An automatic chromatographic sampling system and method based on an AI robot

Abstract

The present application relates to the technical field of robot vision control, solves the technical problem of low sample analysis efficiency caused by manual operation of sample feeding disc replacement in the prior art, and particularly relates to an automatic chromatography sample feeding system and method based on an AI robot.The present application comprises a sample feeding disc for placing chromatography vials, and a gripper arranged at the driving end of a mechanical arm.A positioning assembly for accurately installing the sample feeding disc is arranged on a driving shaft in the chromatograph.The positioning assembly comprises an adapter ring arranged on the driving shaft in the chromatograph, and a wedge-shaped groove formed in the sample feeding disc and cooperating with the adapter ring.Two sides of the adapter ring are respectively provided with a positioning groove along the radial direction.Two sides inside the wedge-shaped groove are respectively provided with a limiting slide block fixed with the sample feeding disc.The limiting slide block slides in the positioning groove, and the limiting slide block is made of neodymium iron boron magnet material.A magnetic attraction assembly for fixing the adapter ring and the sample feeding disc is arranged on the inside of the positioning groove.

Description

Technical Field

[0001] This invention relates to the field of robot vision control technology, and in particular to an automated chromatography injection system and method based on an AI robot. Background Technology

[0002] Chromatographic analysis technology is widely used in chemical analysis, environmental monitoring, food safety, and pharmaceutical research and development. The sample injection process is a crucial step in chromatographic analysis, directly affecting the accuracy and reproducibility of the analytical results. Current chromatographic injection systems suffer from the following problems:

[0003] Most existing sample injection trays are fixed, binding them to the chromatographic instrument. Replacement requires manual operation, resulting in limited automation and low efficiency in sample analysis. Furthermore, the robot systems used are complex, with difficulties in robot positioning and robotic arm positioning when grasping samples. The robotic arm relies on dynamic path planning and adaptive algorithms, which are costly, difficult to deploy, and result in slow sample placement. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an automated chromatographic sample injection system and method based on an AI robot. This solves the technical problem that existing technologies require manual operation of the injection tray during replacement, resulting in limited automation of sample analysis. It achieves batch sample loading, automatic detection, and autonomous injection tray retrieval, thereby reducing manual workload.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an automated chromatographic injection system based on an AI robot, comprising an injection tray for placing chromatographic vials and grippers disposed at the drive end of a robotic arm. A positioning component for precisely installing the injection tray is disposed on the drive shaft inside the chromatograph. The positioning component includes an adapter ring sleeved on the drive shaft inside the chromatograph and a wedge-shaped groove on the injection tray that mates with the adapter ring. Positioning grooves are respectively formed radially on both sides of the adapter ring. Limiting sliders fixed to the injection tray are respectively disposed on both sides inside the wedge-shaped groove. The limiting sliders slide in the positioning grooves and are made of neodymium iron boron magnet material. A magnetic suction component for fixing the adapter ring and the injection tray is disposed on the inner side of the positioning groove. An electromagnetic braking component for precisely controlling the braking of the injection tray is disposed on the drive shaft inside the chromatograph.

[0006] Furthermore, the magnetic suction assembly includes a limiting seat disposed at one end of the positioning groove and fixed to the adapter ring, magnetic suction beads disposed at both ends of the limiting seat, and a magnetic suction slider fixed to the limiting slider disposed between the two magnetic suction beads.

[0007] Furthermore, the electromagnetic braking assembly includes a brake disc mounted on the drive shaft, an excitation coil for pressing the drive shaft is provided on the brake disc, an armature assembly for generating axial displacement under the action of a magnetic field is provided at one end of the excitation coil, and a reset spring for pushing the armature assembly to reset is sleeved on the inner side of the excitation coil.

[0008] An automated chromatographic injection method based on an AI robot involves attaching a reflective strip symmetrically to the center line of the upper surface of the injection plate. The method includes the following steps:

[0009] S1. The robot unit grabs the sample injection plate to be tested according to the task scheduling instructions and moves it to the predetermined position in the area where the target chromatograph is located. Then, it acquires real-time images containing the adapter ring, the sample injection plate and the reflective strip.

[0010] S2. Based on the real-time image, take the side line on the adapter ring opposite to the arc line as the standard line, and the edge line of the reflective strip closest to the wedge groove as the reference line, and calculate the relative position data between the reference line and the standard line.

[0011] S3. Based on the standard lengths of the midpoints of the reference line and the standard line, the standard angle between the reference line and the standard line, and the relative position data when the sample inlet plate and the adapter ring are fully installed, calculate the fine-tuning parameters used to adjust the robotic arm to precisely fix the sample inlet plate onto the adapter ring.

[0012] S4. The robotic arm moves the sample tray according to the fine-tuning parameters. The RFID chip sends a second signal to the AI ​​control unit based on the first signal emitted when the Hall sensor is triggered, releases the electromagnetic brake, and calls the chromatography software through the API to start the detection.

[0013] Preferably, step S1 specifically includes the following steps:

[0014] S11. Obtain the sample tray number and target chromatograph through chromatography software. The robot unit moves to the sample tray storage area, verifies the matching sample tray number information through the vision recognition module, and grabs the corresponding sample tray to be tested.

[0015] S12. The robot unit clamps the sample injection disk to be tested and moves it to the predetermined position in the area where the target chromatograph is located.

