Energy storage battery cell test automated sampling method and system based on robotic arm

Abstract

The present invention discloses an automatic sampling method for energy storage battery cell testing based on a robotic arm, which relates to the field of automation technology, including obtaining cell space parameters; the robotic arm control unit calculates an obstacle avoidance trajectory for grabbing the cell according to the cell space parameters; the robotic arm tool end moves to the top of the cell according to the obstacle avoidance trajectory, and performs vacuum adsorption to grab the cell; after the grabbing is completed, the robotic arm tool end transports the cell to a test area and outputs a test ready signal; when the control system receives the test ready signal, it performs a cell capacitance and resistance value test; after the test is completed, the robotic arm tool end transports the cell to a corresponding storage station according to the classification of the test results, and completes the automatic sampling of the battery cell test. The whole process of grabbing, transporting, testing and classified storage of the cell is automatically completed by the robotic arm, reducing manual intervention and significantly improving production efficiency.

Description

Technical Field

[0001] The invention relates to the technical field of test automation, and in particular to an energy storage battery cell test automation sampling method based on a robotic arm. Background Art

[0002] At present, the process of energy storage battery cell testing generally adopts manual loading and unloading or simple automated equipment to transport and test the cells. Manual operation has problems such as low efficiency, easy errors, poor consistency, and long-term repetitive operations can easily cause worker fatigue, affecting the accuracy of test results; simple automated equipment, although improving efficiency, lacks intelligent perception and precise positioning capabilities, and cannot achieve adaptive grasping and flexible placement of cells. At the same time, improper control can easily cause damage to the cells during the cell handling process.

[0003] Traditional automated testing equipment usually uses fixed fixtures or a single adsorption structure, which cannot adjust the gripping strategy in real time according to the battery cell posture, and lacks effective status monitoring and fault protection mechanisms. In addition, existing equipment often separates the battery cell positioning, gripping, handling and testing links, and insufficient information sharing between links makes it difficult to achieve intelligent collaborative control of the entire process, resulting in difficulty in fundamentally improving test efficiency and reliability. Summary of the invention

[0004] In view of the deficiencies in automation, intelligence and reliability of existing energy storage battery cell testing methods, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is how to achieve visual positioning, intelligent grabbing and handling, and reliable test docking during the energy storage battery cell testing process, thereby improving the automation level and test efficiency of the cell testing.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In the first aspect, an embodiment of the present invention provides an automated sampling method for energy storage battery cell testing based on a robotic arm, which includes obtaining cell space parameters and transmitting the cell space parameters to a robotic arm control unit, wherein the cell space parameters include position parameters and posture parameters; the robotic arm control unit calculates an obstacle avoidance trajectory for grasping the cell according to the cell space parameters, and configures a vacuum adsorption device on the robotic arm tool end; the robotic arm tool end moves to the top of the cell according to the obstacle avoidance trajectory, performs vacuum adsorption to grasp the cell, and monitors the adsorption status in real time; after completing the grasping, the robotic arm tool end transports the cell to a test area, outputs a test ready signal, and completes automatic docking of an electrical performance test; when the control system receives the test ready signal, it performs cell capacitance and resistance value tests; after the test is completed, the robotic arm tool end moves the cell to a corresponding storage station according to the classification of the test results, and completes automated sampling of the battery cell test.

[0008] As a preferred solution of the automatic sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, wherein: obtaining the spatial parameters of the cell includes the following steps: setting an image acquisition device in the cell loading area; configuring an adjustable light source system, and adjusting the illumination according to the reflective characteristics of the cell surface; acquiring image data of the cell through the image acquisition device, and preprocessing the image data; performing feature recognition on the preprocessed image data to extract cell feature data; calculating the position parameters and posture parameters of the cell based on the cell feature data; performing reliability verification on the position parameters and posture parameters, and encapsulating the verified qualified position parameters and posture parameters into a data packet; and transmitting the data packet to the robotic arm control unit through an industrial communication bus.

[0009] As a preferred solution of the automated sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, the calculation of the obstacle avoidance trajectory for grabbing the battery cell includes the following steps: establishing a robotic arm kinematic model according to the DH parameters of the robotic arm to obtain the global kinematic equation of the robotic arm; based on the global kinematic equation of the robotic arm, a hybrid solution strategy is used to solve the joint angle parameters; based on the joint angle parameters, a damped least squares method is used to solve the joint angular velocity parameters; based on the joint angle parameters and the joint angular velocity parameters, an RRT-Connect algorithm is used for path planning to obtain an initial path sequence; based on the initial path sequence, a cubic B-spline algorithm is used for trajectory smoothing to obtain an obstacle avoidance trajectory.

