamping motor and the clamping claw. When the clamping motor is started, the opening and closing action of the clamping claw is realized through the synchronous rotation of the crank. Such a mechanical transmission design ensures the accuracy and reliability of the clamping action. In order to protect the seedlings from damage, force sensors are provided on the inner side of the clamping claw, and these force sensors can monitor the force of the clamping claw in the grasping process in real time. 3. The introduction of the state judgment model of the present invention enables the automated transplanting robot to evaluate the growth status of seedlings, which is of guiding significance for deciding whether to perform transplanting operations. By obtaining the contour information of the seedling branches and stems through OPENCV and combining the range of growth status information, the robot can judge whether the growth status of the seedlings is good, thereby avoiding transplanting seedlings with poor growth status and ensuring that the transplanted seedlings have good growth potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation of the embodiments of the present invention. In the drawings: Figure 1 It is a schematic diagram of the structure of the robot of the present invention; Figure 2 for Figure 1 A is a schematic diagram of the enlarged structure of the middle part; Figure 3 It is a schematic diagram of part of the internal structure of the clamping module; Figure 4 is a schematic diagram of the internal structure of the clamping unit; Figure 5 It is a schematic diagram of an axial structure of the clamping unit of the present invention; Figure 6 Another axial structural schematic diagram of the clamping unit of the present invention; Figure 7 It is a structural schematic diagram of the steering unit of the present invention.

[0008] Marks and corresponding parts names in the attached drawings: 1- cultivation rack, 2- seedling plate, 3- mobile module, 4- clamping module, 5- planting basket; 31-mobile seat, 32-vertical frame, 33-horizontal frame; 41-clamping unit, 42-clamping seat, 43-lifting unit, 44-steering rod, 45-steering motor; 411-clamping block, 412-clamping motor, 413-clamping slot, 414-clamping claw, 415-crank, 416-clamping cover, 417-opening; 431-screw. DETAILED DESCRIPTION

[0009] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and drawings. The schematic implementation modes and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention. It should be noted that the present invention is already in the actual development and use stage.

[0010] Embodiment 1: Please refer to the attached Figures 1 to 6 A multi-layer stereoscopic cultivation automated transplanting robot for a plant factory comprises a cultivation frame 1, a seedling raising board 2 is arranged on the side of the cultivation frame 1, and is characterized in that: the robot comprises a control module, a visual module, a positioning module, a moving module 3 and a clamping module 4, the moving module 3 comprises a moving seat 31, a vertical frame 32 and a horizontal frame 33, the horizontal frame 33 is slidably arranged on the vertical frame 32 and is connected to the control module signal, the moving seat 31 is slidably arranged on the horizontal frame 33 and is connected to the control module signal, the visual module is connected to the control module signal and is used for position information collection, the positioning module is connected to the control module signal and is used for position information determination, the clamping module 4 is arranged on the moving seat 31, and clamps and fixes the planting basket 5 in the seedling raising board 2 through the control module; wherein, the process of determining the position information is: identifying the seedling position through the visual module to scan and obtain 3D point cloud data, then processing the 3D point cloud data to extract the position information and shape information of the planting basket 5 and the seedling, and establishing a coordinate system, and finally transmitting the boundary position information of the planting basket 5 to the control module.

[0011] It should be noted that in plant factories, the multi-layer stereoscopic cultivation mode can effectively improve space utilization and yield. Because the cultivation layer is high, the transplanting process in this mode has greater safety hazards, low efficiency, high labor intensity and other problems compared to traditional transplanting if manual transplanting is used, and it is difficult to ensure the accuracy and consistency of transplanting. Existing automated transplanting equipment is not suitable for the multi-layer stereoscopic cultivation mode. During the transplanting process, it is difficult to accurately locate the planting hole, and the efficiency is low. Therefore, an automated transplanting robot is needed to improve production efficiency and quality.

[0012] Based on the above problems, the applicant proposed a multi-layer stereoscopic cultivation automated transplanting robot for plant factories, which integrates a clamping module 4, including a clamping unit 41 and a clamping seat 42, the clamping seat 42 is fixedly connected to the moving seat 31, and a plurality of lifting units 43 are arranged below the clamping seat 42. The clamping unit 41 includes a clamping block 411 detachably connected to the lifting unit 43, a clamping motor 412 is arranged in the clamping block 411, and a plurality of clamping grooves 413 are opened on the lower end surface of the clamping block 411, and a clamping claw 414 is movably arranged in the clamping groove 413. A crank 415 is also arranged in the clamping block 411, and the two ends of the crank 415 are respectively connected to the output end of the clamping motor 412 and the clamping claw 414, so as to realize the opening and closing action of the clamping claw 414. This design enables the robot to accurately clamp and release the planting basket 5, thereby improving the accuracy and consistency of transplantation.