[0016] S13, the AI ​​control unit sends a position ready signal, and after executing the drive shaft zeroing program in the chromatography control terminal, it starts the electromagnetic braking component and triggers the micro switch, and then sends a braking signal to the AI ​​control unit.

[0017] S14. The AI ​​control unit receives the braking signal, and after the AI ​​control unit moves to the predetermined position, it acquires a real-time image including the adapter ring, the sample inlet plate, and the reflective strip.

[0018] Preferably, step S2 specifically includes the following steps:

[0019] S21. Take the side line on the adaptation ring opposite to the arc line in the real-time image as the standard line, and take the line connecting the midpoint of the reference line and the midpoint of the standard line as the positioning line.

[0020] S22. Mark the midpoint of the standard line as... Point, will The endpoint of the standard line to the right of the point is marked as The midpoint of the reference line is marked as a point. Point, will The endpoint of the reference line to the right of the point is marked as Point, with Establish a three-dimensional coordinate system with point as the center, and... Click to The direction of the point is taken as the positive x-axis. Click to The direction of the point is taken as the positive y-axis, and the normal vector of the upper surface of the fitting ring is taken as the positive z-axis;

[0021] S23. Obtain real-time images Point and The coordinates of the point are given, and the direction vector of the standard line is calculated using the following formula:

[0022]

[0023] In acquiring real-time images Point and The coordinates of the point are used to calculate the direction vector of the reference line. The calculation formula is as follows:

[0024]

[0025] In acquiring real-time images Point and The coordinates of the point are used to calculate the direction vector of the line connecting the points. The calculation formula is as follows:

[0026]

[0027] In the above formula, They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The coordinates of a point along the x-axis, y-axis, and z-axis;

[0028] S24. Based on the real-time image, project the reference line perpendicularly onto the XOZ plane of the three-dimensional coordinate system, and calculate the first angle between the projected reference line and the standard line; the formula for calculating the first angle is:

[0029]

[0030] In the above formula, This represents the first angle between the projection line of the reference line and the standard line in the real-time image. The direction vector of the standard line in the real-time image. This is the direction vector of the projection line of the reference line in the real-time image;

[0031] S25. Based on the real-time image, project the positioning line perpendicularly onto the XOY plane of the three-dimensional coordinate system, and calculate the second angle between the projected line of the positioning line and the standard line. The formula for calculating the second angle is:

[0032]

[0033] In the above formula, This represents the second angle between the projection line of the positioning line in the real-time image and the standard line. The projection direction vector of the connection line located in the real-time image;

[0034] S26. Calculate the actual length of the projection line of the positioning connection based on the real-time image. The calculation formula is:

[0035]

[0036] In the above formula, This represents the actual length of the projected line of the positioning connection in the real-time image. coordinate point The coordinates of the vertical projection point on the XOY plane of the three-dimensional coordinate system;

[0037] S27. Based on the real-time image, obtain the height difference between the midpoint of the standard line and the midpoint of the reference line. The calculation formula is as follows:

[0038]

[0039] In the above formula, This represents the height difference between the line connecting the midpoint of the reference line and the midpoint of the standard line in a real-time image. This represents the coordinates of point A on the z-axis. This represents the coordinates of point B on the z-axis;

[0040] S28. Mark the first included angle, the second included angle, the actual length, and the height difference as the relative position data of the reference line and the standard line.

[0041] Preferably, step S21 specifically includes the following steps:

[0042] S211. Obtain several contour lines on the adaptation ring in the real-time image and construct a contour line set containing several contour lines.

[0043] S212. Extend the coordinates of each contour line in the contour line set to the surrounding four directions by several pixels to form an extension area. Using the extension area as a constraint, search for pixels whose gray values ​​are greater than the gray value change threshold of their neighboring pixels within the extension area, and mark them as feature points. Remove the remaining pixels.

[0044] S213. Divide each contour line into 12 sub-regions centered on the feature points, calculate the sum of pixel gray values ​​in each sub-region, and perform normalization processing.

[0045] S214. Classify the 12 sub-regions into first sub-region, middle sub-region, and last sub-region, and calculate the grayscale difference value of each sub-region based on the sum of the normalized pixel values ​​of each sub-region. ,

[0046] The formula for calculating the grayscale difference value of the first terminal area is:

[0047]

[0048] The formula for calculating the gray-level difference value of the intermediate sub-region is:

[0049]

[0050] The formula for calculating the gray-level difference value of the terminal sub-region is:

[0051]

[0052] In the above formula, Indicates the first Sub-regions Indicates the first The sum of the gray values ​​of all pixels within the sub-region. This represents the grayscale difference value of each sub-region;

[0053] S215. Based on the grayscale difference values ​​of the first terminal region, the middle sub-region, and the last sub-region, locate and extract the candidate contour lines;

[0054] S216. Select the standard line based on the cosine similarity and Euclidean distance between each candidate contour line and the standard contour line.

[0055] Preferably, step S3 specifically includes the following steps:

[0056] S31, Judgment Is it greater than ,

[0057] If so, drive the robotic arm to rotate the sample tray clockwise around point B in a direction parallel to the y-axis by the first included angle. ;

[0058] If not, drive the robotic arm to rotate the sample tray counterclockwise around point B in a direction parallel to the y-axis by the first included angle. ;

[0059] S32, Judgment Is it greater than ,

[0060] If so, the robotic arm drives the sample tray to move along the positive z-axis, moving the height difference. ,

[0061] If not, the robotic arm will drive the sample tray to move in the opposite direction along the z-axis to move the height difference. ;

[0062] S33. Determine whether the second included angle is greater than 90 degrees.