[0010] As a preferred solution of the automated sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, the method comprises the following steps: the tool end of the robotic arm moves to the top of the cell according to the obstacle avoidance trajectory; during the descent of the tool end of the robotic arm, the relative distance between the tool end of the robotic arm and the cell is detected in real time, and the descent speed is adjusted according to the relative distance; the vacuum adsorption mechanism adopts a segmented suction control strategy to perform the adsorption operation; the force and torque during the adsorption process are detected by a force detection device, and the adsorption state is judged according to the detection results; the multiple vacuum suction cups of the vacuum adsorption mechanism are differentially controlled, and the negative pressure value of each group of vacuum suction cups is independently adjusted by a proportional valve; the monitoring parameters during the adsorption process are recorded, and the monitoring parameters are compared with the reference curve in real time; when any monitoring parameter is detected to be out of the preset range, the control system performs protection control.

[0011] As a preferred solution of the automated sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, the automatic docking for completing the electrical performance test includes the following steps: controlling the robotic arm tool end to transfer the cell to the test area; when the robotic arm tool end enters the test area, the robotic arm control unit triggers the visual positioning system to perform position calibration and generate a position compensation instruction; according to the position compensation instruction, an adaptive PID control algorithm is used to adjust the posture of the robotic arm tool end; after the position compensation is completed, an adaptive contact control strategy based on torque sensing is used to make the cell contact smoothly with the detection array; the robotic arm control unit monitors the contact state between the cell and the detection array in real time, and performs pre-test state confirmation; during the test, the robotic arm control unit performs posture stabilization control and state monitoring, and releases the cell after receiving a test completion signal; after each test cycle, the robotic arm control unit generates a test process data packet and uploads it to the robotic arm control unit.

[0012] As a preferred solution of the automatic sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, wherein: executing cell capacitance and resistance value tests includes the following steps: executing open circuit voltage detection, capacity test, internal resistance test, and generating test data; performing Kalman filtering on the test data, and calculating Coulomb efficiency and energy efficiency; based on the test data, using a fuzzy comprehensive evaluation method to score the cell performance and generate a scoring result; dividing the cell into three levels of A, B, and C according to the comprehensive scoring result; and generating a test data report.

[0013] As a preferred solution of the automated sampling method for energy storage battery cell testing based on a robotic arm described in the present invention, wherein: the vacuum adsorption device configured at the tool end of the robotic arm includes: installing an adsorption mechanism having a plurality of annularly distributed vacuum suction cups at the tool end of the robotic arm; and arranging a shock-absorbing buffer layer of silicone material on the adsorption surface of the adsorption mechanism.

[0014] In the second aspect, an embodiment of the present invention provides an automatic sampling system for energy storage battery cell testing based on a robotic arm, which includes a parameter acquisition module for acquiring cell space parameters and transmitting the cell space parameters to a robotic arm control unit, the cell space parameters including position parameters and posture parameters; a path planning module for calculating an obstacle avoidance trajectory for grabbing the cell according to the cell space parameters, and configuring a vacuum adsorption device on the tool end of the robotic arm; a grabbing execution module for moving to the top of the cell according to the obstacle avoidance trajectory, grabbing the cell by vacuum adsorption, and monitoring the adsorption status in real time; a test docking module for transporting the cell to a test area, outputting a test ready signal, and completing automatic docking for electrical performance testing; a performance testing module for executing cell capacitance and resistance value tests upon receiving a test ready signal; a classification storage module for transporting the cell to a corresponding storage station according to the classification of the test results, and completing automatic sampling of the battery cell test.

[0015] The beneficial effects of the present invention are as follows: the present invention realizes accurate recognition and spatial positioning of battery cell features by setting a first camera and a second camera in cooperation with an adjustable light source system, adopts a vacuum adsorption mechanism in cooperation with a silicone shock-absorbing buffer layer and a multi-channel pressure sensor to improve the stability and safety of battery cell grasping, combines a trajectory generation method using quintic polynomial trajectory planning and cubic spline optimization to ensure the stability of the battery cell conveying process, realizes accurate and smooth contact between the battery cell and the detection array through a torque sensor and a visual positioning system, and finally realizes the automation and reliability of the entire battery cell testing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 The framework flow chart of the automated sampling method for energy storage battery cell testing based on a robotic arm.

[0018] Figure 2 Generate a flow chart for obstacle avoidance trajectories for the robotic arm-based automated sampling method for energy storage battery cell testing.

[0019] Figure 3 The transfer test flow chart of the automated sampling method for energy storage battery cell testing based on a robotic arm. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0023] Example 1, reference Figure 1~Figure 3 , which is the first embodiment of the present invention, and provides an automatic sampling method for energy storage battery cell testing based on a robotic arm. The framework flow chart is as follows Figure 1 As shown, including

[0024] S1: Acquire the cell space parameters and transmit the cell space parameters to the robot arm control unit.

[0025] Among them, the cell space parameters include position parameters and posture parameters. The specific implementation plan is as follows:

[0026] S1.1: An image acquisition device is set up in the battery cell loading area.

[0027] Among them, the image acquisition device includes a first camera and a second camera, the first camera is vertically arranged directly above the loading area for collecting the plane features of the battery cells, the second camera is horizontally arranged on the side of the loading area for collecting the height features of the battery cells, and the field of view of the first camera and the second camera covers the entire loading area.

[0028] S1.2: Configure an adjustable light source system, wherein the light source system adjusts the illumination according to the reflective characteristics of the battery cell surface.

