Dual-arm robot powder grinding system and method for automated chemical laboratories

Abstract

The application discloses a double-arm robot powder grinding system and method for an automated chemical laboratory, and belongs to the field of grinding robots. The system comprises the following: a mortar is fixed to a grinding platform; a grinding pestle is installed on the end of a grinding main arm of a double-arm robot through an elastic clamp; a six-dimensional torque sensor is installed on the end of a grinding slave arm of the double-arm robot and is connected to a grinding gripper, and the six-dimensional torque sensor can clamp the mortar and a soft spatula through the grinding gripper; an RGBD camera is arranged on the grinding main arm of the double-arm robot and can shoot and collect an RGB image and a depth map of the grinding platform; a powder screening device can screen the grinding powder in the mortar; and a main controller is in communication connection with the RGBD camera and the double-arm robot respectively and can control the double-arm robot to perform powder grinding, gathering and screening until a grinding task is completed. The system can control the double-arm robot to realize stable and efficient grinding tasks, improve grinding efficiency and guarantee expansion capability.

Description

Technical Field

[0001] This invention relates to the field of grinding robots for powder preparation, and more particularly to a dual-arm robotic powder grinding system and method for automated chemical laboratories. Background Technology

[0002] In chemical laboratories, it is often necessary to grind solid samples into fine powders. Various powder grinding devices for automated experiments are available on the market, including ball mills and automated grinding bowls. These devices are highly efficient for rapid powder grinding, but are primarily designed for large-scale experiments and are not suitable for small-batch powder preparation. Therefore, manual grinding remains a common method of powder preparation in laboratories. However, manual grinding is not only time-consuming and labor-intensive, but also poses potential safety risks for the preparation of toxic or corrosive powders. Recent research has explored the use of soft molds mounted on the end effector of robots, combined with visual and auditory feedback, to achieve effective powder grinding and timely collection. While this grinding system performs excellently in sample preparation and improving the reproducibility of mechanical synthesis reactions, its visual method cannot detect color changes during powder grinding, which may lead to positioning errors. Furthermore, this grinding system relies solely on gel deformation within the soft mold to absorb contact forces, making it difficult to ensure the stability of contact forces during powder grinding. In most experimental tasks, while a single-arm robot can meet the expected requirements, in highly collaborative tasks in the field of chemical materials, especially in operations requiring close contact such as powder grinding, a single-arm robot not only fails to improve grinding efficiency but also cannot guarantee the scalability of the enhanced system.

[0003] Currently, there is no dual-arm robot powder grinding system for use in automated chemical laboratories, which leads to problems such as low grinding efficiency and inability to guarantee scalability in chemical powder grinding experiments.

[0004] In view of this, the present invention is hereby proposed. Summary of the Invention

[0005] The purpose of this invention is to provide a dual-arm robot powder grinding system and method for automated chemical laboratories, which can control the dual-arm robot to achieve stable and efficient grinding tasks, improve grinding efficiency and ensure scalability, thereby solving the above-mentioned technical problems existing in the prior art.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] A dual-arm robotic powder grinding system for automated chemical laboratories includes:

[0008] The equipment includes a mortar and pestle, a flexible clamp, a dual-arm robot, a grinding gripper, a soft shovel, a six-dimensional torque sensor, an RGBD camera, a powder sieving device, and a main controller; among which,

[0009] The mortar and pestle are fixed to the grinding platform;

[0010] The grinding pestle is mounted on the end of the grinding main arm of the dual-arm robot by an elastic clamp;

[0011] The powder sieving device can sieve the grinding powder in the mortar;

[0012] The six-dimensional torque sensor is installed at the end of the grinding arm of the dual-arm robot and connected to the grinding holder, which can hold a mortar, soft shovel or powder sieving device through the grinding holder.

[0013] The RGBD camera is mounted on the grinding main arm of the dual-arm robot and can capture RGB images and depth maps of the grinding platform.

[0014] The main controller is communicatively connected to both the RGBD camera and the dual-arm robot. Upon receiving a start command, it extracts local color features from the RGB image captured by the RGBD camera, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar and optimizes the path using linear interpolation. It then controls the grinding arm of the dual-arm robot to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, it dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot based on the powder distribution information, causing the grinding arm to grip the soft shovel and gather the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, it controls the grinding arm of the dual-arm robot to grip the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements. Finally, if the grinding time exceeds a preset time, it controls the dual-arm robot to stop grinding, completing the grinding task.

[0015] A dual-arm robotic powder grinding method for automated chemical laboratories, employing the system described in this invention, includes:

[0016] Step 1: After receiving the start command, the system extracts local color features from the RGB image captured by the RGBD camera through the main controller, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is a grinding operation, then step 2 is executed; if the determined subsequent operation is a gathering operation, then step 3 is executed; if the determined subsequent operation is a sieving operation, then step 4 is executed; if it is determined that the grinding time is greater than the preset time, then step 5 is executed.

[0017] Step 2: Based on the inner surface model of the mortar, the main controller generates a reference trajectory in the Cartesian coordinate system and optimizes the path using linear interpolation. The grinding arm of the dual-arm robot is controlled to adjust the motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. After the grinding operation is completed, proceed to step 1.

[0018] Step 3: The main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information, so that the grinding arm holds the soft shovel to perform a gathering operation, and gathers the scattered powder to the center of the mortar. After the gathering operation is completed, proceed to step 1.

[0019] Step 4: The grinding arm holds the mortar and uses a powder sieving device to sieve the ground powder, separating the powder that meets the particle size requirements. After the sieving operation is completed, proceed to step 1.

[0020] Step 5: Complete the grinding task.

[0021] Compared with existing technologies, the dual-arm robotic powder grinding system and method for automated chemical laboratories provided by this invention have the following advantages:

[0022] By using the visual feedback from the RGBD camera through the main controller, the operation mode is determined based on the changes in visual characteristics of the powder caused by the particle size during the grinding process. With the help of the set elastic clamps and powder sieving device, force-position mixing control and sieving operation are applied to the robot powder grinding task, which effectively ensures the stability of the grinding process, improves grinding efficiency and experimental repeatability, and realizes stable and efficient fine grinding of powder. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of a dual-arm robotic powder grinding system for an automated chemical laboratory, provided as an embodiment of the present invention.

[0025] Figure 2 This is a side view schematic diagram showing the details of the grinding trajectory motion generation in the system provided in an embodiment of the present invention.

[0026] Figure 3 This is a top view schematic diagram showing the details of the grinding trajectory motion generation in the system provided in an embodiment of the present invention.

[0027] Figure 4This is a schematic diagram of the force-position hybrid control loop of a dual-arm robot in a grinding task, provided in an embodiment of the present invention.

[0028] Figure 5 This is a schematic diagram of the structure of the elastic clamp in the system provided in an embodiment of the present invention.

[0029] Figure 6 This is a schematic diagram of the powder sieving device in the system provided in an embodiment of the present invention.

[0030] Figure 7 A flowchart of a dual-arm robot powder grinding method for an automated chemical laboratory, provided as an embodiment of the present invention. Detailed Implementation

[0031] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the specific content of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments, which do not constitute a limitation of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0032] First, the following explanations are provided for the terms that may be used in this article:

[0033] The term "and / or" means that either or both can be achieved simultaneously. For example, X and / or Y means that it includes both "X" or "Y" as well as the three cases of "X and Y".

