Device and method for automatic sampling inspection and data binding of plasma bags

Abstract

The invention provides a device and method for automatic sampling inspection and data binding of plasma bags, and belongs to the technical field of biopharmaceutical production automation. The device comprises a conveying module, a counting and positioning module, a robot grabbing module, a weighing module, an information acquisition module and a control module; the conveying module comprises a first plasma conveying track and a second plasma conveying track, and further comprises a belt control motor and a belt control encoder; the counting and positioning module comprises a plasma bag counting and positioning device; the robot grabbing module comprises a plasma grabbing mechanical arm and a plasma putting-back mechanical arm; the weighing module comprises a weighing platform; the information acquisition module comprises a photographing lens, a light supplementing lamp and a code scanner; the control module comprises a control screen. The system stably operates at a low temperature, supports flexible setting of a sampling inspection proportion, adapts to production requirements of different batches, and improves the compliance and the intelligent level of plasma sampling inspection in the biopharmaceutical industry.

Description

An apparatus and method for automated sampling and data binding of blood plasma bags. Technical Field

[0001] This application belongs to the field of biopharmaceutical production automation technology, specifically relating to a device and method for automatic sampling and data binding of plasma bags. Background Technology

[0002] In the biopharmaceutical industry, after plasma collection, it is stored in standard plasma bags (approximately 290mm long and 110mm wide), frozen, and then transported to the production base for storage at -20℃. According to industry requirements, plasma is subject to random sampling upon arrival at the warehouse, with the core verification indicator being that the net weight (including packaging) of each plasma bag does not exceed 620g.

[0003] In existing technologies, manual sampling is commonly used. The process involves operators manually selecting plasma bags of the appropriate proportion from the plasma tank, weighing each bag on an electronic scale, and then manually filling out a paper record containing information such as plasma batch number, specifications, unique identification code, and weight. Finally, an administrator manually enters this information into a computer system. This method has the following problems: low efficiency (the manual selection, handling, weighing, and recording process is cumbersome, time-consuming, and labor-intensive); data error-prone (manual copying and entry are prone to errors, omissions, or illegible handwriting, affecting data accuracy); and weak traceability (the correspondence between paper records and physical plasma bags is easily confused, making it difficult to achieve one-item-one-code-one-weight binding).

[0004] Therefore, there is an urgent need for an automated device that integrates conveying, counting and positioning, robotic grasping, weighing, information collection and control modules to achieve automated and high-precision operation of the entire process of plasma bag sampling, weighing and data binding. Summary of the Invention

[0005] The purpose of this invention is to provide an apparatus and method for automatic sampling and data binding of plasma bags, in order to solve the problems of low efficiency, data error-proneness and poor data traceability in the prior art.

[0006] In a first aspect, embodiments of this application provide an apparatus for automatic sampling and data binding of plasma bags. The apparatus includes: a conveying module, a counting and positioning module, a robot gripping module, a weighing module, an information acquisition module, and a control module. The conveying module includes a first plasma conveying track and a second plasma conveying track arranged sequentially in a horizontal direction. The counting and positioning module includes a plasma bag counting and positioning device, which is fixedly installed above the first plasma conveying track by a bracket. The robot gripping module includes a plasma gripping robotic arm and a plasma return robotic arm. The plasma gripping robotic arm is disposed on one side of the first plasma conveying track, and its body is fixed to the outer frame of the first plasma conveying track away from the second plasma conveying track. Initially, the arm hangs down and is positioned directly above the first plasma conveying track. The plasma return robotic arm is positioned on one side of the second plasma delivery track, on the same side as the plasma gripping robotic arm. Its body is fixed to the outer frame of the second plasma delivery track, away from the first plasma delivery track. Initially, the arm hangs down and is directly above the second plasma delivery track. The weighing module includes a weighing platform positioned between the plasma gripping robotic arm and the plasma return robotic arm. The information acquisition module includes a camera lens, a supplementary light, and a barcode scanner, positioned directly above the weighing platform. The control module includes a control screen embedded in the frame panel on the operating side of the device. The control screen contains a main control computer, which is connected to the delivery module, counting and positioning module, robotic gripping module, weighing module, and information acquisition module.