[0029] Specifically, the illumination value of the light source system can be adjusted within the range of 0-1000 Lux, and includes a ring-shaped light source for providing uniform surface illumination and a strip-shaped light source for providing lateral illumination.

[0030] S1.3: Acquire image data of the battery cell through the image acquisition device, and pre-process the image data.

[0031] Specifically, the image acquisition device performs multi-angle imaging of the battery cell according to preset exposure parameters, obtains battery cell surface image data (including grayscale image and depth image), uses a Gaussian filtering algorithm to eliminate noise from the image data, and uses a histogram equalization method to enhance contrast of the image data.

[0032] S1.4: Perform feature recognition on the preprocessed image data to extract battery cell feature data.

[0033] The cell feature data includes cell contour features, cell tab position features, and cell surface defect features. In addition, the feature recognition is implemented using a deep learning target detection algorithm.

[0034] S1.5: Based on the battery cell characteristic data, calculate the position parameters and posture parameters of the battery cell.

[0035] Among them, the position parameters include the three-dimensional coordinates of the center point of the battery cell (X, Y, Z) and the three-dimensional coordinates of the center point of the tab (X', Y', Z'), and the posture parameters include the Euler angles (α, β, γ) of the battery cell relative to the reference coordinate system.

[0036] Moreover, the calculation process of position parameters and attitude parameters is as follows: first, the camera internal and external parameter matrix is ​​obtained through binocular vision calibration, then the three-dimensional coordinates of the feature points are solved using the PnP algorithm, and finally the rigid body transformation matrix is ​​calculated through SVD decomposition to obtain the attitude angle.

[0037] S1.6: Perform reliability verification on the position parameters and posture parameters, and encapsulate the verified qualified position parameters and posture parameters into a data packet.

[0038] Specifically, the reliability verification includes the following steps: performing validity check on the position parameters to ensure that the coordinate values ​​are within the preset working space; performing continuity check on the attitude parameters to ensure that the attitude angle change between two adjacent samples does not exceed the preset threshold; and using Kalman filtering to smooth the position parameters and the attitude parameters to eliminate random fluctuations.

[0039] S1.7: Transmit the data packet to the robot control unit via the industrial communication bus.

[0040] It should be noted that the transmission process uses a redundant verification mechanism to ensure data integrity. After receiving the data packet, the robotic arm control unit parses and stores it through the JSON parser.

[0041] S2: The robotic arm control unit calculates an obstacle avoidance trajectory for grabbing the battery cell according to the battery cell space parameters.

[0042] In order to achieve precise grasping, the tool end of the robot arm is equipped with a vacuum adsorption mechanism, and its adsorption surface is provided with a shock-absorbing buffer layer. The specific implementation plan is as follows:

[0043] S2.1: The robotic arm control unit adopts a motion planning algorithm to convert position parameters and posture parameters into real-time control quantities of each joint of the robotic arm.

[0044] Specifically, the obstacle avoidance trajectory generation flow chart is as follows Figure 2 As shown, the kinematic model of the robot arm is established according to the DH parameters of the robot arm, and the DH parameters include: connecting rod offset d, connecting rod length a, connecting rod torsion angle and joint angle Based on the DH parameters, a homogeneous transformation matrix between adjacent joints is established to obtain the global kinematic equation of the manipulator.

[0045] Furthermore, based on the global kinematics equation of the robot, a hybrid solution strategy is used to solve the joint angle parameters: the first three axes use the geometric analysis method to solve the position, and the back three axes use the iterative method to solve the posture, and the six joint angle parameters of the robot are obtained. Among them, the solution accuracy of the angle parameters of the first three axes is better than 0.1°, and the solution accuracy of the angle parameters of the rear three axes is better than 0.2°.

[0046] Furthermore, based on the joint angle parameters, the damped least squares method is used to solve the joint angular velocity parameters: the Jacobian matrix J is constructed and the damping factor is introduced. , solve the equation , get the angular velocity parameters of each joint .in is the joint velocity vector (indicating the angular velocity parameters of each joint ), I is the identity matrix, and its dimension is Same, v is the target end effector velocity vector, damping factor Adaptively adjust according to the robot arm configuration to avoid singular configurations.

[0047] Furthermore, based on the joint angle parameters and the joint angular velocity parameters, the improved RRT-Connect algorithm is used for path planning to obtain an initial path sequence, specifically including: constructing a starting point tree T in the workspace a and the terminal tree T b ; Adopt adaptive step size from T a and T b Alternately expand nodes; use KD tree to accelerate nearest neighbor search; find the connection nodes and get the initial path sequence. The collision detection accuracy of the planned path is not less than 5mm.

[0048] Furthermore, based on the initial path sequence, a cubic B-spline algorithm is used to smooth the trajectory to obtain a final obstacle avoidance trajectory: key points in the initial path sequence are selected as control points; and B-spline basis functions are calculated to generate a smooth trajectory curve.

[0049] It should be noted that this hybrid solution strategy combines the efficiency of the geometric method and the robustness of the iterative method, and can quickly and reliably complete the inverse kinematics solution. The improved RRT-Connect algorithm has strong real-time and adaptability, and can generate reasonable obstacle avoidance paths in complex working environments.