[0034] The terms "comprising," "including," "containing," "having," or other similar semantic descriptions should be interpreted as non-exclusive inclusion. For example, including a technical feature element (such as raw material, component, ingredient, carrier, dosage form, material, size, part, component, mechanism, device, step, process, method, reaction conditions, processing conditions, parameter, algorithm, signal, data, product or article of manufacture, etc.) should be interpreted as including not only the expressly listed technical feature element, but also other technical feature elements that are not expressly listed and are well-known in the art.

[0035] The term "composed of" excludes any technical features not expressly listed. When used in a claim, it closes the claim to exclude all technical features other than those expressly listed, except for associated conventional impurities. If the term appears only in a clause of a claim, it limits the claim to the elements expressly listed in that clause; elements recited in other clauses are not excluded from the overall claim.

[0036] Unless otherwise explicitly specified or limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this document according to the specific circumstances.

[0037] The terms “center,” “longitudinal,” “lateral,” “length,” “width,” “thickness,” “upper,” “lower,” “front,” “back,” “left,” “right,” “vertical,” “horizontal,” “top,” “bottom,” “inner,” “outer,” “clockwise,” and “counterclockwise” indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience and simplification of description and do not imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this document.

[0038] The solution provided by this invention will be described in detail below. Contents not described in detail in the embodiments of this invention are prior art known to those skilled in the art. Where specific conditions are not specified in the embodiments of this invention, they shall be performed according to conventional conditions in the art or conditions recommended by the manufacturer. Reagents or instruments used in the embodiments of this invention whose manufacturers are not specified are all conventional products that can be purchased commercially.

[0039] like Figure 1 As shown, this invention provides a dual-arm robotic powder grinding system for automated chemical laboratories, comprising:

[0040] The equipment includes a mortar and pestle, a flexible clamp, a dual-arm robot, a grinding gripper, a soft shovel, a six-dimensional torque sensor, an RGBD camera, a powder sieving device, and a main controller; among which,

[0041] The mortar and pestle are fixed to the grinding platform;

[0042] The grinding pestle is mounted on the end of the grinding main arm of the dual-arm robot by an elastic clamp;

[0043] The powder sieving device can sieve the grinding powder in the mortar;

[0044] The six-dimensional torque sensor is installed at the end of the grinding arm of the dual-arm robot and connected to the grinding holder, which can hold a mortar, soft shovel or powder sieving device through the grinding holder.

[0045] The RGBD camera is mounted on the grinding main arm of the dual-arm robot and can capture RGB images and depth maps of the grinding platform.

[0046] The main controller is communicatively connected to both the RGBD camera and the dual-arm robot. Upon receiving a start command, it extracts local color features from the RGB image captured by the RGBD camera, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar and optimizes the path using linear interpolation. It then controls the grinding arm of the dual-arm robot to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, it dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot based on the powder distribution information, causing the grinding arm to grip the soft shovel and gather the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, it controls the grinding arm of the dual-arm robot to grip the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements. Finally, if the grinding time exceeds a preset time, it controls the dual-arm robot to stop grinding, completing the grinding task.

[0047] Preferably, in the above-mentioned grinding system, the main controller includes:

[0048] Visual feedback module and force-controlled grinding module; among which,

[0049] The visual feedback module can extract local color features from the RGB image captured by the RGBD camera after receiving the start command, perform global positioning of the powder in the mortar, and determine the subsequent operation of the dual-arm robot based on the dispersion degree of the powder.

[0050] The force-controlled grinding module is communicatively connected to both the visual feedback module and the dual-arm robot. When the visual feedback module determines that the subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar and optimizes the path using linear interpolation. It then controls the grinding main arm of the dual-arm robot to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, it dynamically adjusts the motion trajectory of the grinding slave arm of the dual-arm robot based on the powder distribution information, causing the grinding slave arm to grip the soft shovel and perform a gathering operation to collect the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, it controls the grinding slave arm of the dual-arm robot to grip the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements.

[0051] Preferably, in the above-mentioned grinding system, the visual feedback module includes:

[0052] The system comprises a visual sampling submodule, a powder positioning submodule, and an operation decision submodule; among which,

[0053] The visual sampling submodule can sample the powder aggregation area in the RGB image captured by the RGBD camera to obtain the sampling area, and calculate the RGB mean and covariance matrix of the powder color in the sampling area to obtain the color range of the powder in the current mortar.

[0054] The powder positioning submodule is communicatively connected to the visual sampling submodule. It can perform binarization positioning on the pixels that satisfy the color range of the powder in the mortar based on the color range of the powder in the current mortar obtained by the visual sampling submodule to obtain the powder positioning range.

[0055] The operation decision submodule is communicatively connected to the powder positioning submodule and can determine the subsequent operations of the dual-arm robot based on the powder positioning range obtained by the powder positioning submodule.

[0056] The force-controlled grinding module includes: a mortar inner surface model construction submodule, a grinding reference trajectory generation submodule, and an end impedance controller; wherein...

[0057] The mortar inner surface model construction submodule can use a semi-circular curved surface to approximate the model of the mortar inner surface, thus obtaining the mortar inner surface model.

[0058] The grinding reference trajectory generation submodule is communicatively connected to the mortar inner surface model construction submodule, and can generate the grinding reference trajectory of the grinding main arm of the dual-arm robot based on the mortar inner surface model constructed by the mortar inner surface model construction submodule.

[0059] The end impedance main controller is communicatively connected to the grinding reference trajectory generation submodule and can adjust the grinding trajectory of the grinding main arm through real-time force feedback based on the grinding reference trajectory generated by the grinding reference trajectory generation submodule.

[0060] Preferably, in the above-mentioned grinding system, the visual sampling submodule samples the powder aggregation area in the RGB image captured by the RGBD camera to obtain a sampling area, and calculates the RGB mean and covariance matrix of the powder color within the sampling area to obtain the color range of the powder in the current mortar, including:

[0061] In RGB image The sampling area was obtained by sampling the powder aggregation area in the sample. for:

[0062] ;

[0063] in, The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The pixel radius of the sampling region;

[0064] Calculate the sampling area RGB mean of internal powder color With covariance matrix The average RGB value of the powder With covariance matrix This indicates the color range of the powder in the mortar, and the RGB mean is calculated. With covariance matrix The calculation formulas are as follows:

[0065] ;

[0066] ;

[0067] in, It is the RGB vector of an RGB image;

[0068] The powder positioning submodule performs binarization positioning on pixels within the current color range of the powder in the mortar, based on the color range of the powder obtained by the visual sampling submodule, to obtain the powder positioning range, including:

[0069] RGB vectors of an RGB image The pixels within the mortar that satisfy the color range of the powder in the current mortar obtained by the visual sampling submodule are processed to obtain a binarized image. for:

[0070] ;

[0071] in, RGB pixels Mahalanobis distance; To set a threshold; This is an indicator function for converting an RGB image to a binary image. If a pixel meets the threshold condition, it is set to 1 to represent white, and otherwise it is set to 0 to represent black.