[0013] As attached Figure 6 As shown, for the lifting unit 43, refer to the attached Figure 3 The lifting unit 43 includes a screw rod 431, the end of which is connected to a lifting motor (not shown in the figure), and the lifting motor is connected to the control module signal. It should also be noted that a clamping cover 416 is provided at the upper end of the clamping block 411, and an opening 417 matching the screw rod 431 is provided in the middle of the clamping cover 416.

[0014] More preferably, the mobile module 3 includes an X-axis mobile unit, a Z-axis mobile unit and a steering unit, wherein the X-axis mobile unit includes a horizontal frame 33, and the Z-axis mobile unit includes a vertical frame 32. In this embodiment, the mobile seat 31 is preferably slidably set on the horizontal frame 33 through an electric slide rail structure. Similarly, the horizontal frame 33 is slidably set on the vertical frame 32 through an electric slide rail, and the clamping seat 42 is movably connected to the mobile seat 31 through a steering unit. The steering unit includes a steering rod 44 and a steering motor. One end of the steering rod 44 is fixedly connected to the clamping seat 42. The steering motor is set in the mobile seat 31 and the output end is connected to the other end of the steering rod 44. When the steering motor is started, the relative rotation of the clamping seat 42 and the mobile seat 31 is realized through the steering unit, as shown in the attached figure. Figure 7 As shown, the upper end surface of the moving seat 31 has a U-shaped opening, the steering rod 44 is rotatably arranged in the U-shaped opening, the interior of the moving seat 31 is provided with a cavity, the steering motor 45 is fixedly arranged in the cavity, and the output end of the steering motor 45 is fixedly connected to the other end of the steering rod 44. Another preferred implementation in this embodiment is that an annular limit groove is provided on the upper and lower end surfaces inside the U-shaped opening, and a pin matching the annular limit groove is provided on the steering rod 44 to prevent the center of gravity of the clamping unit 41 from being unbalanced when rotating. It can be understood that the extension direction of the two ends of the horizontal frame 33 is the X-axis direction, the extension direction of the two ends of the vertical frame 32 is the Z-axis direction, and the extension direction of the steering rod 44 when it is perpendicular to the vertical frame 32 is the Y-axis direction.

[0015] For the cultivation process: The clamping module 4 receives instructions through the control module, the clamping motor 412 is started, and the crank 415 is driven to rotate synchronously, thereby controlling the opening and closing of the clamping jaw 414. When it is necessary to clamp the planting basket 5, the clamping jaw 414 is opened to a size large enough to accommodate the planting basket 5, and then the motor drives the screw rod 431 to descend to the target planting basket 5 plane, the annular clamping jaw 414 is closed, and the planting basket 5 is clamped in the ring, and then the control module controls the lifting motor to rotate, so that the internal thread of the opening 417 matches the external thread of the screw rod 431, and the planting basket 5 is lifted from the seedling plate 2, wherein the maximum lifting amount is the depth of the opening 417. Subsequently, the control module changes the height of the planting basket 5 through the Z-axis moving unit. When the planting basket 5 reaches the predetermined height, the control module changes the angle of the planting basket 5 relative to the cultivation rack 1 through the steering unit, that is, rotates the side close to the seedling plate 2 to the side away from the seedling plate 2, and detects whether it has moved to the predetermined position through the visual module, and detects the empty cultivation holes on the cultivation rack 1. Repeat the above process to transplant the seedlings on the seedling plate 2 into the cultivation holes on the cultivation rack 1.