[0050] S2.2: A suction mechanism having a plurality of annularly distributed vacuum suction cups is installed at the tool end of the robot arm.

[0051] In terms of mechanical structure design, the tool end of the robot arm is provided with a quick-change interface with a flange, and the quick-change interface is connected and fixed to the vacuum adsorption mechanism by bolts, and the body of the vacuum adsorption mechanism is made of aluminum alloy. The adsorption surface of the vacuum adsorption mechanism is evenly distributed with 8 vacuum suction cups along the circumference, and each vacuum suction cup is connected to the vacuum generator through an independent vacuum pipeline, and the vacuum generator uses a proportional regulating valve to control the negative pressure value.

[0052] S2.3: A shock-absorbing buffer layer made of silicone material is provided on the adsorption surface of the vacuum adsorption mechanism.

[0053] To improve the gripping stability, the suction surface of the vacuum suction cup is equipped with a shock-absorbing buffer layer made of silicone material, which has good elastic deformation ability. Through the molding process, the surface of the shock-absorbing buffer layer is made into a microscopic honeycomb structure, which effectively increases the contact area with the surface of the battery cell and improves the adsorption stability.

[0054] S2.4: A pressure sensor is set to monitor the negative pressure value of each vacuum suction cup in real time, and an alarm signal is triggered when the negative pressure value is lower than a preset safety threshold.

[0055] To ensure system reliability, the vacuum adsorption mechanism is equipped with multiple pressure sensors to monitor the negative pressure value of each vacuum suction cup respectively. When the negative pressure values ​​sampled three times in a row are lower than the preset safety threshold, an alarm signal is triggered.

[0056] S2.5: A torque sensor is installed at the tool end of the robot arm, and a deceleration and stop procedure is automatically executed when an abnormal contact force is detected.

[0057] The deceleration and stop procedure is as follows: reduce the speed of the robot arm joint to 20% of the current speed, maintain this speed for 50ms and then stop completely, while cutting off the vacuum generator gas source to ensure the safety of the battery cells and equipment.

[0058] S3: The tool end of the robot arm moves to the top of the battery cell according to the obstacle avoidance trajectory, performs vacuum adsorption to grab the battery cell, and monitors the adsorption status in real time.

[0059] Specifically, the following steps are included:

[0060] S3.1: The tool end of the robot arm moves to the top of the battery cell according to the obstacle avoidance trajectory, and automatically switches to low speed mode when approaching the top of the battery cell.

[0061] Specifically, deceleration points are set at 100 mm and 50 mm away from the battery cell, and the moving speed of the tool end is gradually reduced through three-stage deceleration to ensure that the tool end of the robot arm approaches the battery cell smoothly.

[0062] S3.2: During the descent of the tool end of the robot arm, the relative distance between the tool end of the robot arm and the battery cell is detected in real time, and the descent speed is adjusted according to the relative distance.

[0063] Specifically, the relative distance to the battery cell is detected in real time by a laser ranging sensor. When the relative distance is less than 20 mm, the precise positioning mode is activated and the descent speed is further reduced.

[0064] S3.3: The vacuum adsorption mechanism adopts a segmented suction control strategy to perform the adsorption operation.

[0065] The control system of the vacuum adsorption mechanism adopts a segmented suction control strategy. In the first stage, the pre-inflation is performed at a distance of 5 mm from the surface of the battery cell. The pre-inflation pressure value is the initial negative pressure value and the duration is 200ms. When the shock-absorbing buffer layer first contacts the surface of the battery cell, the control system enters the second stage, and the negative pressure value gradually increases to the maximum negative pressure value in a linear increasing manner to achieve flexible bonding.

[0066] S3.4: Detect the force and torque during the adsorption process through a force detection device, and determine the adsorption state based on the detection results.

[0067] The force detection device includes a three-dimensional force sensor and a torque sensor. The three-dimensional force sensor detects the pressure value in the Z-axis direction, and the torque sensor detects the torsional torque in the X-axis and Y-axis directions. The control system collects data from the mechanical detection unit in real time. When the Z-axis pressure value is stable within the rated pressure range and the X-axis and Y-axis torque values ​​are both less than the maximum allowable torque, the adsorption state is determined to be good.

[0068] S3.5: Differentiately control the multiple vacuum suction cups of the vacuum adsorption mechanism, and independently adjust the negative pressure value of each group of vacuum suction cups through a proportional valve.

[0069] Specifically, the suction cups of the vacuum adsorption mechanism are divided into a center group and an edge group according to the position of the center of gravity of the battery cell for differentiated control. The negative pressure value of each group of vacuum suction cups is adjusted by an independent proportional valve. The response time of the proportional valve is less than the control cycle, thereby achieving precise pressure distribution.

[0070] S3.6: Record the monitoring parameters during the adsorption process, and compare the monitoring parameters with the reference curve in real time.

[0071] The monitoring parameters include the adsorption force curve, the approach displacement curve and the force feedback data. The reference curve is obtained through the standard working condition test, which includes the standard range and fluctuation limit of each parameter.