[0072] Define a binarized image Set of connected white regions within for:

[0073] ;

[0074] From the set The smallest circumcircle of the connected region with the largest area is selected from the options. As the final powder positioning range, the smallest circumcircle Represented as:

[0075] ;

[0076] in, for The smallest enclosing circle.

[0077] The operation decision submodule determines the subsequent operations of the dual-arm robot based on the powder positioning range obtained by the powder positioning submodule in the following manner:

[0078] The pixel coordinates of the center of the inner surface of the mortar in the RGB image The powder positioning range obtained from the powder positioning submodule The furthest distance and the set maximum dispersion distance Calculate the dispersion index of the powder. for:

[0079] ;

[0080] According to the dispersion index The subsequent operations of the dual-arm robot are determined as follows, if the dispersion index is... When the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed, and the subsequent operation of the dual-arm robot is determined to be a gathering operation; if the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed. If the value is less than or equal to 1, then if the powder is still concentrated, the subsequent operation of the dual-arm robot is determined to be a grinding operation.

[0081] Preferably, in the above-mentioned grinding system, the mortar inner surface model construction submodule models the mortar inner surface using a semi-circular surface as an approximation, thereby obtaining the mortar inner surface model, including:

[0082] Constructed mortar inner surface model Represented as:

[0083] ;

[0084] in, The depth of the mortar; The radius of the model of the inner surface of the mortar; Model of the inner surface of the mortar The coordinates of the point on the [the map].

[0085] The grinding reference trajectory generation submodule generates the grinding reference trajectory of the dual-arm robot's grinding main arm based on the mortar inner surface model constructed by the mortar inner surface model construction submodule, including:

[0086] The tip of the grinding pestle is positioned perpendicular to the tangent plane of the inner surface of the mortar. This orientation is obtained by performing a ZYZ Euler transformation on the reference coordinate system. The specific Euler angles are expressed as follows:

[0087] ;

[0088] ;

[0089] ;

[0090] in, , and These are the Euler angles for roll, yaw, and pitch, respectively.

[0091] The path points of the grinding motion are calculated using Euler angles and 6D pose interpolation, and the grinding reference trajectory of the grinding arm of the dual-arm robot is generated using an open-source robot operation and motion planning framework.

[0092] The end impedance controller adjusts the grinding trajectory of the grinding arm through real-time force feedback based on the grinding reference trajectory generated by the input grinding reference trajectory generation submodule, including:

[0093] The end-effector impedance controller designed for the dual-arm robot adjusts the grinding trajectory of the grinding arm through real-time force feedback. The end-effector impedance controller is as follows:

[0094] ;

[0095] ;

[0096] in, , and The matrix is ​​a diagonal matrix, representing the mass, damping, and stiffness matrices of the spring-damped-mass system, respectively. The control objective of the force control subspace is selected by the diagonal matrix. Decide, The value can be 0 or 1. Representing the task space degrees of freedom It is the absolute value of the contact force error response. This represents the reference pose under no contact force. The desired pose under the influence of contact force. Indicates the current actual pose With reference pose The pose error between them; and Let represent the acceleration and velocity of the pose error, respectively. The control law is expressed as:

[0097] ;

[0098] ;

[0099] Acceleration of the end effector of the grinding arm under force-position hybrid control based on external force determination And by selecting the matrix Adjust the forces and moments within the force-controlled subspace;

[0100] The entire control cycle of the end effector impedance controller of the dual-arm robot is within the control period. Inside, the desired joint angle is obtained from the reference trajectory. and angular velocity , and the actual joint angle and angular velocity Comparison using forward kinematics and Jacobian matrix Calculate the pose error and the speed of pose error And obtain the acceleration of the pose error. To obtain the acceleration of the pose error Then, using pseudo-inverse Jacobian matrix Calculate the joint velocity induced by external force. Then, the joint velocity With the desired joint angular velocity Add them together to get the actual joint velocity, integrate the actual joint velocity, and then compare it with the desired joint angle. Add them together to obtain the final joint control amount. Then, the joint control amount The signal is sent to the grinding main arm of the dual-arm robot, thus completing one control cycle.

[0101] Preferably, in the above-mentioned grinding system, if the determined subsequent operation is a gathering operation, the main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information in the following manner, so that the grinding arm grips the soft shovel to perform a gathering operation, collecting the scattered powder to the center of the mortar, including:

[0102] The main controller calculates the positive polarity of the powder in the pixel coordinate system using the following formula. Axis, negative Axis, positive Axis and negative Dispersion index in the four directions of the axis , The formula is:

[0103] ;

[0104] Where i=1, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=2, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system The minimum value of the axis; when i=3, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=4, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system Minimum value of the axis; The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The maximum dispersion distance is set.

[0105] When the dispersion index of the corresponding direction When the value is greater than 1, the powder in that direction is gathered. The main controller controls the grinding arm to hold the soft shovel and gathers the powder in the corresponding direction in the mortar according to the direction obtained above and through the given gathering trajectory.

[0106] If a fixed sieving cycle is reached, the main controller controls the grinding process by using a powder sieving device to sieve the ground powder from the mortar held by the grinding arm, separating powder that meets the particle size requirements, including:

[0107] A powder sieving device using a vibrating motor-driven screen is employed.

[0108] The main controller controls the grinding arm to pour the powder from the mortar into the screen of the powder sieving device. The vibrating motors on both sides of the screen drive the screen to vibrate, shaking the small powder particles to the bottom tray, while the large powder particles remain above the screen, thus completing the separation of powders of different particle sizes. Afterward, the grinding arm is controlled to hold the screen and pour the large powder particles back into the mortar along the guide groove, completing the sieving operation.

[0109] like Figure 7 As shown, this invention also provides a dual-arm robot powder grinding method for automated chemical laboratories, employing the above-described system, including:

[0110] Step 1: After receiving the start command, the system extracts local color features from the RGB image captured by the RGBD camera through the main controller, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is a grinding operation, then step 2 is executed; if the determined subsequent operation is a gathering operation, then step 3 is executed; if the determined subsequent operation is a sieving operation, then step 4 is executed; if it is determined that the grinding time is greater than the preset time, then step 5 is executed.

[0111] Step 2: Based on the inner surface model of the mortar, the main controller generates a reference trajectory in the Cartesian coordinate system and optimizes the path using linear interpolation. The grinding arm of the dual-arm robot is controlled to adjust the motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. After the grinding operation is completed, proceed to step 1.

[0112] Step 3: The main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information, so that the grinding arm holds the soft shovel to perform a gathering operation, and gathers the scattered powder to the center of the mortar. After the gathering operation is completed, proceed to step 1.

[0113] Step 4: The grinding arm holds the mortar and uses a powder sieving device to sieve the ground powder, separating the powder that meets the particle size requirements. After the sieving operation is completed, proceed to step 1.

[0114] Step 5: Complete the grinding task.