[0063] If so, drive the robotic arm to rotate the sample tray counterclockwise around the midpoint of the standard line. ;

[0064] If not, the robotic arm will drive the sample tray to rotate clockwise around the midpoint of the standard line. ;

[0065] S34. Based on the fully installed state of the injection tray and adapter ring, obtain the length of the standard positioning line as follows: Based on the calculated length of the positioning line The robotic arm drives the sample tray to move along the positioning line toward the adapter ring. The length.

[0066] Preferably, after the chromatography software starts and the detection is completed, the AI ​​control unit controls the robotic arm to disassemble the injection tray along a fixed trajectory, and the robot unit delivers the injection tray to the recovery area along a predetermined path and updates the task status to the chromatography software.

[0067] By employing the above technical solution, the present invention provides an automated chromatographic injection system and method based on an AI robot, which has at least the following beneficial effects:

[0068] 1. This invention uses a positioning component to precisely install the injection plate onto the adapter ring, and a magnetic suction component further secures the adapter ring and the injection plate. It employs a dual-fixing structure of mechanical cooperation and magnetic induction between the limiting slider and the positioning groove. After the limiting slider of the injection plate slides into the positioning groove of the adapter ring, it forms a physical lock, which helps to achieve precise positioning, ensures the accuracy of injection, strengthens the mechanical fixation stability, reduces experimental interference, improves the efficiency of automatic chromatography injection, shortens the sample placement time, and reduces human intervention and operational errors.

[0069] 2. This invention uses an electromagnetic braking component to precisely control the braking of the sample inlet disc. The neodymium iron boron magnet and the Hall element work together to accurately determine the installation status of the sample inlet disc, preventing sample spillage due to lack of locking. The electromagnetic brake provides real-time feedback on the braking status through a micro switch, avoiding equipment collisions caused by drive shaft slippage. At the same time, the AI ​​control unit monitors abnormal signals in each stage in real time, which can trigger a shutdown warning in a timely manner, reducing the risk of equipment damage and experimental accidents.

[0070] 3. This invention achieves high-precision positioning of the sample tray from grasping to installation through the synergy of multi-dimensional visual recognition and mechanical control. The visual recognition module scans the QR code to verify the sample information, avoiding the risk of mis-picking and ensuring the accuracy of the operation object. The combination of reflective strip visual detection and Hall element sensing greatly improves the installation alignment accuracy of the sample tray. Through multiple safety mechanisms, the operational risks and losses are reduced. Attached Figure Description

[0071] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0072] Figure 1 This is a schematic diagram of the structure of the robotic arm of the present invention;

[0073] Figure 2 This is a schematic diagram of the positioning component of the present invention;

[0074] Figure 3 This is a schematic diagram of the adapter ring structure of the present invention;

[0075] Figure 4 This is a schematic diagram of the magnetic suction assembly of the present invention;

[0076] Figure 5 This is a schematic diagram of the electromagnetic braking assembly of the present invention;

[0077] Figure 6 This is a flowchart of the automated chromatographic injection method of the present invention;

[0078] Figure 7This is a schematic diagram of the standard line, positioning line, and reference line of the automated chromatographic injection method of the present invention.

[0079] Figure 8 This is a three-dimensional coordinate system schematic diagram of the automated chromatographic injection method of the present invention.

[0080] In the diagram: 1. Sample inlet tray; 2. Robotic arm; 3. Gripper;

[0081] 4. Positioning component; 41. Adapter ring; 42. Wedge groove; 43. Positioning groove; 44. Limiting slider;

[0082] 45. Magnetic suction assembly; 451. Limiting seat; 452. Magnetic ball catch; 453. Magnetic slider;

[0083] 46. ​​Electromagnetic braking assembly; 461. Brake disc; 462. Excitation coil; 463. Armature assembly; 464. Return spring. Detailed Implementation

[0084] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0085] Traditionally, the process involves manually carrying the chromatographic vials containing the samples to the chromatograph area and placing them into the corresponding injection tray 1. In scenarios with large sample volumes and numerous chromatographic instruments, this method is inefficient and prone to errors. To reduce manual operation, improve the automation of sample analysis, and enhance efficiency through AI scheduling and decision-making, the following approach is adopted: Figure 1 - Figure 5 As shown, this invention provides an automated chromatographic injection system based on an AI robot, comprising an injection disk 1, a robotic arm 2, grippers 3, and a positioning component 4. The top surface of the injection disk 1 has multiple slots for placing chromatographic vials. The grippers 3 are mounted on the drive end of the robotic arm 2 to hold the injection disk 1. The positioning component 4 is mounted on the drive shaft of the chromatograph to position the injection disk 1. The drive shaft is a traditional component of the chromatograph, its function being to drive the injection disk 1 to rotate to different positions. The system then controls the injection needle above the injection disk 1 to draw samples from different chromatographic vials within the injection disk 1 for detection. The injection disk 1 is then held in place by the grippers 3 on the drive end of the robotic arm 2 and inserted into the chromatograph. The positioning component 4 on the drive shaft automatically and precisely fixes the injection disk 1, and then chromatographic analysis is performed.

[0086] Reflective strips are attached symmetrically to the center line of the upper surface of the sample inlet tray 1 using 3M adhesive to assist in robot visual positioning; waterproof QR code labels containing instrument serial number information are attached to the outer surface of the sample inlet tray 1.