[0016] For the visual module, in this embodiment, its preferred structure includes a binocular visual camera and a multi-degree-of-freedom pan-tilt structure. More preferably, the camera uses the FL3-U3-13S2C model camera. The visual module can be set below the clamping seat 42, on the side of the horizontal frame 33 close to the seedling board 2, and on the side of the vertical frame 32 close to the seedling board 2. In this embodiment, its preferred setting position is on the side of the horizontal frame 33 close to the seedling board 2. The initial position of the clamping module 4 within the visual range of the visual module is that the vertical frame 32 is at the lowest point, and the horizontal frame 33 is at the far right (facing the clamping cultivation frame 1). For the positioning of the target position information, it adopts 3D point cloud data calculation, which will not be repeated here. For the control module, it is a high-performance microcontroller or embedded computer system. In this embodiment, it is preferably an ARMCortex-A series processor, which has sufficient computing power to process data from the visual module and control each module of the robot in real time. The control module should have good interface expansibility and be able to exchange data and transmit commands efficiently with the visual module, positioning module, mobile module 3 and clamping module 4; for the positioning module, it is preferably a laser radar or a stereoscopic vision system, which can provide high-precision 3D point cloud data for accurately locating the position of the planting basket 5 and the seedlings. In this embodiment, the Velodyne series of laser radars are selected, which can generate accurate three-dimensional maps of the surrounding environment in real time with their high precision and fast scanning capabilities; for the mobile module 3, the preferred structure is an electric slide rail system, including a mobile seat 31, a vertical frame 32 and a horizontal frame 33, which can achieve precise movement on the X and Z axes. The mobile seat 31 and the horizontal frame 33 are driven by servo motors to ensure the accuracy and stability of the movement. The vertical frame 32 is made of high-strength material to ensure the stability and bearing capacity of the structure.

[0017] Embodiment 2: This embodiment only describes the parts that are different from Embodiment 1, specifically: the clamping module 4 includes a clamping unit 41 and a clamping seat 42, the clamping seat 42 is movably connected to the movable seat 31, a plurality of lifting units 43 are arranged below the clamping seat 42, the clamping unit 41 includes a clamping block 411 detachably connected to the lifting unit 43, a clamping motor 412 is arranged in the clamping block 411, a plurality of clamping grooves 413 are opened on the lower end surface of the clamping block 411, a clamping jaw 414 is movably arranged in the clamping groove 413, a crank 415 is also arranged in the clamping block 411, and the two ends of the crank 415 are respectively connected to the output end of the clamping motor 412 and the clamping jaw 414, when the clamping motor 412 is started, the opening and closing of the clamping jaw 414 is realized by the synchronous rotation of the crank 415. A force sensor is disposed inside the clamping jaw 414 , and the force sensor is connected to a control module signal. A force feedback model is installed inside the control module, and the force feedback model is used to monitor the force applied during the gripping process of the clamping jaw 414 .

[0018] Based on the above mechanism, it can be understood that the technical problem addressed by this embodiment is how to achieve accurate and safe clamping and transplanting operations on the planting basket 5, while ensuring the protection of the seedlings during the transplanting process and improving the efficiency and accuracy of the operation. This embodiment provides a solution with a complex structure but precise operation through the design of the clamping module 4. More specifically, the clamping module 4 is composed of a clamping unit 41 and a clamping seat 42. The clamping seat 42 is fixedly connected to the movable seat 31, and a number of lifting units 43 are arranged under the clamping seat 42. The above structural design enables the clamping module 4 to not only move in the horizontal direction, but also to make precise position adjustments in the vertical direction; the clamping unit 41 and the lifting unit 43 are connected in a detachable manner, which increases the flexibility of the equipment and the convenience of maintenance. A clamping motor 412 is arranged in the clamping block 411, which is the core power source for realizing the clamping action. A plurality of clamping grooves 413 are formed on the lower end surface of the clamping block 411 , and clamping claws 414 are movably disposed in the clamping grooves 413 . The clamping claws 414 are responsible for directly contacting the planting basket 5 to achieve the clamping function.

[0019] In addition, a crank 415 is provided in the clamping block 411, and the two ends of the crank 415 are respectively connected to the output end of the clamping motor 412 and the clamping jaw 414. When the clamping motor 412 is started, the opening and closing action of the clamping jaw 414 is realized through the synchronous rotation of the crank 415. Such a mechanical transmission design ensures the accuracy and reliability of the clamping action. In order to protect the seedlings from damage, force sensors are provided on the inner side of the clamping jaw 414. These force sensors can monitor the force applied by the clamping jaw 414 during the grasping process in real time and transmit the data to the control module.