[0115] Preferably, in step 1 of the above grinding method, after receiving the start command, the system extracts local color features from the RGB image captured by the RGBD camera through the main controller in the following manner, performs global localization of the powder in the mortar, and determines the subsequent operations of the dual-arm robot based on the dispersion degree of the powder, including:

[0116] Step 11, RGB image captured by RGBD camera The sampling area was obtained by sampling the powder aggregation area in the sample. for:

[0117] ;

[0118] in, The pixel coordinates of the center of the inner surface of the mortar in the RGB image; Sampling area The pixel radius;

[0119] Calculate the sampling area RGB mean of internal powder With covariance matrix The average RGB value of the powder With covariance matrix This indicates the color range of the powder in the mortar, RGB average. With covariance matrix The calculation formulas are as follows:

[0120] ;

[0121] ;

[0122] in, It is the RGB vector of the image;

[0123] Step 12, process the RGB vectors of the RGB image. The pixels that satisfy the color range of the powder in the mortar obtained in step 11 are processed to obtain a binarized image. for:

[0124] ;

[0125] in, RGB image pixels Mahalanobis distance; To set a threshold; This is an indicator function for converting an RGB image into a binary image. If a pixel meets a set threshold, it is set to 1 to represent white; otherwise, it is set to 0 to represent black.

[0126] Define a binarized image Set of connected white regions within for:

[0127] ;

[0128] From the set The smallest circumcircle of the connected region with the largest area is selected from the options. As the final powder positioning range, the smallest circumcircle Represented as:

[0129] ;

[0130] in, for The smallest enclosing circle;

[0131] Step 13, using the pixel coordinates of the center of the inner surface of the mortar in the RGB image. The powder positioning range obtained in step 12 The furthest distance and the set maximum dispersion distance Calculate the dispersion index of the powder. for:

[0132] ;

[0133] According to the dispersion index The subsequent operations of the dual-arm robot are determined as follows, if the dispersion index is... When the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed, and the subsequent operation of the dual-arm robot is determined to be a gathering operation; if the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed. If the value is less than or equal to 1, then if the powder is still concentrated, the subsequent operation of the dual-arm robot is to perform a grinding operation.

[0134] Preferably, in step 2 of the above grinding method, the main controller generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar, and optimizes the path using linear interpolation. The main arm of the dual-arm robot is controlled to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. This includes:

[0135] Step 21: The inner surface of the mortar is approximated by a semi-circular surface using the main controller to model the mortar's inner surface, thus obtaining the mortar's inner surface model. Represented as:

[0136] ;

[0137] in, The depth of the mortar; The radius of the model of the inner surface of the mortar; Model of the inner surface of the mortar coordinates on;

[0138] Step 22: Based on the mortar inner surface model constructed by the mortar inner surface model construction submodule, generate the grinding reference trajectory of the grinding main arm of the dual-arm robot, specifically as follows:

[0139] The tip of the grinding pestle is positioned perpendicular to the tangent plane of the inner surface of the mortar. This orientation is obtained by performing a ZYZ Euler transformation on the reference coordinate system. The specific Euler angles are expressed as follows:

[0140] ;

[0141] ;

[0142] ;

[0143] in, , and These are the Euler angles for roll, yaw, and pitch, respectively.

[0144] The path points of the grinding motion are calculated using Euler angles and 6D pose interpolation, and the grinding reference trajectory of the grinding main arm of the dual-arm robot is generated using an open-source robot operation and motion planning framework.

[0145] Step 23: Based on the grinding reference trajectory generated by the input grinding reference trajectory generation submodule, the grinding trajectory of the grinding arm is adjusted through real-time force feedback, specifically as follows:

[0146] The end-effector impedance controller designed for the dual-arm robot adjusts the grinding trajectory of the grinding arm through real-time force feedback. The end-effector impedance controller is as follows:

[0147] ;

[0148] ;

[0149] in, , and The matrix is ​​a diagonal matrix, representing the mass, damping, and stiffness matrices of the spring-damped-mass system, respectively. The control objective of the force control subspace is selected by the diagonal matrix. Decide, The value can be 0 or 1. Representing the task space degrees of freedom It is the absolute value of the contact force error response. This represents the reference pose under no contact force. The desired pose under the influence of contact force. Indicates the current actual pose With reference pose The error between; and Let represent the acceleration and velocity of the pose error, respectively. The control law is expressed as:

[0150] ;

[0151] ;

[0152] Acceleration of the end effector of the grinding arm under force-position hybrid control based on external force determination And by selecting the matrix Adjust the forces and torques within the force-controlled subspace.

[0153] Preferably, in step 3 of the above grinding method, the main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information in the following manner, so that the grinding arm clamps the soft shovel to perform a gathering operation, collecting the scattered powder to the center of the mortar, including:

[0154] Step 31: The main controller calculates the positive coordinates of the powder in the pixel coordinate system according to the following formula. Axis, negative Axis, positive Axis and negative Dispersion index in the four directions of the axis , The formula is:

[0155] ;

[0156] Where i=1, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=2, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system The minimum value of the axis; when i=3, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=4, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system Minimum value of the axis; The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The maximum dispersion distance is set.

[0157] Step 32, when the dispersion index of the corresponding direction When the value is greater than 1, the powder in that direction is gathered together. The main controller controls the grinding arm to pick up the soft shovel and gathers the powder in the corresponding direction in the mortar according to the direction obtained above and the given gathering trajectory.

[0158] In step 4, the main controller controls the grinding process by using a powder sieving device to sieve the ground powder from the mortar held by the grinding arm, separating powder that meets the particle size requirements, including:

[0159] A powder sieving device using a vibrating motor-driven screen is employed.

[0160] The main controller controls the grinding arm to pour the powder from the mortar into the screen of the powder sieving device. The vibrating motors on both sides of the screen drive the screen to vibrate, shaking the small powder particles to the bottom tray, while the large powder particles remain above the screen, thus completing the separation of powders of different particle sizes. Afterward, the grinding arm is controlled to hold the screen and pour the large powder particles back into the mortar along the guide groove, completing the sieving operation.

[0161] In summary, the grinding system and method provided by the embodiments of the present invention, through the visual feedback of the RGBD camera by the main controller, determines the operation mode based on the changes in visual characteristics of the powder caused by the particle size during the grinding process. With the help of the set elastic clamps and powder sieving device, force-position mixing control and sieving operation are applied to the robot powder grinding task, which effectively ensures the stability of the grinding process, improves the grinding efficiency and experimental repeatability, and realizes stable and efficient fine grinding of powder.

[0162] To more clearly demonstrate the technical solution and its effects provided by the present invention, the following detailed description of the solution provided by the embodiments of the present invention is provided with reference to specific examples.

[0163] Example 1

[0164] like Figure 1 As shown, this embodiment provides a dual-arm robotic powder grinding system for automated chemical laboratories, capable of achieving stable and efficient grinding of chemical powders. The system includes:

[0165] The system includes a dual-arm robot, a mortar and pestle, a grinding pestle, a flexible gripper, a grinding holder, a soft shovel, a six-dimensional torque sensor, an RGBD camera, a powder sieving device, and a main controller; among these components...

[0166] A mortar, pestle, and soft spatula are each used to form a grinding tool set. This tool set works in conjunction with a dual-arm robot to form a dual-arm grinding robot. The mortar is fixed to the grinding platform, and the pestle is mounted on the end of the robot's arm via a flexible clamp, with this arm designated as the main grinding arm. The other arm is the secondary grinding arm, with a grinding gripper installed at its end to hold and fix the mortar and gather the powder for grinding using the soft spatula.