[0072] S3.7: When it is detected that any monitoring parameter exceeds the preset range, the control system starts the emergency protection program and executes protection control.

[0073] Among them, when the adsorption force, displacement or force exceeds the allowable range, the emergency protection program is triggered. The emergency protection program executes actions in the order of pressure relief, lifting and evacuation, and sends out an alarm signal.

[0074] S4: After completing the grabbing, the tool end of the robot arm transports the battery cell to the test area, outputs a test ready signal, and completes the automatic docking of the electrical performance test.

[0075] Among them, the test area is provided with a detection array for detecting the electrical performance of the battery cell, the contact position of the detection array corresponds to the electrode terminal position of the battery cell, and the robot arm tool end places the battery cell in the detection array according to the preset test position parameters.

[0076] Specifically, the transmission test flow chart is as follows: Figure 3 As shown, the following steps are included:

[0077] S4.1: The robot control unit generates a transmission trajectory according to the cell space parameters and the test area position parameters using a quintic polynomial trajectory planning algorithm, and controls the robot tool end to transfer the cell to the test area.

[0078] Among them, the steps for generating the transmission trajectory are as follows: divide the motion process into three sections: acceleration, uniform speed, and deceleration, and set the position, velocity, and acceleration boundary conditions at the starting and ending points of each section; use a quintic polynomial to fit each section of the trajectory, and solve the polynomial coefficients through the boundary conditions; use cubic B-splines for smooth connection in the transition area between sections to ensure the continuity of the acceleration; set the speed upper limit and acceleration upper limit according to the dynamic characteristics of the robot arm.

[0079] It should be noted that this trajectory planning scheme can ensure the smooth movement of the robot arm when carrying the battery cell, avoiding vibration and unstable grasping caused by sudden changes in acceleration.

[0080] S4.2: When the tool end of the robotic arm enters the test area, the robotic arm control unit triggers the visual positioning system to perform position calibration and generate a position compensation instruction.

[0081] Specifically, the visual positioning system includes a 3D vision sensor and adopts a multi-feature fusion positioning algorithm: extracts multiple features such as the edge, tab, and identification point of the battery cell; uses the RANSAC algorithm to eliminate mismatched points; uses the weighted ICP algorithm to align the current posture with the target posture; calculates the posture compensation amount; and ultimately achieves a positioning accuracy better than 0.05mm.

[0082] S4.3: According to the position compensation instruction, an adaptive PID control algorithm is used to adjust the posture of the tool end of the robot arm.

[0083] The adjustment process uses a cascade PID control structure: the outer loop is the position loop, and the inner loop is the speed loop. The position loop uses an adaptive PID algorithm: fuzzy rules are designed based on the position error e to adaptively adjust the PID parameters; when | Increase K p To improve response speed; when Increase K i To eliminate static error; introduce feedforward compensation to reduce tracking error. The speed loop adopts PI control and adjusts parameters according to the dynamic characteristics of the robot arm. The control cycle is 2ms and the position sampling cycle is 10ms.

[0084] S4.4: After the position compensation is completed, an adaptive contact control strategy based on torque sensing is adopted to make the battery cell contact smoothly with the detection array.

[0085] Specifically, torque sensors are installed at the joints of the robotic arm to estimate the end contact state in real time through the joint torque. A hierarchical control structure is adopted, with the upper layer being for contact state judgment and task planning, and the lower layer being for torque feedforward compensation control. The contact speed and stiffness parameters are adaptively adjusted according to the torque change rate to achieve stable and compliant contact.

[0086] S4.5: The robot arm control unit monitors the contact status between the battery cell and the detection array in real time and performs pre-test status confirmation.

[0087] The contact state monitoring includes: contact resistance detection, contact pressure detection and contact position detection. The contact resistance monitoring adopts the four-wire method, the sampling frequency is 1kHz, the threshold is less than 10mΩ and the fluctuation range is less than ±1mΩ; the contact pressure monitoring adopts the torque sensor, the sampling frequency is 2kHz, the pressure range is 1.8N to 2.2N, and the fluctuation is less than ±0.1N; the contact position monitoring combines the visual system and the torque sensor data, the XY plane position deviation is less than 0.05mm, and the attitude angle deviation is less than 0.1°.

[0088] When all monitoring parameters meet the preset conditions, the robot control unit sends a test ready signal to the control system via industrial real-time Ethernet.

[0089] S4.6: During the test, the robot arm control unit performs posture stabilization control and status monitoring, and executes the battery cell release procedure after receiving the test completion signal.

[0090] Specifically, the battery cell release adopts a three-order gradient control strategy: in the first stage, the negative pressure value is reduced to 80% of the initial negative pressure value (-64kPa) and lasts for 100ms; in the second stage, it is reduced to 50% of the initial negative pressure value (-40kPa) and lasts for 100ms; in the third stage, the pressure is completely released (0kPa) to allow the battery cell to detach smoothly.

[0091] S4.7: After each test cycle is completed, the robot control unit generates a test process data packet and uploads it to the robot control unit.