[0020] Embodiment 3: This embodiment only describes the parts that are different from Embodiment 1, specifically: The force feedback model satisfies: ; ; in, is the control signal; is the proportional gain, which is used to adjust the immediate response of the error; is the integral gain, which is used to adjust the cumulative effect of the error; is the differential gain, which is used to adjust the rate of change of the error; is the target force, i.e., the force applied by the clamp 414 set by the control module; To measure the force, that is, the actual force applied by the clamping jaw 414 is measured by a force sensor; is the error, i.e. the difference between the target force and the measured force; is the integral of the error, indicating the accumulation of errors from the beginning to the current moment; is the rate of change of the error, which indicates how fast the error changes over time.

[0021] A safety threshold is set in the force sensor. When the measured force exceeds the safety threshold, the control module generates a feedback signal for reducing the actual force applied by the clamp 414 .

[0022] It should be noted that the control module is equipped with a force feedback model, which is a key technical component. It calculates the difference between the current force applied by the clamp 414 and the preset target force through an algorithm based on the data provided by the force sensor, and adjusts the clamping force of the clamp 414 accordingly. This force feedback control mechanism can ensure that the force applied by the clamp 414 is always kept within a safe range during the process of grasping and releasing the planting basket 5, avoiding damage to the seedlings. The working principle of the force feedback model is based on the PID control algorithm in control theory. By adjusting the three parameters of proportion (P), integration (I) and differentiation (D), fine control of the force of the clamp 414 is achieved.

[0023] Specifically, the proportional gain is used to adjust the immediate response of the error, the integral gain is used to adjust the cumulative effect of the error, and the differential gain is used to adjust the rate of change of the error. The control module sets a target force value, and the difference between the actual force value measured by the force sensor and the target force value is the error. The control module calculates the control signal based on the error and adjusts the driving signal of the clamp 414 to maintain the target force. If the measured force exceeds the set safety threshold, the control module will generate a feedback signal to reduce the actual force applied by the clamp 414 to protect the plant from damage.

[0024] Embodiment 3: This embodiment only describes the parts that are different from the first embodiment. Specifically, the control module generates a motion signal through the visual module, and the motion signal satisfies: ; in, The instant location coordinates of the planting basket; is the target position coordinate of the planting basket.

[0025] It should be noted that, in this embodiment, due to the presence of the steering unit, when the height reaches the predetermined position, the steering action of the steering unit replaces the movement on the Y-axis, so there is no coordinate movement in the sense of the Y-axis. The essence of the movement signal formula is the Pythagorean theorem, that is, in the established three-axis coordinate system, in the plane formed by the X-axis and the Z-axis, when the instantaneous position coordinates and the target position are determined, the distance calculated by the movement signal is the straight-line distance in the plane. Since the X-axis and the Z-axis must be perpendicular, the shortest distance moved on the X-axis and the Z-axis can also be directly calculated by the Pythagorean theorem.

[0026] Embodiment 4: This embodiment only describes the parts that are different from Embodiment 1, specifically: The visual module is equipped with a target detection model, which extracts features through a feature extraction network to obtain a feature map; and inputs the feature map into a region proposal network to generate multiple seedling stem target candidate regions; then the feature map and the candidate regions are jointly input into the RoIAlign layer, so that each target candidate region is normalized to the same scale, and the pixels in the corresponding feature map are completely aligned with the pixels in the original image; finally, the target features are extracted from the feature map, and then output to the fully connected layer and the fully convolutional network respectively for target classification and instance segmentation to generate categories, bounding boxes and binary images.

[0027] The visual module is also equipped with a state judgment model, which is used to judge the growth status of the seedlings. The judgment process includes: using OPENCV to obtain the contour information of the seedling branches and stems; obtaining the coordinates of the center of mass of the seedlings and determining the stems; determining the center point of the branch bifurcation position; locating the stem state operation point; and judging the range of the stem state operation point.

[0028] The state judgment model inputs a range of growth state information, and when the stem state operation point belongs to the range, the growth state of the seedling is judged to be good, and when the stem state operation point does not belong to the range, the growth state of the seedling is judged to be abnormal. The feature extraction network is a ResNet50 structure and an FPN feature pyramid network.