[0167] Each of the robotic arms of the dual-arm robot is a multi-degree-of-freedom manipulator;

[0168] The elastic clamp consists of upper and lower flanges 2 and four springs 4 with slide rails 3 in the middle. Figure 5 As shown;

[0169] The six-dimensional torque sensor is installed at the end of the grinding arm of the dual-arm robot and connected to the grinding gripper, which can hold the mortar and soft spatula.

[0170] The RGBD camera is mounted on the grinding main arm of the dual-arm robot and can capture RGB images and depth maps of the grinding platform; if a single camera is used, it is mounted at the end of the grinding main arm in a "eye on hand" manner.

[0171] like Figure 6 As shown, the powder sieving device can sieve the ground powder in the mortar. The sieving device mainly includes a vibrating motor 50, a screen 40, and a main support 10. The vibrating motor 50 is mounted on the main support 10, and the screen 40 is mounted inside the main support 10. The vibrating motor 50 can drive the screen 40 to vibrate inside the main support 10. A powder splash guard 30 is provided at the upper end of the main support 10, and a powder guide channel 20 is provided on one side. A tray 60 is provided at the bottom of the screen 40 inside the main support 10. The main support 10 can be printed by a 3D printer.

[0172] The main controller is communicatively connected to both the RGBD camera and the dual-arm robot. Upon receiving a start command, it extracts local color features from the RGB image captured by the RGBD camera, globally locates the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the powder dispersion. If the determined subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the mortar's inner surface model and optimizes the path using linear interpolation. The main grinding arm of the dual-arm robot is controlled to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, the motion trajectory of the grinding slave arm of the dual-arm robot is dynamically adjusted based on the powder distribution information, allowing the grinding slave arm to hold a soft shovel and perform a gathering operation, collecting the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, the grinding slave arm is controlled to hold the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements. Finally, if the grinding time exceeds a preset time, the dual-arm robot is controlled to stop grinding, completing the grinding task.

[0173] In this embodiment, the dual-arm robotic powder grinding system operates as follows: Upon receiving a grinding command, the main controller's visual feedback module extracts local color features from the RGB image captured by the camera to achieve global powder localization and selects subsequent operations based on the powder's dispersion. If grinding is required, the main controller's force-controlled grinding module generates a reference trajectory in a Cartesian coordinate system based on the mortar's inner surface model and optimizes the path using linear interpolation. Subsequently, the main arm adjusts its motion trajectory in real time through force-position hybrid control to maintain stable grinding contact force. The movement of the grinding pestle inevitably leads to powder dispersion. The system dynamically adjusts the trajectory of the slave arm based on the powder distribution and uses a soft shovel to gather the scattered powder to the center of the mortar. Furthermore, to optimize particle size distribution and improve grinding efficiency, after each fixed sieving cycle T, the slave arm, in conjunction with a sieving device, separates powder that meets the particle size requirements.

[0174] Example 2

[0175] This embodiment provides a dual-arm robotic powder grinding system for automated chemical laboratories, using the system of Embodiment 1, and specifically includes the following steps:

[0176] Step 1: The powder grinding system uses visual feedback to obtain information about the powder distribution within the mortar and makes optimal operational decisions based on this information. This step mainly includes a visual sampling step, a powder localization step, and an operational decision-making step, comprising the following steps:

[0177] Step 11: Construct a visual threshold by extracting local color features of the powder in the RGB image. Sampling was performed on the powder aggregation area, and the sampling area was defined as follows:

[0178] ;

[0179] in, Here are the pixel coordinates of the center of the inner surface of the mortar in the image. is the pixel radius of the sampling area.

[0180] Next, calculate the sampling area. RGB mean of internal powder With covariance matrix This represents the color range of the powder currently in the mortar. (RGB mean) With covariance matrix Represented as:

[0181] ;

[0182] ;

[0183] in, It is the RGB vector of the image.

[0184] Step 12: Based on the local color characteristics of the powder, locate the overall powder distribution range. Since the powder disperses during the grinding process with the movement of the grinding pestle, it may not be concentrated in the sampling area during visual positioning. Therefore, the calculated... and The color of the powder may be affected by the color of the inner surface of the mortar. Therefore, we chose to perform visual sampling after the powder had been agglomerated, i.e., when the powder was concentrated at the bottom of the mortar, to obtain an accurate powder color range. Then we... Pixels within this range are processed to achieve visual positioning of the powder. The processed pixels... Represented as:

[0185] ;

[0186] in, RGB pixels Mahalanobis distance, To set a threshold. This is an indicator function that converts an RGB image into a binary image. If a pixel meets the threshold condition, it is set to 1 (white); otherwise, it is set to 0 (black). Figure 2 and Figure 3 The result of this binarization process is shown.

[0187] To avoid the influence of specks or small amounts of powder far from the center of the mortar on the accurate localization of powder-dense areas, we define a binarized image. Set of connected white regions within for:

[0188] ;

[0189] From the set The smallest circumcircle of the connected region with the largest area is selected from the options. The minimum circumcircle ℛ, representing the final positioning range of the powder, is denoted as:

[0190] ;

[0191] in, for The smallest enclosing circle.

[0192] Step 13, based on the powder positioning range The decision is made regarding the subsequent operation of the grinding system. Firstly, based on... arrive The furthest distance and the set maximum dispersion distance Calculate the dispersion index of the powder. :

[0193] ;

[0194] when When the value is greater than 1, it indicates that the powder distribution is relatively dispersed. In order to ensure sufficient contact between the powder and the grinding pestle, a gathering operation should be performed. Conversely, if the powder is still concentrated, the grinding operation should continue.

[0195] Step 2: When the powder is concentrated, the grinding arm system performs the grinding operation.

[0196] Step 21: When modeling the inner surface of the mortar, a semi-circular surface is used for approximation. This model of the inner surface of the mortar... It can be represented as:

[0197]

[0198] in, For the depth of the mortar, Let be the radius of the model of the inner surface of the mortar. For the model The coordinates of the point on the [the map].

[0199] Step 22: Generate a grinding reference trajectory based on the inner surface model of the mortar. For example... Figure 2 and Figure 3 As shown, the direction of the pestle tip is set perpendicular to the tangent plane of the inner surface of the mortar. This orientation is obtained by performing a ZYZ Euler transformation on the reference coordinate system. The specific Euler angles are expressed as follows:

[0200] ;

[0201] ;

[0202] ;

[0203] Subsequently, 6D pose interpolation was used to calculate the path points of the grinding motion, and MoveIt!, an open-source robot manipulation and motion planning framework based on ROS (Robot Operating System), was used to generate a reference trajectory.