[0092] S5: When the control system receives the test ready signal, it performs a cell capacity and resistance value test.

[0093] Specifically, perform open circuit voltage detection, the open circuit voltage detection uses a digital voltmeter to measure the voltage at the positive and negative terminals of the battery cell, performs continuous sampling at a preset sampling frequency, and takes the average value as the open circuit voltage value. Perform capacity testing, the capacity testing uses a constant current charge and discharge method, specifically including: constant current charging at a preset rate to the charging cut-off voltage, and standing for a preset time; constant current discharging at a preset rate to the discharge cut-off voltage, and standing for a preset time; recording the voltage, current and time data of the charging and discharging process. Perform an internal resistance test procedure, apply an AC signal of a preset frequency to the battery cell for an AC internal resistance test; apply a pulse current of a preset pulse width to the battery cell for a DC internal resistance test.

[0094] Furthermore, the test data is processed by Kalman filtering to calculate the Coulomb efficiency and energy efficiency. The processed test data is stored in the battery cell performance database, and the battery cell performance database establishes a corresponding relationship between the battery cell number and the test data. Based on the test data, the fuzzy comprehensive evaluation method is used to score the battery cell performance and generate a scoring result, wherein the scoring indicators include internal resistance consistency (e.g., weight 0.3), capacity retention rate (e.g., weight 0.3), charge and discharge efficiency (e.g., weight 0.2) and temperature characteristics (e.g., weight 0.2). According to the comprehensive scoring results, the battery cells are divided into three levels: A, B, and C, among which Class A (e.g., score ≥ 85 points) has excellent performance and can be used directly, Class B (e.g., 70 points ≤ score < 85 points) has slight performance deviations and needs to be re-inspected, and Class C (e.g., score < 70 points) does not meet the performance standards and needs to be scrapped.

[0095] Furthermore, a test data report is generated, including: original data of electrical performance test, statistical analysis results, classification level determination results and suggested processing opinions.

[0096] S6: After the test is completed, the robot arm tool end moves the battery cells to the corresponding storage station according to the classification of the test results, completing the automatic sampling of the battery cell test.

[0097] Specifically, the target storage station is determined according to the classification result, and the A, B, and C grade battery cells correspond to three independent storage areas respectively, and each area is provided with a multi-layer pallet structure; the control system reads the three-dimensional coordinates and posture parameters of the target station, and uses a quintic polynomial trajectory planning algorithm to generate a storage trajectory; the robot tool end executes the battery cell release action, and adopts a progressive pressure relief method: the robot tool end executes the battery cell release program, and adopts a third-order gradient control strategy for release; the placement status of the battery cell is detected by the visual positioning system, and the placement is determined to be successful when the position deviation is less than 0.05mm; after the robot arm completes the battery cell placement, the robot tool end returns to the preparation position and updates the pallet status information at the same time, and the pallet status information includes the number of placed battery cells; the control system sends a test completion signal to the robot control unit via the industrial real-time Ethernet.

[0098] It should be noted that only the quintic polynomial trajectory planning technology solution is used in this stage. This is mainly because the storage process path is simple and fixed. The quintic polynomial can provide sufficient smoothness and has a small amount of calculation. It can efficiently complete the precise storage of battery cells and meet the accuracy requirements of classified placement.

[0099] Furthermore, the present embodiment also provides an automatic sampling system for energy storage battery cell testing based on a robotic arm, including a parameter acquisition module, used to acquire cell space parameters and transmit the cell space parameters to a robotic arm control unit, the cell space parameters including position parameters and posture parameters; a path planning module, used to calculate an obstacle avoidance trajectory for grabbing the cell according to the cell space parameters, and configure a vacuum adsorption device on the tool end of the robotic arm; a grabbing execution module, used to move to the top of the cell according to the obstacle avoidance trajectory, perform vacuum adsorption to grab the cell, and monitor the adsorption status in real time; a test docking module, used to transport the cell to a test area, output a test ready signal, and complete automatic docking for electrical performance testing; a performance testing module, used to perform cell capacitance and resistance value testing upon receiving a test ready signal; a classification storage module, used to transport the cell to a corresponding storage station according to the classification of the test results, and complete automatic sampling of the battery cell test.

[0100] In summary, the present invention realizes accurate identification and spatial positioning of battery cell features by setting the first camera and the second camera in conjunction with an adjustable light source system, adopts a vacuum adsorption mechanism in conjunction with a silicone shock-absorbing buffer layer and a multi-channel pressure sensor to improve the stability and safety of battery cell grabbing, combines the trajectory generation method of quintic polynomial trajectory planning and cubic spline optimization to ensure the stability of the battery cell conveying process, and realizes accurate and smooth contact between the battery cell and the detection array through a torque sensor and a visual positioning system, ultimately realizing the automation and reliability of the entire battery cell testing process. The entire process of grabbing, transporting, testing, and classified storage of battery cells is automatically completed by a robotic arm, reducing manual intervention and significantly improving production efficiency.