[0007] Furthermore, the conveying module also includes a belt-controlled motor and a belt-controlled encoder. The belt-controlled motor is connected to the belt of the first plasma conveying track and the pulley of the second plasma conveying track respectively. The belt-controlled encoder is coaxially connected to the output shaft of the belt-controlled motor, and both the belt-controlled motor and the belt-controlled encoder are connected to the main control computer.

[0008] Furthermore, the installation height of the plasma bag counter locator is 30-50cm away from the conveying surface of the first plasma conveying track, and a camera is installed on the counter locator.

[0009] Furthermore, both the plasma grabbing robot arm and the plasma return robot arm include a robot arm body and an end effector, with the end effector facing the corresponding conveying track or weighing platform.

[0010] Furthermore, the weighing module also includes a weighing sensor, which is embedded inside the platform surface of the weighing platform, and the sensing area of ​​the weighing sensor coincides with the center of the platform surface of the weighing module.

[0011] Furthermore, the shooting direction of the camera lens, the illumination direction of the fill light, and the scanning direction of the barcode scanner are all oriented towards the center of the weighing platform of the weighing module.

[0012] Secondly, this application also provides a method for automatic sampling and data binding of plasma bags. The method is based on the device described in the first aspect and includes the following steps: S1. Inputting or confirming the plasma information and sampling parameters of the current batch through the control screen of the control module. The plasma information includes batch number and specifications, and the sampling parameters include sampling ratio or sampling quantity. The main control computer built into the control screen receives the parameters and automatically calculates the grab interval to complete the system startup; S2. Placing the entire batch of plasma bags to be inspected sequentially at the feed end of the first plasma conveying track of the conveying module. The main control computer sends a start command to the first plasma conveying track, and the first plasma conveying track drives the plasma bags to move along the conveying direction; When the plasma bag enters the detection area of ​​the plasma bag counter locator of the counting and positioning module, the camera on the counter locator collects the 2D image and 3D point cloud data of the plasma bag; S3. YOLOv8-OBB is used to process the 2D image, outputting the bounding box coordinates, counting signal, attitude tilt angle θ, and category confidence of the plasma bag. Simultaneously, the 3D point cloud data is denoised, and local planes on the bag surface are searched as candidate adsorption locations. All data is synchronously transmitted to the main control computer. S4. When the counting result counted by the main control computer reaches the preset grasping interval, an action command is sent to the plasma grasping robot arm, transmitting the attitude tilt angle θ and candidate adsorption locations to the robot arm. S5. The main control computer calculates the compensation angle and drives the plasma grasping robot arm to rotate to the corresponding compensation angle. The plasma grasping robot arm moves above the first plasma delivery track, grasps the target plasma bag, and transfers it to the center of the weighing platform of the weighing module before resetting. The main control computer controls the supplementary light to turn on, the camera lens to capture images of the plasma bag, the barcode scanner to scan the plasma bag identification code, and the weighing platform's built-in weighing sensor to acquire the plasma bag's weight data. The main control computer automatically binds and stores the identification code, weight data, image, attitude tilt angle θ, point cloud data, and timestamp, and simultaneously determines whether the weight meets the preset weight, outputting a yes or no result; S6. The main control computer sends a command to the plasma return robot arm, which picks up the inspected plasma bag from the weighing platform and moves it above the second plasma delivery track; the main control computer controls the first plasma delivery track to pause operation, and after the second plasma delivery track clears its placement area, the plasma return robot arm releases the plasma bag; after a delay of m seconds, the first plasma delivery track resumes operation.