[0087] To achieve high-precision positioning of the injection disk 1 throughout the entire process from gripping to installation, and to reduce manual intervention, the positioning component 4 specifically includes an adapter ring 41 fitted onto the drive shaft inside the chromatograph. The adapter ring 41 is made of aluminum alloy, and its inner diameter matches the outer diameter of the drive shaft with a tolerance of ±0.05mm. A wedge-shaped groove 42 is formed on the injection disk 1 to mate with the adapter ring 41, which also ensures that the robotic arm 2 maintains a fixed direction when gripping. By moving the injection disk 1, the wedge-shaped groove 42 of the injection disk 1 engages with the adapter ring 41, thus ensuring that the injection disk 1 and the adapter ring 41 are aligned. The fitting ring 41 forms a disc. Positioning grooves 43 are respectively provided radially on both sides of the fitting ring 41. Limiting sliders 44 fixed to the sample injection disc 1 are respectively provided on both sides inside the wedge-shaped groove 42. The limiting sliders 44 slide in the positioning grooves 43 and are made of neodymium iron boron magnet material. When the sample injection disc 1 approaches the fitting ring 41, the limiting sliders 44 are inserted into the positioning grooves 43 and slide in them. When the limiting sliders 44 move to the end of the positioning grooves 43, the wedge-shaped grooves 42 of the sample injection disc 1 and the fitting ring 41 are engaged with each other.

[0088] To improve the stability of the mounting ring 41 of the sample tray 1 and prevent sample spillage due to loose locking, a magnetic suction assembly 45 for fixing the mounting ring 41 and the sample tray 1 is specifically provided on the inner side of the positioning groove 43. The magnetic suction assembly 45 includes a limiting seat 451 fixed at one end of the positioning groove 43 and fixed to the mounting ring 41, magnetic beads 452 at both ends of the limiting seat 451, and a magnetic slider 453 fixed to the limiting slider 44 between the two magnetic beads 452. A Hall element and an RFID chip are also installed at the magnetic beads 452 at the bottom of the positioning groove 43 for signal transmission when the sample tray 1 is installed. The neodymium iron boron magnet and the Hall element work together to accurately determine the installation status of the sample tray 1. At the same time, the AI ​​control unit monitors the abnormal signals of each link in real time and can trigger a shutdown warning in time to reduce the risk of equipment damage and experimental accidents.

[0089] To provide real-time feedback on braking status and prevent equipment collisions caused by drive shaft slippage, an electromagnetic braking assembly 46 is specifically installed on the drive shaft inside the chromatograph for precise control of the injection disc 1's braking. The electromagnetic braking assembly 46 includes a brake disc 461 mounted on the drive shaft inside the chromatograph. The surface of the brake disc 461 is coated with a ceramic layer. An excitation coil 462 is installed on the brake disc 461 to press against the drive shaft. An armature assembly 463 is installed at one end of the excitation coil 462 to generate axial displacement under the influence of a magnetic field. A return spring 464 is sleeved inside the excitation coil 462 to push the armature assembly 463 back to its original position. When the excitation coil 462 is energized, it generates a magnetic field that attracts the armature assembly 463, driving the brake disc 461 to press against the drive shaft. Simultaneously, the compression causes elastic deformation, compressing the armature assembly 463. When the excitation coil 462 is de-energized, it pushes the armature assembly 463 back to its original position, releasing the brake disc 461 and loosening the drive shaft.

[0090] This invention also proposes an automated chromatographic injection method based on an AI robot, achieving fully unmanned operation from sample tray grabbing, path navigation, precise alignment to installation and fixation. The AI ​​control unit quickly generates robotic arm adjustment instructions by dynamically comparing pre-stored standard parameters with real-time visual feedback, significantly saving preparation time for each injection. Simultaneously, automated disassembly, retrieval, and task status synchronization after detection further reduce waiting time between processes, greatly improving the overall efficiency of automated chromatographic injection. Figure 6 As shown, the method includes the following steps:

[0091] S1. Based on the registration information of the chromatography software, the robot unit starts the detection task scheduling autonomously, obtains the QR code information of the sample injection tray 1 to be tested, matches the instrument number and the target chromatograph, moves the robot unit along the preset path to the sample injection tray 1 storage area, and the vision recognition module on the robot unit scans the QR code of the sample injection tray 1. After verifying the information matching, it ensures that the operation object is accurate and controls the pneumatic gripper 3 of the robotic arm 2 to grab the corresponding sample injection tray 1.

[0092] The robot unit uses a visual recognition module to identify laboratory environmental markers and AI visual recognition to avoid obstacles. The robot unit holds the sample tray 1 to be tested and moves it to the target chromatograph along a preset path. The AI ​​control unit scans the chromatograph's QR code through the visual recognition module to confirm again. This dual recognition helps to avoid the risk of mistakenly taking the sample tray 1.

[0093] The AI ​​control unit on the robot unit sends a position ready signal to the chromatography control terminal, which is the chromatograph. The chromatograph executes the drive shaft zeroing program. After the AI ​​control unit recognizes the signal, it sends a start command to the electromagnetic braking assembly 46. The excitation coil 462 is energized, and the adsorption armature assembly 463 presses the drive shaft through the brake disc 461 and triggers the micro switch. The chromatography control terminal sends a braking signal to the AI ​​control unit.