[0101] Example 2, reference Figure 1~Figure 3 , which is the second embodiment of the present invention, provides an automated sampling method for energy storage battery cell testing based on a robotic arm. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0102] To verify the feasibility and superiority of the present invention, the researchers conducted a three-month experimental verification on the battery production line of a new energy technology company. The experiment used a six-axis industrial robot equipped with an industrial camera system with a resolution of 2048×1536 pixels and a structured light sensor with an accuracy of 0.01mm. The test sample was a 21700 cylindrical battery with a rated capacity of 5000mAh, and a total of 10,000 cells were processed.

[0103] During the image acquisition phase, the researchers used a 1000 Lux ring light source and a 500 Lux strip light source to achieve a 99.7% accuracy rate in identifying battery cell features. The image preprocessing used a Gaussian filter with σ=1.2, and the image contrast was increased by 45% after histogram equalization. The deep learning target detection algorithm achieved a detection accuracy rate of 98.9% on a test set of 1,000 samples, with an average processing time of 47ms.

[0104] The calculation results of the position parameters show that the measurement accuracy of the center point of the battery cell reaches ±0.03mm, the measurement accuracy of the tab position reaches ±0.02mm, and the measurement accuracy of the attitude angle reaches ±0.05°. In the 1000 repeated positioning tests, the position repetition accuracy is ±0.025mm, and the attitude repetition accuracy is ±0.03°. After Kalman filtering, the standard deviation of the measurement data is reduced by 67%.

[0105] In terms of robot arm trajectory planning, the improved RRT-Connect algorithm is used, and the average planning time of the generated obstacle avoidance trajectory is 85ms, and the trajectory smoothness error is less than 0.1mm / s 2The vacuum adsorption system uses a maximum negative pressure value of -80kPa and achieves a 98.4% grasping success rate through a segmented suction control strategy. During dynamic movement, the maximum vibration acceleration measured is 0.15g, which is much lower than the 0.45g of the traditional solution.

[0106] During the test docking process, the position compensation accuracy of the visual positioning system reached ±0.008mm, and the angle compensation accuracy reached ±0.02°. After adopting dynamic impedance control, the contact pressure between the battery cell and the test probe was stabilized within the range of 2.0±0.05N, and the contact resistance was always kept below 5mΩ. The average cycle time of the test process was 42 seconds, which was 63% more efficient than manual operation.

[0107] After three months of continuous operation tests, the system showed extremely high reliability and efficiency. Among the 10,000 cells processed, the pass rate of a single test reached 95.2%, and the repeatability error of the test data was less than 0.5%. The mean trouble-free operation time (MTBF) of the system reached 720 hours, far exceeding the expected target of 500 hours. In order to fully evaluate the advantages of the solution of the present invention, the researchers compared it in detail with the traditional method. The comparison results are shown in Table 1.

[0108] Table 1 Performance comparison between the present invention and the traditional method

[0109] Evaluation indicators The present invention Traditional methods Positioning accuracy ±0.03mm ±0.15mm Attitude angle accuracy ±0.05° ±0.3° Test cycle time 42s 114s One test pass rate 95.2% 87.3% Crawl success rate 98.4% 92.1% Measurement repeatability error Less than 0.5% Less than 2.3% Daily processing capacity 2000 740 ;

[0110] It can be seen from the comparative data in Table 1 that the solution of the present invention is superior to the traditional manual solution in all key performance indicators. In particular, in terms of positioning accuracy and attitude angle accuracy, they have been improved by 80% and 83.3% respectively, which directly leads to a significant improvement in test reliability. At the same time, the test cycle time was shortened by 63.2%, the daily processing volume was increased by 170.3%, and production efficiency was improved. The stability of the system has also been significantly improved, which reduces equipment maintenance requirements and downtime. On the whole, the solution of the present invention has achieved a qualitative leap in efficiency, accuracy and reliability, and provides a reliable technical solution for the automation upgrade of battery production lines.

[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An automatic sampling method for energy storage battery cell testing based on a robotic arm, characterized in that: include, Acquire cell space parameters, and transmit the cell space parameters to a robot control unit, wherein the cell space parameters include position parameters and posture parameters; The robot arm control unit calculates an obstacle avoidance trajectory for grabbing the battery cell according to the battery cell space parameters, and configures a vacuum adsorption device at the tool end of the robot arm; The tool end of the robot arm moves to the top of the battery cell according to the obstacle avoidance trajectory, performs vacuum adsorption to grab the battery cell, and monitors the adsorption status in real time; After the grabbing is completed, the tool end of the robot arm transports the battery cell to the test area, outputs a test ready signal, and completes the automatic docking of the electrical performance test; When the control system receives the test ready signal, it performs the cell capacity and resistance value test; After the test is completed, the robot arm tool end moves the battery cells to the corresponding storage station according to the classification of the test results, completing the automatic sampling of the battery cell test; The automatic docking for completing the electrical performance test comprises the following steps: Control the tool end of the robot arm to transfer the battery cells to the test area; When the tool end of the robot arm enters the test area, the robot arm control unit triggers the visual positioning system to perform position calibration and generate a position compensation instruction; According to the position compensation instruction, an adaptive PID control algorithm is used to adjust the posture of the tool end of the robot arm; After the position compensation is completed, an adaptive contact control strategy based on torque sensing is adopted to make the battery cell contact smoothly with the detection array; The robot control unit monitors the contact status between the battery cell and the detection array in real time and performs pre-test status confirmation; During the test, the manipulator control unit performs posture stabilization control and status monitoring, and releases the battery cell after receiving a test completion signal; After each test cycle is completed, the robot control unit generates a test process data packet and uploads it to the robot control unit.