[0013] Further, the specific steps for preprocessing the 2D image using YOLOv8-OBB in step S3 are as follows: S31. Perform Gaussian filtering preprocessing on the 2D image to suppress random noise caused by low-temperature frost and retain the outline features of the blood plasma bag; S32. Adjust the gain of the RGB channels of the image to offset the color gamut shift caused by uneven light reflection in the cold storage environment and unify the color features of the image; S33. Scale the preprocessed image to 640×640 pixels and use letterbox proportional padding to maintain the original aspect ratio of the image to obtain the model input image; S34. [The text abruptly ends here, likely due to an incomplete sentence or a formatting error.] The input image is fed into the YOLOv8-OBB feature extraction network, which sequentially fuses shallow details and deep semantic features through the C2f module and enhances global feature capture capabilities through the SPPF module; S35. The classification branch outputs the class confidence score in the 0~1 interval using the sigmoid activation function, the regression branch outputs the bounding box coordinates in the pixel coordinate system using CIoU loss, and the pose branch outputs the pose tilt angle θ using MAE loss; S36. The number of valid detection boxes is counted to generate a counting signal, and the bounding box coordinates, counting signal, pose tilt angle θ, and class confidence score are synchronously transmitted to the main control computer.

[0014] Further, in step S5, the formula for calculating the compensation angle is: Δθ = -θ + 0.5°.

[0015] Furthermore, the method also includes the following steps: S7. Repeat steps S2-S6 until the preset sampling quantity is completed, and the main control computer generates a complete sampling report.

[0016] As can be seen from the above technical solutions, this invention has the following advantages: Relying on the collaborative design of dual conveyor tracks and dual robotic arms, it achieves fully automated, unmanned operation of plasma bag feeding, counting, grasping, weighing, and return, avoiding repetitive manual operations in low-temperature environments, reducing the risk of human contamination, and improving sampling efficiency; The information acquisition module simultaneously completes barcode scanning, photography, and weighing, and the control module automatically binds the identification code, weight, image, and timestamp, ensuring high data integrity; Precise positioning and posture correction of the plasma bags further enhance the uniqueness of data association, providing a complete electronic evidence chain for quality traceability; Each module is arranged in a reasonable position to work collaboratively, with seamless connection between robotic arm grasping and track conveying, avoiding process congestion. This invention operates stably at low temperatures and supports flexible setting of sampling ratios to adapt to different batch production needs, improving the compliance and intelligence level of plasma sampling in the biopharmaceutical industry. Attached Figure Description

[0017] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying 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.

[0018] Figure 1 is a schematic diagram of the device for automatic sampling and data binding of plasma bags according to this application; Figure 2 is a schematic diagram of the information acquisition module according to this application; Figure 3 is a flowchart of the method for automatic sampling and data binding of plasma bags according to this application.

[0019] The components include: 1. First plasma transport track; 2. Second plasma transport track; 3. Plasma bag counter and locator; 4. Plasma grabbing robot arm; 5. Plasma return robot arm; 6. Weighing platform; 7. Camera lens; 8. Fill light; 9. Barcode scanner; 10. Control panel; 11. Belt-controlled motor; 12. Belt-controlled encoder. Detailed Implementation

[0020] Various embodiments of the invention will be described more fully in the apparatus and method for automated sampling and data binding of plasma bags, which will be described in detail below. The invention may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the invention to the specific embodiments disclosed herein, but rather the invention should be understood to cover all modifications, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of the invention.

[0021] It should be understood that, when used in this specification, the term "comprising" indicates the presence of the described feature, integral, step, operation, element, and / or component, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0022] The terms "one embodiment" or "some embodiments" used in this application mean that one or more embodiments of this application include the specific features, structures, or characteristics described in that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this application do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0023] To facilitate a clear description of the technical solutions of this application, the terms "first" and "second" are used to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" do not necessarily imply that they are different.