[0029] It should be noted that the working principle of the target detection model is based on deep learning. First, through the feature extraction network, such as the ResNet50 structure and the FPN feature pyramid network, key features are extracted from the input image to form a feature map. This step can capture important information in the image, such as shape, texture, and color, etc., laying the foundation for subsequent target recognition. Further, the feature map is input into the region proposal network (RPN), which is a network specially designed to generate potential target regions (i.e. candidate regions). RPN can identify areas in the image that may contain targets and generate multiple candidate regions for these areas. These candidate regions are then compared with the feature Figure 1 The image is then input to the RoIAlign layer, which normalizes each candidate region to the same scale and ensures that the pixels in the feature map are completely aligned with the pixels in the original image. This process is crucial to improving the accuracy of target detection because it ensures that targets at different scales and positions can be accurately identified.

[0030] After feature extraction and alignment, the model further extracts target features in the feature map and outputs them to the fully connected layer and the fully convolutional network respectively. The fully connected layer is responsible for target classification and determining the target category in the candidate area; while the fully convolutional network is responsible for instance segmentation and generates an accurate bounding box and binary image of the target. The target area in the binary image is marked as white and the background is black, which provides accurate target location information for subsequent transplantation operations.

[0031] The state judgment model is used to evaluate the growth status of seedlings. Its workflow includes using the OPENCV library to obtain the contour information of the seedling branches and stems, determine the center of mass coordinates and stem position of the seedlings, identify the center point of the branch bifurcation, locate the stem state operation point, and determine whether the stem state operation point is within the preset growth status information range. This model determines whether the growth status of the seedlings is good or abnormal based on the input growth status information range, thereby providing the automated transplanting robot with a decision basis for whether transplanting operations are needed.

[0032] It should also be noted that, since the optimal number of layers required for the neural network structure is unknown during the training process, and its depth depends on the complexity of the data set, it is possible to add skip connections to the network structure, allowing the training of useless layer parameters to be skipped when training the network structure, ignoring the parameters of the skipped layers of the network structure, effectively reducing the number of parameters and improving the effectiveness of the model. Therefore, the ResNet network structure is dynamic during training, and the number of layers is optimally adjusted during the training process. In the seedling detection, the categories are bifurcated and non-bifurcated, and the detection categories are simple. Therefore, in this embodiment, ResNet50 is preferably used as the backbone neural network for model training.

[0033] In addition, since there are many seedlings on the seedling plate, there is also the problem of multi-scale target detection. FPN adopts a top-down hierarchical structure with lateral connections, from single-scale input to building a network feature pyramid, fusing different feature maps so that each layer of the fused network has deep and shallow features.

[0034] For image preprocessing, preprocessing was performed before training. The images were uniformly processed to a size not exceeding 800×800 pixels. The data set was expanded to at least 3,000 images by adding Gaussian white noise, random cropping, and changing the color tone.

[0035] For the target detection process: by reading the binary image separately, calling the contour extraction function in OPENCV to retrieve the contour in the binary image, and obtaining the contour information of the branch stem and the main stem respectively.

[0036] After obtaining the contour information of the branch and stem, the stem and branch are determined by obtaining the centroid coordinates of each contour, so as to perform the processing operation. First, the coordinates of the centroid point a of the branch are determined, and the coordinates of the mass points b and c of the two nearest stems are obtained. The stem above the branch is determined among the adjacent stems by comparing the ordinates of points b and c. Finally, the branch a and stem b to be processed are determined.

[0037] The contour fitting straight line method was used to fit straight lines to the branch stems and the main stems respectively and obtain the straight line equations. The point where the two straight lines intersected was found, which was the center point of the bifurcation of the seedling stems.

[0038] In the two-dimensional image, the width of the identified stem segmentation area is assumed to be a constant. According to the contour information of the stem and branch, the stem centroid A and branch centroid B can be obtained. The angles between the fitted stem centerline and branch centerline and the horizontal axis can be determined, thereby determining the equation of the fitted centerline and the coordinates of the intersection with the centerline (branch center point). The entire branch operation point can be obtained through the relationship between the branch center points. The introduction of the state judgment model enables the automated transplanting robot to evaluate the growth state of the seedlings, which is of guiding significance for deciding whether to perform the transplanting operation. By obtaining the contour information of the seedling branches and stems through OPENCV, and combining the range of the growth state information, the robot can judge whether the growth state of the seedlings is good, thereby avoiding transplanting seedlings with poor growth status and ensuring that the transplanted seedlings have good growth potential.