2. The automatic sampling method for energy storage battery cell testing based on a robotic arm as claimed in claim 1, characterized in that: The obtaining of the cell space parameters comprises the following steps: An image acquisition device is provided in the battery cell loading area; An adjustable light source system is configured, and the light source system adjusts the illumination according to the reflective characteristics of the battery cell surface; Acquire image data of the battery cell through the image acquisition device, and pre-process the image data; Perform feature recognition on the preprocessed image data and extract cell feature data; Based on the battery cell characteristic data, calculating the position parameters and posture parameters of the battery cell; Performing reliability verification on the position parameters and posture parameters, and encapsulating the verified qualified position parameters and posture parameters into a data packet; The data packet is transmitted to the robot control unit via the industrial communication bus.

3. The automatic sampling method for energy storage battery cell testing based on a robotic arm as claimed in claim 1, characterized in that: The calculation of the obstacle avoidance trajectory for grabbing the battery cell comprises the following steps: The kinematic model of the robot arm is established according to the DH parameters of the robot arm, and the global kinematic equation of the robot arm is obtained; Based on the global kinematic equations of the robot, a hybrid solution strategy is used to solve the joint angle parameters; Based on the joint angle parameters, the damped least square method is used to solve the joint angular velocity parameters; Based on the joint angle parameters and the joint angular velocity parameters, an RRT-Connect algorithm is used to perform path planning to obtain an initial path sequence; Based on the initial path sequence, a cubic B-spline algorithm is used to smooth the trajectory to obtain an obstacle avoidance trajectory.

4. The automatic sampling method for energy storage battery cell testing based on a robotic arm as claimed in claim 1, characterized in that: The vacuum adsorption process for grabbing the battery cell comprises the following steps: The tool end of the robot arm moves to the top of the battery cell according to the obstacle avoidance trajectory; During the descent of the tool end of the robot arm, the relative distance between the tool end of the robot arm and the battery cell is detected in real time, and the descent speed is adjusted according to the relative distance; The vacuum adsorption mechanism adopts a segmented suction control strategy to perform the adsorption operation; The force and torque during the adsorption process are detected by a force detection device, and the adsorption state is determined based on the detection results; Differentiated control is performed on the multiple vacuum suction cups of the vacuum adsorption mechanism, and the negative pressure value of each group of vacuum suction cups is independently adjusted through a proportional valve; Recording monitoring parameters during the adsorption process, and comparing the monitoring parameters with a reference curve in real time; When it is detected that any monitoring parameter exceeds the preset range, the control system performs protection control.

5. The automatic sampling method for energy storage battery cell testing based on a robotic arm as claimed in claim 1, characterized in that: The method of performing the cell capacitance and resistance value test comprises the following steps: Perform open circuit voltage detection, capacity test, internal resistance test, and generate test data; Perform Kalman filtering on the test data and calculate the Coulomb efficiency and energy efficiency; Based on the test data, a fuzzy comprehensive evaluation method is used to score the performance of the battery cell to generate a scoring result; According to the comprehensive scoring results, the battery cells are divided into three levels: A, B, and C; Generate test data report.

6. The automatic sampling method for energy storage battery cell testing based on a robotic arm as claimed in claim 1, characterized in that: The vacuum adsorption device is configured at the tool end of the robot arm, comprising: A suction mechanism having a plurality of annularly distributed vacuum suction cups is installed at the tool end of the robot arm; A shock-absorbing and buffering layer made of silica gel is arranged on the adsorption surface of the adsorption mechanism.

7. An automatic sample loading system for energy storage battery cell testing based on a robotic arm, based on the automatic sample loading method for energy storage battery cell testing based on a robotic arm according to any one of claims 1 to 6, characterized in that: Also includes, A parameter acquisition module, used for acquiring cell space parameters and transmitting the cell space parameters to a robot control unit, wherein the cell space parameters include position parameters and posture parameters; A path planning module, used to calculate an obstacle avoidance trajectory for grabbing the battery cell according to the battery cell space parameters, and configure a vacuum adsorption device at the tool end of the robot arm; A grasping execution module is used to move to the top of the battery cell according to the obstacle avoidance trajectory, perform vacuum adsorption to grasp the battery cell, and monitor the adsorption status in real time; A test docking module, used to transport the battery cell to the test area, output a test ready signal, and complete automatic docking for electrical performance testing; A performance test module, used to perform a cell capacitance and resistance value test upon receiving a test ready signal; The classification storage module is used to move the battery cells to the corresponding storage stations according to the classification of the test results, and complete the automatic sampling of battery cell testing.

Images