[0024] To make the objectives, features, and advantages of this invention more apparent and understandable, specific embodiments and accompanying drawings will be used to clearly and completely describe the technical solutions protected by this invention. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0025] Please refer to Figure 1 for a device for automatic sampling and data binding of plasma bags. The device includes: a conveying module, a counting and positioning module, a robot gripping module, a weighing module, an information acquisition module, and a control module. The conveying module includes a first plasma conveying track 1 and a second plasma conveying track 2 arranged sequentially in a horizontal direction. The counting and positioning module includes a plasma bag counting and positioning device 3, which is fixedly installed above the first plasma conveying track 1 by a bracket. The robot gripping module includes a plasma gripping robotic arm 4 and a plasma return robotic arm 5. The plasma gripping robotic arm 4 is located on one side of the first plasma conveying track 1, and its body is fixed to the outer frame of the first plasma conveying track 1 away from the second plasma conveying track 2. In the initial state, the arm hangs down and is located directly above the first plasma conveying track 1. The plasma return robotic arm 5 is located on the first plasma conveying track 1... The second plasma delivery track 2 is located on one side, on the same side as the plasma grabbing robot arm 4. Its body is fixed to the outer frame of the second plasma delivery track 2 away from the first plasma delivery track 1. In the initial state, the arm hangs down and is located directly above the second plasma delivery track 2. The weighing module includes a weighing platform 6, which is set between the plasma grabbing robot arm 4 and the plasma return robot arm 5. As shown in Figure 2, the information acquisition module includes a camera lens 7, a supplementary light 8, and a barcode scanner 9, which are set directly above the weighing platform 6. The control module includes a control screen 10, which is embedded in the frame panel on the operating side of the device. The control screen 10 has a built-in main control computer, which is connected to the delivery module, the counting and positioning module, the robot grabbing module, the weighing module, and the information acquisition module.

[0026] It should be noted that the mounting racks of all components of the device described in this embodiment are adapted to a -20℃ cold storage environment. The connecting cables between the components are made of cold-resistant and waterproof cables, and the cables are routed inside the rack to avoid external interference.

[0027] As a refinement and extension of the specific implementation of the above embodiments, in order to fully illustrate the specific implementation process of this embodiment, another device for automatic sampling and data binding of plasma bags is provided. This device includes: a conveying module, a counting and positioning module, a robot grasping module, a weighing module, an information acquisition module, and a control module. The conveying module includes a first plasma conveying track 1 and a second plasma conveying track 2 arranged sequentially in the horizontal direction. In some embodiments, the gap width between the first plasma conveying track 1 and the second plasma conveying track 2 is 1.2-1.5 times the width of the weighing platform of the weighing module to ensure that the robot grasping system operates without interference.

[0028] The counting and positioning module includes a plasma bag counting and positioning device 3, which is fixedly installed above the first plasma delivery track 1 via a bracket. The robotic gripping module includes a plasma gripping robotic arm 4 and a plasma return robotic arm 5. The plasma gripping robotic arm 4 is located on one side of the first plasma delivery track 1, with its body fixed to the outer frame of the first plasma delivery track 1 away from the second plasma delivery track 2. In its initial state, the arm hangs down and is located directly above the first plasma delivery track 1. The plasma return robotic arm 5 is located on one side of the second plasma delivery track 2, on the same side as the plasma gripping robotic arm 4, and its body is fixed to the second plasma delivery track. 2. At the outer frame away from the first plasma delivery track 1, the arm is initially hanging down and positioned directly above the second plasma delivery track 2; the weighing module includes a weighing platform 6, which is located between the plasma grabbing robot arm 4 and the plasma return robot arm 5; the information acquisition module includes a camera lens 7, a supplementary light 8, and a barcode scanner 9, which are located directly above the weighing platform 6; in some embodiments, the barcode scanner 9 is fixed by an adjustable bracket, the height of which can be adjusted vertically to keep the distance between the scanning window of the barcode scanner 9 and the surface of the weighing platform 6 of the weighing module 20-30cm.

[0029] The control module includes a control panel 10, which has a built-in main control computer. The main control computer is connected to the conveying module, the counting and positioning module, the robot grasping module, the weighing module, and the information acquisition module.

[0030] The conveying module also includes a belt control motor 11 and a belt control encoder 12. The belt control motor 11 is connected to the belt of the first plasma conveying track 1 and the pulley of the second plasma conveying track 2 respectively. The belt control encoder 12 is coaxially connected to the output shaft of the belt control motor 11, and both the belt control motor 11 and the belt control encoder 12 are connected to the main control computer.