[0039] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-layer stereoscopic cultivation automated transplanting robot for a plant factory, comprising a cultivation rack (1), wherein a seedling raising plate (2) is arranged on the side of the cultivation rack (1), and characterized in that: The robot comprises a control module, a vision module, a positioning module, a moving module (3) and a clamping module (4); the moving module (3) comprises a moving seat (31), a vertical frame (32) and a horizontal frame (33); the horizontal frame (33) is slidably arranged on the vertical frame (32) and is connected to the control module by signal; the moving seat (31) is slidably arranged on the horizontal frame (33) and is connected to the control module by signal; the vision module is connected to the control module by signal and is used for collecting position information; the positioning module is connected to the control module by signal and is used for determining position information; the clamping module (4) is arranged on the moving seat (31) and clamps and fixes the planting basket (5) in the seedling raising plate (2) through the control module; The process of determining the position information is as follows: identifying the position of the seedlings through a visual module and scanning to obtain 3D point cloud data, then processing the 3D point cloud data to extract the position information and shape information of the planting basket (5) and the seedlings, and establishing a coordinate system, and finally transmitting the boundary position information of the planting basket (5) to the control module.

2. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 1, characterized in that: The clamping module (4) comprises a clamping unit (41) and a clamping seat (42), the clamping seat (42) being movably connected to the movable seat (31), a plurality of lifting units (43) being arranged below the clamping seat (42), the clamping unit (41) comprising a clamping block (411) detachably connected to the lifting unit (43), a clamping motor (412) being arranged in the clamping block (411), a plurality of clamping grooves (413) being arranged on the lower end surface of the clamping block (411), a clamping claw (414) being movably arranged in the clamping groove (413), a crank (415) being further arranged in the clamping block (411), two ends of the crank (415) being respectively connected to the output end of the clamping motor (412) and the clamping claw (414), and when the clamping motor (412) is started, the opening and closing of the clamping claw (414) is realized by the synchronous rotation of the crank (415).

3. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 2, characterized in that: A force sensor is arranged on the inner side of the clamp (414), and the force sensor is connected to a control module signal. A force feedback model is installed in the control module, and the force feedback model is used to monitor the force applied by the clamp (414) during the grasping process.

4. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 2, characterized in that: The force feedback model satisfies: ； ； in, is the control signal; is the proportional gain, which is used to adjust the immediate response of the error; is the integral gain, which is used to adjust the cumulative effect of the error; is the differential gain, which is used to adjust the rate of change of the error; is the target force, i.e., the force applied by the clamping jaw (414) set by the control module; To measure the force, that is, the actual force applied by the clamping jaw (414), measured by a force sensor; is the error, i.e. the difference between the target force and the measured force; is the integral of the error, indicating the accumulation of errors from the beginning to the current moment; is the rate of change of the error, which indicates how fast the error changes over time.

5. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 4, characterized in that: A safety threshold is set in the force sensor. When the measured force exceeds the safety threshold, the control module generates a feedback signal for reducing the actual force applied by the clamp.

6. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 1, characterized in that: The control module generates a movement signal through the visual module, and the movement signal satisfies: ； in, The instant location coordinates of the planting basket; is the target position coordinate of the planting basket.

7. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 1, characterized in that: The visual module is equipped with a target detection model, and the target detection model extracts features through a feature extraction network to obtain a feature map; The feature map is input into the region proposal network to generate multiple candidate regions of seedling stem targets; The feature map and the candidate region are then input into the RoIAlign layer together, so that each target candidate region is normalized to the same scale, and the pixels in the corresponding feature map are completely aligned with the pixels in the original image; Finally, the target features are extracted from the feature map and then output to the fully connected layer and the fully convolutional network for target classification and instance segmentation to generate categories, bounding boxes, and binary images.

8. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 7, characterized in that: The visual module is also equipped with a state judgment model, which is used to judge the growth status of the seedlings. The judgment process includes: using OPENCV to obtain the contour information of the seedling branches and stems; obtaining the coordinates of the center of mass of the seedlings and determining the stems; determining the center point of the branch bifurcation position; locating the stem state operation point; and judging the range of the stem state operation point.

9. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 8, characterized in that: A range of growth status information is input into the state judgment model. When the stem state operation point belongs to the range, the growth state of the seedling is judged to be good. When the stem state operation point does not belong to the range, the growth state of the seedling is judged to be abnormal.

10. The multi-layer stereoscopic cultivation automated transplanting robot for a plant factory according to claim 7, characterized in that: The feature extraction network is a ResNet50 structure and an FPN feature pyramid network.

Images