[0094] The AI ​​control unit receives the braking signal and confirms the braking signal. The visual recognition module detects that the reflective strip has moved to the predetermined position, so that the adapter ring 41, the reflective strip and the sample inlet plate 1 are all in the same image, thereby obtaining the current real-time image.

[0095] S2. Based on the real-time image, using the side line on the adapter ring 41 opposite to the arc line as the standard line and the edge line of the reflective strip closest to the wedge groove 42 as the reference line, calculate the relative position data between the reference line and the standard line. The standard line provides a clear and unique spatial reference for the positioning of the sample inlet plate 1. The position of this side line is accurately captured by the visual recognition module, and this is used as the coordinate origin. The fixed posture of the sample inlet plate 1 is calibrated by the reflective strip to ensure that the installation position of the sample inlet plate 1 of different batches and specifications on the adapter ring 41 is more accurate, so that the spatial coordinates of the sample inlet plate 1 can be kept highly consistent each time it is fixed, laying a stable positional foundation for the subsequent operation of the robotic arm 2. The specific steps include the following:

[0096] S21. Based on the real-time image, using the line connecting the midpoint of the reference line and the midpoint of the standard line as the positioning line, and the side line on the adapter ring 41 opposite to the arc line as the standard line, the core of moving the reflective strip to the predetermined position is to achieve reference alignment with the standard line. The standard line serves as a fixed spatial reference on the adapter ring, ensuring the accuracy and uniqueness of the standard line identification. Specifically, this includes the following steps:

[0097] S211. Obtain several contour lines on the adaptation ring 41 in the real-time image and construct a contour line set containing several contour lines.

[0098] S212. Extend the coordinates of each contour line in the contour line set to the surrounding four directions by several pixels to form an extension area. Using the extension area as a constraint, search for pixels whose gray values ​​are greater than the gray value change threshold of their neighboring pixels within the extension area, and mark them as feature points. Remove the remaining pixels.

[0099] S213. Divide each contour line into 12 sub-regions centered on the feature points, calculate the sum of pixel gray values ​​in each sub-region, and perform normalization processing.

[0100] S214. Number the 12 sub-regions sequentially, for example, from left to right or from top to bottom. Then classify them into first sub-region, middle sub-region, and last sub-region. Calculate the grayscale difference value of each sub-region based on the sum of the normalized pixel values. ,

[0101] The formula for calculating the grayscale difference value of the first terminal region, i.e., the first sub-region, is as follows:

[0102]

[0103] The formula for calculating the grayscale difference value of the middle sub-regions, namely the second to eleventh sub-regions, is as follows:

[0104]

[0105] The formula for calculating the grayscale difference value of the twelfth sub-region (the terminal sub-region) is as follows:

[0106]

[0107] In the above formula, Indicates the first Sub-regions This represents the sum of the grayscale values ​​of all pixels within the first sub-region. This represents the sum of the grayscale values ​​of all pixels within the second sub-region. This represents the sum of the grayscale values ​​of all pixels within the eleventh sub-region. This represents the sum of the grayscale values ​​of all pixels within the twelfth sub-region. Indicates the first The sum of the gray values ​​of all pixels within each sub-region, Indicates the first Gray-level difference values ​​of each sub-region;

[0108] S215. Based on the grayscale difference values ​​of the first terminal region, the middle sub-region, and the last sub-region, the grayscale difference values ​​quantify the intensity of grayscale changes between sub-regions. Analyze the distribution pattern and spatial correlation of these grayscale difference values. For example, if the grayscale difference values ​​of the first terminal region, the middle sub-region, and the last sub-region are all within a certain threshold range, then the feature points of the first terminal region, the middle sub-region, and the last sub-region are constructed as points on the candidate contour line. Otherwise, they are not points on the candidate contour line. Thus, several feature points on the candidate contour line are selected. A connection can be established between these selected adjacent feature points. By connecting the beginning and end of each connection line, several candidate contour lines can be located and extracted.

[0109] S216. Calculate the cosine similarity between the feature points on the candidate contour line and each feature point on the preset standard contour line in the image library. The cosine similarity can be calculated by calculating the cosine value of the angle between the normal vectors of the two feature points on the corresponding contour lines. For feature points whose cosine similarity is within the similarity threshold, further calculate the Euclidean distance between the two feature points. If the Euclidean distance between the two feature points is within the distance threshold, mark the two feature points as successfully matched. Finally, count the number of feature points that are successfully matched between the candidate contour line and the standard contour line, and select the candidate contour line with the most successfully matched feature points and the smallest Euclidean distance as the standard line.

[0110] S22. Mark the midpoint of the standard line as... Point, will The endpoint of the standard line to the right of the point is marked as The midpoint of the reference line is marked as a point. Point, will The endpoint of the reference line to the right of the point is marked as Point, with Establish a three-dimensional coordinate system with point as the center, and... Click to The direction of the point is taken as the positive x-axis. Click to The direction of the point is taken as the positive y-axis, and the normal vector of the upper surface of the fitting ring 41 is taken as the positive z-axis, passing through the origin. The direction vector of a point that is perpendicular to the XOY plane is the normal vector;

[0111] S23. Obtain real-time images Point and The coordinates of the point are given, and the direction vector of the standard line is calculated using the following formula:

[0112]

[0113] In acquiring real-time images Point and The coordinates of the point are used to calculate the direction vector of the reference line. The calculation formula is as follows:

[0114]