[0031] The installation height of the plasma bag counter locator 3 is 30-50cm away from the conveying surface of the first plasma conveying track 1, and a camera is installed on the counter locator 3.

[0032] Both the plasma grabbing robot arm 4 and the plasma return robot arm 5 include a robot arm body and an end effector, with the end effector facing the corresponding conveying track or weighing platform 6.

[0033] The weighing module also includes a weighing sensor, which is embedded inside the platform 6 of the weighing platform. The sensing area of ​​the weighing sensor coincides with the center of the platform of the weighing module.

[0034] The shooting direction of the camera lens 7, the illumination direction of the fill light 8, and the scanning direction of the barcode scanner 9 are all oriented towards the center of the weighing platform 6 of the weighing module.

[0035] Please refer to Figure 3, which is a flowchart of a method for automatic sampling and data binding of plasma bags. The method includes the following steps: S1. Input or confirm the plasma information and sampling parameters of the current batch through the control screen 10 of the control module. The plasma information includes the batch number and specifications. The sampling parameters include the sampling ratio or sampling quantity. After receiving the parameters, the main control computer built into the control screen 10 automatically calculates the grab interval and completes the system startup. In some embodiments, at the plasma storage sampling site in a -20℃ cold storage, the operator stands on the operating side of the device, facing the control screen 10 embedded in the rack panel, and inputs the plasma information of the current batch through touch operation: batch number "20250601-XJ", specifications "290mm×110mm standard plasma bag", and confirms the sampling parameter as a sampling ratio of 10% (corresponding to a total batch quantity of 500 bags, of which 50 bags need to be sampled). The main control computer built into the control panel 10 receives the above parameters in real time, automatically calculates the grab interval of 8 seconds based on the preset running speed of the first plasma transport track 1 (0.3 m / s) and the distance between plasma bags, and synchronously completes the initial self-test of each module. Finally, the control panel 10 displays that the system is ready, completes the overall startup, and waits for the subsequent feeding and sampling process.

[0036] S2. The entire batch of plasma bags to be inspected is placed sequentially at the feed end of the first plasma transport track 1 of the transport module. The main control computer sends a start command to the first plasma transport track 1, and the first plasma transport track 1 moves the plasma bags along the transport direction. When the plasma bag enters the detection area of ​​the plasma bag counter locator 3 of the counting and positioning module, the camera on the counter locator 3 acquires the 2D image and 3D point cloud data of the plasma bag. In some embodiments, the entire batch of 500 plasma bags to be inspected (290mm×110mm) is placed sequentially and stably at the feed end of the first plasma transport track 1 of the transport module to avoid bag stacking. After receiving the system ready signal, the main control computer sends a start command to the first plasma transport track 1, and the track moves the plasma bags at a constant speed of 0.3m / s along the transport direction. When the first plasma bag enters the detection area of ​​the plasma bag counter locator 3 (40cm from the transport surface) of the counting and positioning module, the camera coaxially mounted with the locator starts synchronously and quickly acquires the 2D image and 3D point cloud data of the plasma bag. S3. YOLOv8-OBB is used to process the 2D image, outputting the bounding box coordinates, count signal, attitude tilt angle θ, and category confidence of the plasma bag. Simultaneously, denoising of the 3D point cloud data is performed, and local planes on the bag surface are searched as adsorption candidate locations. All data is synchronously transmitted to the main control computer. The specific steps of preprocessing the 2D image using YOLOv8-OBB in step S3 are as follows: S31. Perform Gaussian filtering preprocessing on the 2D image to suppress random noise caused by low-temperature frost and preserve the outline features of the plasma bag; S32. Adjust the RGB channel gain of the image to compensate for color gamut shift caused by uneven light reflection in the cold storage environment and unify the image color features; S33. Scale the preprocessed image to 640×6. S34. The model input image is obtained by using letterbox proportional padding to maintain the original aspect ratio of the image, with a resolution of 40 pixels. S35. The model input image is input into the YOLOv8-OBB feature extraction network, which sequentially fuses shallow details and deep semantic features through the C2f module and enhances global feature capture capability through the SPPF module. S36. The classification branch outputs the class confidence score in the range of 0 to 1 using the sigmoid activation function, the regression branch outputs the bounding box coordinates in the pixel coordinate system using CIoU loss, and the pose branch outputs the pose tilt angle θ using MAE loss. S37. The number of valid detection boxes is counted to generate a counting signal, and the bounding box coordinates, counting signal, pose tilt angle θ, and class confidence score are synchronously transmitted to the main control computer.