[0204] Step 23: The reference trajectory is input to the end impedance controller, which adjusts the grinding trajectory based on real-time force feedback. This end impedance controller is designed as follows:

[0205] ;

[0206] ;

[0207] in, , and The matrix is ​​a diagonal matrix, representing the mass, damping, and stiffness matrices of the spring-damped-mass system, respectively. The control objective of the force control subspace is selected by the diagonal matrix. Decide, The value can be 0 or 1. Representing the task space degrees of freedom It is the absolute value of the contact force error response. This represents the reference pose under no contact force. The desired pose under the influence of contact force, and Indicates the current actual pose With reference pose The error between them. Furthermore. and Let represent the acceleration and velocity of the pose error, respectively. The control law can be expressed as:

[0208] ;

[0209] ;

[0210] Acceleration of the end effector under force-position hybrid control based on external force determination And by selecting the matrix Adjust the forces and torques within the force control subspace. The entire control loop is as follows: Figure 4 As shown, the system in the control cycle Inside, the desired joint angle is obtained from the reference trajectory. and angular velocity , and the actual joint angle and angular velocity Comparison using forward kinematics and Jacobian matrix Calculate and and obtain .get Then, we first utilize forward kinematics and the Jacobian matrix. pseudo-inverse Jacobian matrix Calculate the joint velocity induced by external force. Then, the joint velocity is compared with... Add them together to get the actual joint velocity. Further integrate the actual joint velocity and then... Add them together to obtain the joint control quantity. Then The signal is sent to the grinding robotic arm to complete one control cycle.

[0211] Step 3: When the powder is dispersed, the grinding arm performs a gathering operation.

[0212] Step 31: Calculate the powder dispersion index in each direction and decide whether to perform a powder agglomeration operation in that direction. For powder in the negative pixel coordinate system... Axis, positive Axis, negative Axis and positive The degree of dispersion in the four directions of the axis is determined by... , , and The calculation yielded:

[0213] ;

[0214] in, , , , They are Inner pixel coordinates The sum of the maximum and minimum values ​​of the axis The maximum and minimum values ​​of the axis. To improve collection efficiency, only when the dispersion of the corresponding orientation is... When the value is greater than 1, the powder in that direction is gathered together.

[0215] Step 32: Grind the small shovel picked up from the arm and, based on the decision in step 31, gather the powder in the mortar according to the given gathering trajectory.

[0216] Step 4: After each fixed sieving cycle, the dual-arm robot performs sieving operations on the powder in the mortar. The sieving device mainly consists of a vibrating motor, a screen, a main support frame, and auxiliary structures, such as... Figure 6As shown. During each sieving cycle T, the robotic arm, in conjunction with the powder sieving device, filters the powder in the mortar, returning only powder with a particle size larger than the desired value to the mortar for further grinding. The sieving operation is completed collaboratively by the robotic arm and the sieving device. First, the robotic arm pours the powder from the mortar onto the screen. Motors on both sides of the screen vibrate, shaking small powder particles to the bottom tray, while larger particles remain above the screen, achieving separation of powders of different sizes. Next, the robotic arm holds the screen and pours the larger powder particles back into the mortar along the guide channel, completing the sieving operation. Furthermore, to reduce material loss and platform contamination, the vibrating motor ensures that no powder remains on the screen, and the robotic arm holds a splash guard to prevent powder from splashing.

[0217] Step 5: If the grinding time exceeds the set time, stop the grinding operation and complete the grinding task.

[0218] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0219] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims. The information disclosed in the background section is intended only to enhance the understanding of the overall background technology of the present invention and should not be construed as an admission or implication in any way that such information constitutes prior art known to those skilled in the art.

Claims

1. A dual-arm robotic powder grinding system for automated chemical laboratories, characterized in that, include: The equipment includes a mortar and pestle, a flexible clamp, a dual-arm robot, a grinding gripper, a soft shovel, a six-dimensional torque sensor, an RGBD camera, a powder sieving device, and a main controller; among these, The mortar and pestle are fixed to the grinding platform; The grinding pestle is mounted on the end of the grinding main arm of the dual-arm robot by an elastic clamp; The powder sieving device can sieve the grinding powder in the mortar; The six-dimensional torque sensor is installed at the end of the grinding arm of the dual-arm robot and connected to the grinding holder, which can hold a mortar, soft shovel or powder sieving device through the grinding holder. The RGBD camera is mounted on the grinding main arm of the dual-arm robot and can capture RGB images and depth maps of the grinding platform. The main controller is communicatively connected to both the RGBD camera and the dual-arm robot. Upon receiving a start command, it extracts local color features from the RGB image captured by the RGBD camera, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar and optimizes the path using linear interpolation. It then controls the grinding arm of the dual-arm robot to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, it dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot based on the powder distribution information, causing the grinding arm to grip the soft shovel and gather the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, it controls the grinding arm of the dual-arm robot to grip the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements. Finally, if the grinding time exceeds a preset time, it controls the dual-arm robot to stop grinding, completing the grinding task.

2. The dual-arm robotic powder grinding system for automated chemical laboratories according to claim 1, characterized in that, The main controller includes: Visual feedback module and force-controlled grinding module; among them, The visual feedback module can extract local color features from the RGB image captured by the RGBD camera after receiving the start command, perform global positioning of the powder in the mortar, and determine the subsequent operation of the dual-arm robot based on the dispersion degree of the powder. The force-controlled grinding module is communicatively connected to both the visual feedback module and the dual-arm robot. When the visual feedback module determines that the subsequent operation is grinding, it generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar and optimizes the path using linear interpolation. It then controls the grinding main arm of the dual-arm robot to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. If the determined subsequent operation is gathering, it dynamically adjusts the motion trajectory of the grinding slave arm of the dual-arm robot based on the powder distribution information, causing the grinding slave arm to grip the soft shovel and perform a gathering operation to collect the scattered powder to the center of the mortar. If a fixed sieving cycle is reached, it controls the grinding slave arm of the dual-arm robot to grip the mortar and use a powder sieving device to sieve the ground powder, separating powder that meets the particle size requirements.

3. The dual-arm robotic powder grinding system for automated chemical laboratories according to claim 2, characterized in that, The visual feedback module includes: The system comprises a visual sampling submodule, a powder positioning submodule, and an operation decision submodule; among which, The visual sampling submodule can sample the powder aggregation area in the RGB image captured by the RGBD camera to obtain the sampling area, and calculate the RGB mean and covariance matrix of the powder color in the sampling area to obtain the color range of the powder in the current mortar. The powder positioning submodule is communicatively connected to the visual sampling submodule. It can perform binarization positioning on the pixels that satisfy the color range of the powder in the mortar based on the color range of the powder in the current mortar obtained by the visual sampling submodule to obtain the powder positioning range. The operation decision submodule is communicatively connected to the powder positioning submodule and can determine the subsequent operations of the dual-arm robot based on the powder positioning range obtained by the powder positioning submodule. The force-controlled grinding module includes: a mortar inner surface model construction submodule, a grinding reference trajectory generation submodule, and an end impedance controller; wherein... The mortar inner surface model construction submodule can use a semi-circular curved surface to approximate the model of the mortar inner surface, thus obtaining the mortar inner surface model. The grinding reference trajectory generation submodule is communicatively connected to the mortar inner surface model construction submodule, and can generate the grinding reference trajectory of the grinding main arm of the dual-arm robot based on the mortar inner surface model constructed by the mortar inner surface model construction submodule. The end impedance controller is communicatively connected to the grinding reference trajectory generation submodule and can adjust the grinding trajectory of the grinding arm through real-time force feedback based on the grinding reference trajectory generated by the grinding reference trajectory generation submodule.