[0115] In acquiring real-time images Point and The coordinates of the point are used to calculate the direction vector of the line connecting the points. The calculation formula is as follows:

[0116]

[0117] In the above formula, They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The coordinates of a point along the x-axis, y-axis, and z-axis;

[0118] S24. Project the reference line perpendicularly onto the XOZ plane of the three-dimensional coordinate system, and calculate the first angle between the projected reference line and the standard line; the formula for calculating the first angle is:

[0119]

[0120] In the above formula, This represents the first angle between the projection line of the reference line and the standard line in the real-time image. The direction vector of the standard line in the real-time image. This is the direction vector of the projection line of the reference line in the real-time image;

[0121] S25. Project the positioning line perpendicularly onto the XOY plane of the three-dimensional coordinate system, and calculate the second angle between the projected line of the positioning line and the standard line. The formula for calculating the second angle is:

[0122]

[0123] In the above formula, This represents the second angle between the projection line of the positioning line in the real-time image and the standard line. The projection direction vector of the connection line located in the real-time image;

[0124] S26. Calculate the actual length of the projection line of the positioning connection. The calculation formula is:

[0125]

[0126] In the above formula, This represents the actual length of the projected line of the positioning connection in the real-time image. coordinate point The coordinates of the vertical projection point on the XOY plane of the three-dimensional coordinate system;

[0127] S27. Obtain the height difference between the midpoint of the standard line and the midpoint of the reference line. The calculation formula is:

[0128]

[0129] In the above formula, This represents the height difference between the line connecting the midpoint of the reference line and the midpoint of the standard line in a real-time image. This represents the coordinates of point A on the z-axis. This represents the coordinates of point B on the z-axis;

[0130] S28. Mark the first included angle, the second included angle, the actual length, and the height difference as the relative position data of the reference line and the standard line.

[0131] S3. Based on the standard lengths of the midpoints of the reference line and the standard line, the standard angle between the reference line and the standard line, and the relative position data when the injection plate 1 and the adapter ring 41 are fully installed, calculate and adjust the fine-tuning parameters for precisely fixing the injection plate to the adapter ring using the robotic arm 2. The spatial orientation of the injection plate is calibrated to match the state based on the standard line of the adapter ring, ensuring that the injection operation of the robot unit can accurately correspond to the position and meet the requirements of chromatographic analysis for injection accuracy and high efficiency. The specific steps include the following:

[0132] S31, Judgment Is it greater than ,

[0133] If so, drive the robotic arm 2 to rotate the sample tray 1 clockwise around point B in a direction parallel to the y-axis by a first included angle. ;

[0134] If not, the robotic arm 2 will drive the sample tray 1 to rotate counterclockwise around point B in a direction parallel to the y-axis by a first included angle. ;

[0135] S32, Judgment Is it greater than ,

[0136] If so, the robotic arm 2 drives the sample tray 1 to move along the positive z-axis, moving the height difference. ,

[0137] If not, the robotic arm 2 will drive the sample tray 1 to move in the opposite direction along the z-axis to move the height difference. ;

[0138] S33. Determine whether the second included angle is greater than 90 degrees.

[0139] If so, drive the robotic arm 2 to rotate the sample feed plate 1 counterclockwise around the midpoint of the standard line. ;

[0140] If not, the robotic arm 2 will drive the sample tray 1 to rotate clockwise around the midpoint of the standard line. ;

[0141] S34. Based on the fully installed state of the sample inlet plate 1 and the adapter ring 41, obtain the length of the standard positioning connection line as follows: Based on the calculated length of the positioning line The robotic arm 2 drives the sample feeding plate 1 to move along the positioning line towards the adapter ring 41. The length.

[0142] S4. The robot unit moves the sample tray 1 to ensure that the positioning reference is without deviation. The pre-detection function of the Hall element confirms the magnetic field threshold range of its sensing area, providing a judgment standard for subsequent magnet positioning detection. The limiting slider 44 slides into the positioning groove 43 on the side of the adapter ring 41. When the limiting slider 44 contacts the edge of the entrance of the positioning groove 43, the contact force detected by the force sensor reaches the preset threshold. The AI ​​control unit immediately reduces the moving speed and guides the parallelism between the edge of the limiting slider 44 and the edge of the positioning groove 43 through vision, ensuring that the limiting slider 44 slides in along the axial direction of the positioning groove 43.

[0143] As the limiting slider 44 continues to slide into the positioning groove 43, the AI ​​control unit issues a deceleration command until the limiting slider 44 is fully embedded in the positioning groove 43. The magnetic bead 452 on the inner side of the positioning groove 43 clamps the magnetic slider 453, thereby mechanically locking the limiting slider 44 to prevent the slider from accidentally coming out during movement. The locking state is fed back to the control unit through the micro switch at the end of the buckle.

[0144] When the neodymium iron boron magnet enters the sensing area of ​​the Hall element, the Hall element sends a high-level signal to the AI ​​control unit after the magnet is in place. At the same time, the visual recognition module captures an image of the fit between the adapter ring 41 and the sample injection plate 1. The image analysis confirms that there is no obvious offset between the two.

[0145] S8, once the injection tray 1 is fully installed, the Hall sensor is triggered, the RFID chip sends a signal to the AI ​​control unit, the control unit releases the electromagnetic brake, and the chromatography software is called via API to start the detection.