[0037] In some embodiments, 5000 2D images of blood plasma bags under a -20℃ cold storage environment are acquired, covering conditions such as frost, wrinkles, and tilting from -15° to 15°. The OBB rotation bounding boxes, pose tilt angle θ, and category labels are manually labeled. The training, validation, and test sets are divided in an 8:1:1 ratio. Preprocessing steps S31-S33 are performed on the training set images (Gaussian filter parameters are set to 3×3 convolution kernels with a standard deviation of 0.8; RGB channels are corrected using the grayscale world algorithm; the images are scaled to 640×640 pixels and filled with letterboxes). Based on the YOLOv8-OBB architecture, the C2f module branch number is set to 3, the SPPF module pooling kernel is set to 1×1 / 2×2 / 3×3 / 6×6, the multi-task loss function weights are Lcls:Lreg:Lpose=1:5:2, and the AdamW optimizer (initial learning rate 1e-4) is used. Training was iterated for 300 rounds with a size of 32. Training was stopped when the confidence level of the validation set class was ≥0.95 and the attitude tilt angle MAE was ≤0.5°. The optimal weights were then exported to the main control computer.

[0038] S4. When the counting result counted by the main control computer reaches the preset grasping interval, an action command is sent to the plasma grasping robot arm 4, transmitting the attitude tilt angle θ and the candidate adsorption positions to the plasma grasping robot arm 4. In some embodiments, the main control computer counts the detection results of the plasma bag counter locator 3 in real time. When the count reaches the preset grasping interval (8 seconds / bag), it immediately sends a grasping action command to the plasma grasping robot arm 4 of the robot grasping module. At the same time, the output attitude tilt angle θ (3°) of the first bag of plasma and the two vacuum adsorption candidate positions obtained by 3D point cloud processing are synchronously transmitted to the plasma grasping robot arm 4. The main body of the robot arm begins to rotate, and the attitude adjustment joint prepares to calculate the compensation angle according to the θ value to prepare for precise grasping.

[0039] S5. The main control computer calculates the compensation angle and drives the plasma grabbing robot arm 4 to rotate to the corresponding compensation angle; the plasma grabbing robot arm 4 moves to above the first plasma transport track 1, grabs the target plasma bag and transfers it to the center of the weighing platform 6 of the weighing module, and then resets; the main control computer controls the supplementary light 8 to light up, the camera lens 7 to capture the image of the plasma bag, the barcode scanner 9 to scan the plasma bag identification code, and the weighing sensor built into the weighing platform 6 to acquire the weight data of the plasma bag. The main control computer automatically binds and stores the identification code, weight data, image, attitude tilt angle θ, point cloud data and timestamp, and at the same time determines whether the weight meets the preset weight, and outputs the judgment result as yes or no; in some embodiments, the formula for calculating the compensation angle is: Δθ=-θ+0.5°.

[0040] S6. The main control computer sends a command to the plasma return robot arm 5, which grabs the inspected plasma bag on the weighing platform 6 and moves it above the second plasma delivery track 2. The main control computer controls the first plasma delivery track 1 to pause operation. After the second plasma delivery track 2 clears the placement area, the plasma return robot arm 5 releases the plasma bag. After a delay of m seconds, the first plasma delivery track 1 resumes operation.