4. The dual-arm robotic powder grinding system for automated chemical laboratories according to claim 3, characterized in that, The visual sampling submodule samples the powder aggregation area in the RGB image captured by the RGBD camera to obtain the sampling area, and calculates the RGB mean and covariance matrix of the powder color within the sampling area to obtain the color range of the powder in the current mortar, including: In RGB image The sampling area was obtained by sampling the powder aggregation area in the sample. for: ； in, The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The pixel radius of the sampling region; Calculate the sampling area RGB mean of internal powder color With covariance matrix The average RGB value of the powder With covariance matrix This indicates the color range of the powder in the mortar, and the RGB mean is calculated. With covariance matrix The calculation formulas are as follows: ； ； in, It is the RGB vector of an RGB image; The powder positioning submodule performs binarization positioning on pixels within the current color range of the powder in the mortar, based on the color range of the powder obtained by the visual sampling submodule, to obtain the powder positioning range, including: RGB vectors of an RGB image The pixels within the mortar that satisfy the color range of the powder in the current mortar obtained by the visual sampling submodule are processed to obtain a binarized image. for: ； in, RGB pixels Mahalanobis distance; To set a threshold; This is an indicator function for converting an RGB image to a binary image. If a pixel meets the threshold condition, it is set to 1 to represent white, and otherwise it is set to 0 to represent black. Define a binarized image Set of connected white regions within for: ； From the set The smallest circumcircle of the connected region with the largest area is selected from the options. As the final powder positioning range, the smallest circumcircle Represented as: ； in, for The smallest enclosing circle; The operation decision submodule determines the subsequent operations of the dual-arm robot based on the powder positioning range obtained by the powder positioning submodule in the following manner: The pixel coordinates of the center of the inner surface of the mortar in the RGB image The powder positioning range obtained from the powder positioning submodule The furthest distance and the set maximum dispersion distance Calculate the dispersion index of the powder. for: ； According to the dispersion index The subsequent operations of the dual-arm robot are determined as follows, if the dispersion index is... When the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed, and the subsequent operation of the dual-arm robot is determined to be a gathering operation; if the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed. If the value is less than or equal to 1, then if the powder is still concentrated, the subsequent operation of the dual-arm robot is determined to be a grinding operation.

5. The dual-arm robotic powder grinding system for automated chemical laboratories according to claim 4, characterized in that, The mortar inner surface model construction submodule uses a semi-circular surface to approximate the mortar inner surface in the following manner to obtain the mortar inner surface model, including: Constructed mortar inner surface model Represented as: ； in, The depth of the mortar; The radius of the model of the inner surface of the mortar; Model of the inner surface of the mortar coordinates on; The grinding reference trajectory generation submodule generates the grinding reference trajectory of the dual-arm robot's grinding main arm based on the mortar inner surface model constructed by the mortar inner surface model construction submodule, including: The tip of the grinding pestle is set to be perpendicular to the tangent plane of the inner surface of the mortar. This orientation is obtained by performing a ZYZ Euler transformation on the reference coordinate system. The specific Euler angles are expressed as follows: ； ； ； in, , and These are the Euler angles for roll, yaw, and pitch, respectively. The path points of the grinding motion are calculated using Euler angles and 6D pose interpolation, and the grinding reference trajectory of the grinding main arm of the dual-arm robot is generated using an open-source robot operation and motion planning framework. The end impedance controller adjusts the grinding trajectory of the grinding arm through real-time force feedback based on the grinding reference trajectory generated by the input grinding reference trajectory generation submodule, including: The end impedance controller adjusts the grinding trajectory of the grinding arm through real-time force feedback. The end impedance controller is as follows: ； ； in, , and The matrix is ​​a diagonal matrix, representing the mass, damping, and stiffness matrices of the spring-damped-mass system, respectively. The control objective of the force control subspace is selected by the diagonal matrix. Decide, The value can be 0 or 1. Representing the task space degrees of freedom It is the absolute value of the contact force error response. This represents the reference pose under no contact force. The desired pose under the influence of contact force. Indicates the current actual pose With reference pose The pose error between them; and Let the acceleration and velocity represent the pose error, respectively. The acceleration is expressed as: ； Joint control amount for: ； Acceleration of the end effector of the grinding arm under force-position hybrid control based on external force determination And by selecting the matrix Adjust the forces and moments within the force-controlled subspace; The entire control cycle of the end effector impedance controller of the dual-arm robot is within the control period. Inside, the desired joint angle is obtained from the reference trajectory. and angular velocity , and the actual joint angle and angular velocity Comparison using forward kinematics and Jacobian matrix Calculate the pose error and the speed of pose error And obtain the acceleration of the pose error. To obtain the acceleration of the pose error Then, using forward kinematics and the Jacobian matrix... pseudo-inverse Jacobian matrix Calculate the joint velocity induced by external force. Then, the joint speed With the desired joint angular velocity Add them together to get the actual joint velocity, integrate the actual joint velocity, and then compare it with the desired joint angle. Add them together to obtain the final joint control amount. Then, the joint control amount The signal is sent to the grinding main arm of the dual-arm robot, thus completing one control cycle.

6. The dual-arm robotic powder grinding system for automated chemical laboratories according to claim 5, characterized in that, If the determined subsequent operation is a gathering operation, the main controller dynamically adjusts the motion trajectory of the grinding slave arm of the dual-arm robot according to the powder distribution information in the following manner, so that the grinding slave arm clamps the soft shovel to perform a gathering operation, collecting the scattered powder to the center of the mortar, including: The main controller calculates the positive polarity of the powder in the pixel coordinate system using the following formula. Axis, negative Axis, positive Axis and negative Dispersion index in the four directions of the axis , The formula is: ； Where i=1, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=2, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system The minimum value of the axis; when i=3, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=4, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system Minimum value of the axis; The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The maximum dispersion distance is set. When the dispersion index of the corresponding direction When the value is greater than 1, the powder in that direction is gathered. The main controller controls the grinding arm to pick up the soft shovel and gathers the powder in the corresponding direction in the mortar according to the direction obtained above and through the given gathering trajectory. If a fixed sieving cycle is reached, the main controller controls the grinding process by using a powder sieving device to sieve the ground powder from the mortar held by the grinding arm, separating powder that meets the particle size requirements, including: A powder sieving device using a vibrating motor-driven screen is employed. The main controller controls the grinding arm to pour the powder from the mortar into the screen of the powder sieving device. The vibrating motors on both sides of the screen drive the screen to vibrate, shaking the small powder particles to the bottom tray, while the large powder particles remain above the screen, thus completing the separation of powders of different particle sizes. Afterward, the grinding arm is controlled to hold the screen and pour the large powder particles back into the mortar along the guide groove, completing the sieving operation.