[0146] After the chromatography software starts and the detection is completed, the AI ​​control unit controls the robotic arm 2 to disassemble the injection tray 1 along a fixed trajectory. The trajectory parameters are optimized by AI algorithms and pre-stored to ensure that the force, angle, and speed of each disassembly action are consistent, avoiding damage to the injection tray 1 or spillage of residual sample due to uneven manual operation. At the same time, the fixed trajectory design avoids the precision components at the edge of the adapter ring 41, reducing the risk of mechanical collision and protecting the core components of the chromatograph from damage.

[0147] The robot unit grasps the sample injection tray 1 to be tested along a predetermined path. Combined with AI visual obstacle avoidance function, it can dynamically avoid obstacles in the transfer path and move to the predetermined position in the area where the target chromatograph is located. Then, it acquires real-time images including the adapter ring 41, the sample injection tray 1 and the reflective strip. Based on the real-time images, it calculates the relative position data. Based on the standard length of the midpoint of the reference line and the midpoint of the standard line, the standard angle between the reference line and the standard line and the relative position data when the sample injection tray 1 and the adapter ring 41 are fully installed, it calculates the fine-tuning parameters. The robotic arm 2 moves the sample injection tray 1 according to the fine-tuning parameters. The RFID chip sends a second signal to the AI ​​control unit according to the first signal emitted when the Hall sensor is triggered, releases the electromagnetic brake, and calls the chromatography software through the API to start the detection.

[0148] The robot unit delivers sample tray 1 to the recovery area along a predetermined path and updates the task status to the chromatography software, making the connection time from the end of the detection to the recovery of sample tray 1 faster. This greatly improves efficiency compared to traditional manual operation, effectively reduces the manual waiting links in the experimental process, and realizes fully automated closed-loop management from sample injection to recovery.

[0149] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0150] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0151] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An automated chromatographic sample injection system based on an AI robot, comprising an injection tray (1) for placing chromatographic vials, and grippers (3) disposed at the drive end of a robotic arm (2), characterized in that, The drive shaft inside the chromatograph is equipped with a positioning component (4) for accurately installing the injection plate (1). The positioning component (4) includes an adapter ring (41) fitted on the drive shaft inside the chromatograph and a wedge-shaped groove (42) formed on the injection plate (1) to cooperate with the adapter ring (41). The adapter ring (41) has positioning grooves (43) formed on both sides radially. The wedge-shaped groove (42) has a limiting slider (44) fixed to the injection plate (1) on both sides. The limiting slider (44) slides in the positioning groove (43) and is made of neodymium iron boron magnet material. The positioning groove (43) has a magnetic suction component (45) for fixing the adapter ring (41) and the injection plate (1) on the inner side. The drive shaft inside the chromatograph has an electromagnetic braking component (46) for precisely controlling the braking of the injection plate (1). The magnetic suction assembly (45) includes a limiting seat (451) disposed at one end of the positioning groove (43) and fixed to the adapter ring (41), magnetic suction beads (452) respectively disposed at both ends of the limiting seat (451), and a magnetic suction slider (453) fixed to the limiting slider (44) disposed between the two magnetic suction beads (452). The electromagnetic braking assembly (46) includes a brake disc (461) mounted on a drive shaft. An excitation coil (462) for pressing the drive shaft is provided on the brake disc (461). An armature assembly (463) for generating axial displacement under the action of a magnetic field is provided at one end of the excitation coil (462). A reset spring (464) for pushing the armature assembly (463) to reset is sleeved on the inner side of the excitation coil (462).

2. A method for use in the automated chromatographic sample injection system of claim 1, wherein a reflective strip is affixed symmetrically to the center line of the upper surface of the sample injection plate, characterized in that, The method includes the following steps: S1. The robot unit grabs the sample injection plate to be tested according to the task scheduling instructions and moves it to the predetermined position in the area where the target chromatograph is located. Then, it acquires real-time images containing the adapter ring, the sample injection plate and the reflective strip. S2. Using the side line on the adaptation ring opposite to the arc line in the real-time image as the standard line, and the edge line of the reflective strip closest to the wedge groove as the reference line, calculate the relative position data between the reference line and the standard line. S3. Based on the standard lengths of the midpoints of the reference line and the standard line, the standard angle between the reference line and the standard line, and the relative position data when the sample inlet plate and the adapter ring are fully installed, calculate the fine-tuning parameters used to adjust the robotic arm to precisely fix the sample inlet plate onto the adapter ring. S4. The robotic arm moves the sample tray according to the fine-tuning parameters. The RFID chip sends a second signal to the AI ​​control unit based on the first signal emitted when the Hall sensor is triggered, releases the electromagnetic brake, and calls the chromatography software through the API to start the detection.

3. The chromatographic injection method according to claim 2, characterized in that, Step S1 specifically includes the following steps: S11. Obtain the sample tray number and target chromatograph through chromatography software. The robot unit moves to the sample tray storage area, verifies the matching sample tray number information through the vision recognition module, and grabs the corresponding sample tray to be tested. S12. The robot unit clamps the sample injection disk to be tested and moves it to the predetermined position in the area where the target chromatograph is located. S13, the AI ​​control unit sends a position ready signal, and after executing the drive shaft zeroing program in the chromatography control terminal, it starts the electromagnetic braking component and triggers the micro switch, and then sends a braking signal to the AI ​​control unit. S14. The AI ​​control unit receives the braking signal, and after the AI ​​control unit moves to the predetermined position, it acquires a real-time image including the adapter ring, the sample inlet plate, and the reflective strip.