[0041] In some embodiments, after the first blood plasma bag is weighed and its information is collected, the main control computer immediately sends a return command to the plasma return robotic arm 5 of the robot gripping module. The plasma return robotic arm 5 quickly moves above the weighing platform 6, accurately grips the inspected blood plasma bag through its end effector, and then smoothly rotates it to stand by directly above the second blood plasma delivery track 2. Simultaneously, the main control computer sends a pause command to the first blood plasma delivery track 1, and the track immediately stops running; at the same time, it detects the placement area of ​​the second blood plasma delivery track 2 and sends a feedback signal after confirming that there are no obstructions. The plasma return robotic arm 5 releases the blood plasma bag, and after a 0.5-second delay to ensure that the bag is stably placed, the main control computer controls the first blood plasma delivery track 1 to resume its running speed of 0.3 m / s, and begins the next bag sampling cycle.

[0042] S7. Repeat steps S2-S6 until the preset sampling quantity is completed, and the main control computer generates a complete sampling report.

[0043] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0044] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A device for automatic sampling and data binding of blood plasma bags, characterized in that, The device includes: a conveying module, a counting and positioning module, a robot gripping module, a weighing module, an information acquisition module, and a control module; the conveying module includes a first plasma conveying track (1) and a second plasma conveying track (2) arranged sequentially in the horizontal direction; the counting and positioning module includes a plasma bag counting and positioning device (3), which is fixedly installed above the first plasma conveying track (1) by a bracket; the robot gripping module includes a plasma gripping robot arm (4) and a plasma return robot arm (5), the plasma gripping robot arm (4) is located on one side of the first plasma conveying track (1), and its body is fixed to the outer frame of the first plasma conveying track (1) away from the second plasma conveying track (2), and in the initial state the arm hangs down and is located above the first plasma conveying track (1); the plasma return robot arm (5) is located on one side of the second plasma conveying track (2), and is connected to the plasma... The gripping robot arm (4) is located on the same side, and its body is fixed to the outer frame of the second plasma delivery track (2) away from the first plasma delivery track (1). In the initial state, the arm hangs down and is located directly above the second plasma delivery track (2). The weighing module includes a weighing platform (6), which is set between the plasma gripping robot arm (4) and the plasma return robot arm (5). The information acquisition module includes a camera lens (7), a fill light (8), and a barcode scanner (9), which are set directly above the weighing platform (6). The control module includes a control screen (10), which is embedded in the frame panel on the operating side of the device. The control screen (10) has a built-in main control computer, which is connected to the delivery module, the counting and positioning module, the robot gripping module, the weighing module, and the information acquisition module.

2. The device for automatic sampling and data binding of plasma bags according to claim 1, characterized in that, The conveying module also includes a belt control motor (11) and a belt control encoder (12). The belt control motor (11) is connected to the belt of the first plasma conveying track (1) and the pulley of the second plasma conveying track (2) respectively. The belt control encoder (12) is coaxially connected to the output shaft of the belt control motor (11), and both the belt control motor (11) and the belt control encoder (12) are connected to the main control computer.

3. The device for automatic sampling and data binding of plasma bags according to claim 1, characterized in that, The installation height of the plasma bag counter (3) and the distance between the delivery surface of the first plasma delivery track (1) are 30-50cm. A camera is installed on the counter (3).

4. The device for automatic sampling and data binding of plasma bags according to claim 1, characterized in that, The plasma grabbing robot arm (4) and the plasma return robot arm (5) both include a robot arm body and an end effector, with the end effector facing the corresponding conveying track or weighing platform (6).

5. The device for automatic sampling and data binding of plasma bags according to claim 1, characterized in that, The weighing module also includes a weighing sensor, which is embedded inside the platform (6) of the weighing platform. The sensing area of ​​the weighing sensor coincides with the center of the platform of the weighing module.

6. The device for automatic sampling and data binding of plasma bags according to claim 1, characterized in that, The shooting direction of the camera lens (7), the illumination direction of the fill light (8), and the scanning direction of the barcode scanner (9) are all oriented toward the center of the weighing platform (6) of the weighing module.