7. A dual-arm robot powder grinding method for automated chemical laboratories, characterized in that, The system according to any one of claims 1-6 comprises: Step 1: After receiving the start command, the system extracts local color features from the RGB image captured by the RGBD camera through the main controller, performs global positioning of the powder in the mortar, and determines the subsequent operation of the dual-arm robot based on the dispersion of the powder. If the determined subsequent operation is a grinding operation, then step 2 is executed; if the determined subsequent operation is a gathering operation, then step 3 is executed; if the determined subsequent operation is a sieving operation, then step 4 is executed; if it is determined that the grinding time is greater than the preset time, then step 5 is executed. Step 2: Based on the inner surface model of the mortar, the main controller generates a reference trajectory in the Cartesian coordinate system and optimizes the path using linear interpolation. The grinding arm of the dual-arm robot is controlled to adjust the motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. After the grinding operation is completed, proceed to step 1. Step 3: The main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information, so that the grinding arm holds the soft shovel to perform a gathering operation, and gathers the scattered powder to the center of the mortar. After the gathering operation is completed, proceed to step 1. Step 4: The grinding arm holds the mortar and uses a powder sieving device to sieve the ground powder, separating the powder that meets the particle size requirements. After the sieving operation is completed, proceed to step 1. Step 5: Complete the grinding task.

8. The dual-arm robot powder grinding method for automated chemical laboratories according to claim 7, characterized in that, In step 1, after receiving the start command, the system extracts local color features from the RGB image captured by the RGBD camera using the main controller, performs global localization of the powder in the mortar, and determines the subsequent operations of the dual-arm robot based on the dispersion of the powder, including: Step 11, RGB image captured by RGBD camera The sampling area was obtained by sampling the powder aggregation area in the sample. for: ； in, The pixel coordinates of the center of the inner surface of the mortar in the RGB image; Sampling area The pixel radius; Calculate the sampling area RGB mean of internal powder With covariance matrix The average RGB value of the powder With covariance matrix This indicates the color range of the powder in the mortar, RGB average. With covariance matrix The calculation formulas are as follows: ； ； in, It is the RGB vector of the image; Step 12, process the RGB vectors of the RGB image. The pixels that satisfy the color range of the powder in the mortar obtained in step 11 are processed to obtain a binarized image. for: ； in, RGB image pixels Mahalanobis distance; To set a threshold; This is an indicator function for converting an RGB image into a binary image. If a pixel meets a set threshold, it is set to 1 to represent white; otherwise, it is set to 0 to represent black. Define a binarized image Set of connected white regions within for: ； From the set The smallest circumcircle of the connected region with the largest area is selected from the options. As the final powder positioning range, the smallest circumcircle Represented as: ； in, for The smallest enclosing circle; Step 13, using the pixel coordinates of the center of the inner surface of the mortar in the RGB image. The powder positioning range obtained in step 12 The furthest distance and the set maximum dispersion distance Calculate the dispersion index of the powder. for: ； According to the dispersion index The subsequent operations of the dual-arm robot are determined as follows, if the dispersion index is... When the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed, and the subsequent operation of the dual-arm robot is determined to be a gathering operation; if the dispersion index is greater than 1, it indicates that the powder distribution is relatively dispersed. If the value is less than or equal to 1, then if the powder is still concentrated, the subsequent operation of the dual-arm robot is to perform a grinding operation.

9. The dual-arm robot powder grinding method for automated chemical laboratories according to claim 8, characterized in that, In step 2, the main controller generates a reference trajectory in a Cartesian coordinate system based on the inner surface model of the mortar, and optimizes the path using linear interpolation. The main arm of the dual-arm robot is then controlled to adjust its motion trajectory in real time through force-position hybrid control to stabilize the grinding contact force and grind the powder. This includes: Step 21: The inner surface of the mortar is approximated by a semi-circular surface using the main controller to model the mortar's inner surface, thus obtaining the mortar's inner surface model. Represented as: ； in, The depth of the mortar; The radius of the model of the inner surface of the mortar; Model of the inner surface of the mortar coordinates on; Step 22: Based on the mortar inner surface model constructed by the sub-module of the force-controlled grinding module of the main controller, the grinding reference trajectory of the grinding main arm of the dual-arm robot is generated, specifically as follows: The tip of the grinding pestle is set to be perpendicular to the tangent plane of the inner surface of the mortar. This orientation is obtained by performing a ZYZ Euler transformation on the reference coordinate system. The specific Euler angles are expressed as follows: ； ； ； in, , and These are the Euler angles for roll, yaw, and pitch, respectively. The path points of the grinding motion are calculated using Euler angles and 6D pose interpolation, and the grinding reference trajectory of the grinding main arm of the dual-arm robot is generated using an open-source robot operation and motion planning framework. Step 23: Based on the grinding reference trajectory generated by the grinding reference trajectory generation sub-module of the force control grinding module of the main controller, the grinding trajectory of the grinding arm is adjusted through real-time force feedback, specifically as follows: The end impedance controller of the force-controlled grinding module of the main controller adjusts the grinding trajectory of the grinding arm through real-time force feedback. The end impedance controller of the force-controlled grinding module of the main controller is as follows: ； ； in, , and The matrix is ​​a diagonal matrix, representing the mass, damping, and stiffness matrices of the spring-damped-mass system, respectively. The control objective of the force control subspace is selected by the diagonal matrix. Decide, The value can be 0 or 1. Representing the task space degrees of freedom It is the absolute value of the contact force error response. This represents the reference pose under no contact force. The desired pose under the influence of contact force. Indicates the current actual pose With reference pose The error between; and Let the acceleration and velocity represent the pose error, respectively. The acceleration is expressed as: ； Joint control amount for: ； Acceleration of the end effector of the grinding arm under force-position hybrid control based on external force determination And by selecting the matrix Adjust the forces and torques within the force-controlled subspace.

10. The dual-arm robot powder grinding method for automated chemical laboratories according to claim 9, characterized in that, In step 3, the main controller dynamically adjusts the motion trajectory of the grinding arm of the dual-arm robot according to the powder distribution information in the following manner, so that the grinding arm grips the soft shovel to perform a gathering operation, collecting the scattered powder to the center of the mortar, including: Step 31: The main controller calculates the positive coordinates of the powder in the pixel coordinate system according to the following formula. Axis, negative Axis, positive Axis and negative Dispersion index in the four directions of the axis , The formula is: ； Where i=1, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=2, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system The minimum value of the axis; when i=3, positive The dispersion index of axial orientation. = , It is the powder positioning range The positive inner pixel coordinate system The maximum value of the axis; when i=4, negative The dispersion index of axial orientation. = , It is the powder positioning range Negative inner pixel coordinate system Minimum value of the axis; The pixel coordinates of the center of the inner surface of the mortar in the RGB image; The maximum dispersion distance is set. Step 32, when the dispersion index of the corresponding direction When the value is greater than 1, the powder in that direction is gathered. The main controller controls the grinding arm to hold the soft shovel and gathers the powder in the corresponding direction in the mortar according to the direction obtained above and the given gathering trajectory. In step 4, the main controller controls the grinding process by using a powder sieving device to sieve the ground powder from the mortar held by the grinding arm, separating powder that meets the particle size requirements, including: A powder sieving device using a vibrating motor-driven screen is employed. The main controller controls the grinding arm to pour the powder from the mortar into the screen of the powder sieving device. The vibrating motors on both sides of the screen drive the screen to vibrate, shaking the small powder particles to the bottom tray, while the large powder particles remain above the screen, thus completing the separation of powders of different particle sizes. Afterward, the grinding arm is controlled to hold the screen and pour the large powder particles back into the mortar along the guide groove, completing the sieving operation.

Images