4. The chromatographic injection method according to claim 2, characterized in that, Step S2 specifically includes the following steps: S21. Take the side line on the adaptation ring opposite to the arc line in the real-time image as the standard line, and take the line connecting the midpoint of the reference line and the midpoint of the standard line as the positioning line. S22. Mark the midpoint of the standard line as... Point, will The endpoint of the standard line to the right of the point is marked as The midpoint of the reference line is marked as a point. Point, will The endpoint of the reference line to the right of the point is marked as Point, with Establish a three-dimensional coordinate system with point as the center, and... Click to The direction of the point is taken as the positive x-axis. Click to The direction of the point is taken as the positive y-axis, and the normal vector of the upper surface of the fitting ring is taken as the positive z-axis; S23. Obtain real-time images Point and The direction vector of the standard line is calculated from the coordinates of a point using the following formula: In acquiring real-time images Point and The direction vector of the reference line is calculated from the coordinates of a point using the following formula: In acquiring real-time images Point and The direction vector of the line connecting the coordinates of a point is calculated using the following formula: In the above formula, They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The x-axis, y-axis, and z-axis coordinates of a point. They are respectively The coordinates of a point along the x-axis, y-axis, and z-axis; S24. Project the reference line perpendicularly onto the XOZ plane of the three-dimensional coordinate system, and calculate the first angle between the projected reference line and the standard line; the formula for calculating the first angle is: In the above formula, This represents the first angle between the projection line of the reference line and the standard line in the real-time image. The direction vector of the standard line in the real-time image. This is the direction vector of the projection line of the reference line in the real-time image; S25. Project the positioning line perpendicularly onto the XOY plane of the three-dimensional coordinate system, and calculate the second angle between the projected line of the positioning line and the standard line. The formula for calculating the second angle is: In the above formula, This represents the second angle between the projection line of the positioning line in the real-time image and the standard line. The projection line direction vector of the location connection in the real-time image; S26. Calculate the actual length of the projection line of the positioning connection. The calculation formula is: In the above formula, This represents the actual length of the projected line of the positioning connection in the real-time image. coordinate point The coordinates of the vertical projection point on the XOY plane of the three-dimensional coordinate system; S27. Obtain the height difference between the midpoint of the standard line and the midpoint of the reference line. The calculation formula is: In the above formula, This represents the height difference between the line connecting the midpoint of the reference line and the midpoint of the standard line in a real-time image. This represents the coordinates of point A on the z-axis. This represents the coordinates of point B on the z-axis; S28. Mark the first included angle, the second included angle, the actual length, and the height difference as the relative position data of the reference line and the standard line.

5. The chromatographic injection method according to claim 4, characterized in that, Step S21 specifically includes the following steps: S211. Obtain several contour lines on the adaptation ring in the real-time image and construct a contour line set containing several contour lines. S212. Extend the coordinates of each contour line in the contour line set to the surrounding four directions by several pixels to form an extension area. Using the extension area as a constraint, search for pixels whose gray values ​​are greater than the gray value change threshold of their neighboring pixels within the extension area, and mark them as feature points. Remove the remaining pixels. S213. Divide each contour line into 12 sub-regions centered on the feature points, calculate the sum of pixel gray values ​​in each sub-region, and perform normalization processing. S214. Classify the 12 sub-regions into first sub-region, middle sub-region, and last sub-region, and calculate the grayscale difference value of each sub-region based on the sum of the normalized pixel values ​​of each sub-region. , The formula for calculating the grayscale difference value of the first terminal area is: The formula for calculating the gray-level difference value of the intermediate sub-region is: The formula for calculating the gray-level difference value of the terminal sub-region is: In the above formula, Indicates the first Sub-regions Indicates the first The sum of the gray values ​​of all pixels within the sub-region. This represents the grayscale difference value of each sub-region; S215. Based on the grayscale difference values ​​of the first terminal region, the middle sub-region, and the last sub-region, locate and extract the candidate contour lines; S216. Select the standard line based on the cosine similarity and Euclidean distance between each candidate contour line and the standard contour line.

6. The chromatographic injection method according to claim 4, characterized in that, Step S3 specifically includes the following steps: S31, Judgment Is it greater than , If so, drive the robotic arm to rotate the sample tray clockwise around point B in a direction parallel to the y-axis by the first included angle. ; If not, drive the robotic arm to rotate the sample tray counterclockwise around point B in a direction parallel to the y-axis by the first included angle. ; S32, Judgment Is it greater than , If so, the robotic arm drives the sample tray to move along the positive z-axis, moving the height difference. , If not, the robotic arm will drive the sample tray to move in the opposite direction along the z-axis to move the height difference. ; S33. Determine whether the second included angle is greater than 90 degrees. If so, drive the robotic arm to rotate the sample tray counterclockwise around the midpoint of the standard line. ; If not, the robotic arm will drive the sample tray to rotate clockwise around the midpoint of the standard line. ; S34. Based on the fully installed state of the injection tray and adapter ring, obtain the length of the standard positioning line as follows: Based on the calculated length of the positioning line The robotic arm drives the sample tray to move along the positioning line toward the adapter ring. The length.

7. The chromatographic injection method according to claim 2, characterized in that, After the chromatography software starts and the detection is completed, the AI ​​control unit controls the robotic arm to disassemble the injection tray along a fixed trajectory. The robot unit then delivers the injection tray to the recovery area along a predetermined path and updates the task status to the chromatography software.

Images