7. A method for automated sampling and data binding of blood plasma bags, characterized in that, The method is based on the apparatus of any one of claims 1-6, and the method includes the following steps: S1. Input or confirm the plasma information and sampling parameters of the current batch through the control screen (10) of the control module. The plasma information includes the batch number and specifications. The sampling parameters include the sampling ratio or sampling quantity. The main control computer built into the control screen (10) automatically calculates the grab interval after receiving the parameters and completes the system startup; S2. Place the entire batch of plasma bags to be inspected sequentially at the feed end of the first plasma conveying track (1) of the conveying module. The main control computer sends a start command to the first plasma conveying track (1). The first plasma conveying track (1) drives the plasma bags to move along the conveying direction; When the plasma bag enters the detection area of ​​the plasma bag counting locator (3) of the counting and positioning module, the camera on the counting locator (3) collects the 2D image and 3D point cloud data of the plasma bag; S3. YOLOv8-OBB is used to process the 2D image, outputting the bounding box coordinates, counting signal, attitude tilt angle θ, and category confidence of the plasma bag. At the same time, the 3D point cloud data is denoised and the local plane on the surface of the bag is searched as the adsorption candidate position. All data are synchronously transmitted to the main control computer. S4. When the counting result counted by the main control computer reaches the preset grasping interval, an action command is sent to the plasma grasping robot arm (4), transmitting the attitude tilt angle θ and adsorption candidate position to the plasma grasping robot arm (4). S5. The main control computer calculates the compensation angle and drives the plasma grasping robot arm (4) to rotate to the corresponding compensation angle. The plasma grasping robot arm (4) moves to the top of the first plasma transport track (1), grasps the target plasma bag and transfers it to the center of the weighing platform (6) of the weighing module and then resets. The main control computer controls the supplementary lighting. The light (8) is turned on, the camera lens (7) captures the image of the plasma bag, the barcode scanner (9) scans the plasma bag identification code, the weighing platform (6) built-in weighing sensor acquires the plasma bag weight data, the main control computer automatically binds and stores the identification code, weight data, image, attitude tilt angle θ, point cloud data and timestamp, and at the same time judges whether the weight meets the preset weight, and outputs the judgment result as yes or no; S6. The main control computer sends an instruction to the plasma return robot arm (5), the plasma return robot arm (5) grabs the inspected plasma bag on the weighing platform (6) and moves it above the second plasma transport track (2); the main control computer controls the first plasma transport track (1) to stop running, and after the second plasma transport track (2) clears the placement area, the plasma return robot arm (5) releases the plasma bag; after a delay of m seconds, the first plasma transport track (1) resumes running.

8. The method for automatic sampling and data binding of plasma bags according to claim 7, characterized in that, The specific steps for preprocessing the 2D image using YOLOv8-OBB in step S3 are as follows: S31. Perform Gaussian filtering preprocessing on the 2D image to suppress random noise caused by low-temperature frost and retain the outline features of the blood plasma bag; S32. Adjust the gain of the RGB channel of the image to offset the color gamut shift caused by uneven light reflection in the cold storage environment and unify the color features of the image. S33. Scale the preprocessed image to 640×640 pixels, and use letterbox proportioning to maintain the original aspect ratio of the image to obtain the model input image; S34. Input the model input image into the YOLOv8-OBB feature extraction network, and successively fuse shallow details and deep semantic features through the C2f module and enhance the global feature capture capability through the SPPF module; S35. The classification branch outputs the class confidence in the range of 0~1 using the sigmoid activation function, the regression branch outputs the bounding box coordinates in the pixel coordinate system using CIoU loss, and the pose branch outputs the pose tilt angle θ using MAE loss; S36. Count the number of valid detection boxes to generate a counting signal, and synchronously transmit the bounding box coordinates, counting signal, pose tilt angle θ, and class confidence to the main control computer.

9. The method for automatic sampling and data binding of plasma bags according to claim 7, characterized in that, In step S5, the formula for calculating the compensation angle is: Δθ = -θ + 0.5°.

10. The method for automatic sampling and data binding of plasma bags according to claim 7, characterized in that, The method further includes the following steps: S7. Repeat steps S2-S6 until the preset sampling quantity is completed, and the main control computer generates a complete sampling report.

Images