Cooperative control method for automatic assembly station of forceps holder

Abstract

The invention discloses a clamp automatic assembly station cooperative control method, and particularly relates to the technical field of precision assembly automation. On the basis of the three-dimensional characteristic model of the forceps holder assembly, assembly parameter extraction, precision threshold calculation, state signal collection, action matching analysis and time sequence strategy adjustment are sequentially executed, and the high-precision and high-stability automatic assembly process is achieved through multi-station cooperative control; constructing an assembly model by extracting geometrical characteristics of a clamping jaw, a connecting pin and a driving arm; dynamically comparing the action state of the assembly station with the assembly result, and calculating a matching coefficient; if the matching is insufficient, a delay or advance control instruction is sent, and the assembly rhythm is adjusted; and finally, the assembly quality is confirmed through error comparison, and an assembly completion state is output or manual reinspection is triggered. The method can effectively avoid multi-station action conflict and mistaken assembly, improves the assembly precision and beat flexibility, and is suitable for an automatic high-precision assembly line of miniature clamp products.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of precision assembly automation, in particular to a cooperative control method for automatic assembly stations of pincers. BACKGROUND

[0002] In modern manufacturing, pincer assemblies are widely used in mechanical structures for high-precision clamping, positioning or clamping, especially in the fields of electronic products, medical devices and micro robots. The requirements for assembly precision, assembly efficiency and cooperative action among multiple stations of pincer assemblies are particularly stringent.

[0003] Currently, in the process of automatic assembly of pincers, a multi-station assembly line layout is often used, and each assembly station operates independently, lacking unified coordination control. Especially in the case where the pincer parts are extremely small (such as the length of the clamping jaw being less than 5mm and the diameter of the connecting pin being less than 1mm) and the assembly tolerance is extremely low (less than ±0.02mm), the timing mismatch, action conflict or misassembly phenomenon frequently occurs among stations, which not only seriously affects the overall assembly efficiency, but also easily causes quality defects such as clamping jaw interference, misplacement and damage.

[0004] In addition, due to the fine and complex structure of pincer parts, multi-axis linkage is often required during assembly, including clamping, rotating, inserting, pressing and other operations. When these actions are performed independently by multiple stations, if there is a lack of effective assembly state perception and information interaction mechanism, the following technical problems may occur: if a station enters assembly in advance, it may cause misplacement assembly of upstream and downstream stations; if the pincer pin is not in place, pressing action is performed, causing damage to the workpiece; in addition, action misjudgment of the station causes frequent system shutdown, seriously affecting the beat control. SUMMARY

[0005] The purpose of the present application is to provide a cooperative control method for automatic assembly stations of pincers to solve the problems in the background art.

[0006] In order to achieve the above-mentioned purpose, the present application provides the following technical scheme: a cooperative control method for automatic assembly stations of pincers, comprising: S100, collecting geometric dimension feature data of a pincer assembly part to be assembled, including microstructure parameters of pincer clamping jaws, connecting pins and driving arms, and constructing a three-dimensional feature model W0 of the pincer assembly; S200, determining the minimum action tolerance range Tn required by each assembly station according to the three-dimensional feature model W0 of the pincer assembly, n being the station number, and obtaining a precision coordination threshold sequence T={T1, T2, …, Tn} of each station; S300, collecting real-time assembly state signals S of each assembly station under the current production beat, including station action completion identifier, pose confirmation signal and clamping jaw position feedback signal; S400: Dynamically compare the status signal S with the accuracy coordination threshold T, calculate the action matching coefficient R between workstations, and if R is lower than the preset threshold R0, send a workstation delay or early execution command. S500: Adjust the action timing control strategy of the corresponding workstation according to the delayed or advanced execution instructions to complete the assembly cycle refactoring. S600. Based on the revised motion control strategy, each assembly station executes the corresponding assembly task to form the clamping component collaborative assembly completion state W1. S700: Compare the assembly completion status W1 with the three-dimensional feature model W0 of the clamping component to check for errors. If the assembly accuracy requirements are met, proceed to the next workpiece assembly cycle. If not, output a misassembly prompt signal to trigger manual re-inspection.

[0007] Preferably, S100 includes: S101. Acquire image data of the clamping jaws, connecting pins, and the parts of the drive arm to be assembled. S102. Perform edge extraction and feature point fitting on the image data to identify the geometric feature parameters of each part to be assembled, including length, cylindricity, coaxiality and tilt angle. S103. Based on the extracted geometric feature parameters, generate a three-dimensional feature model W0 of the clamping component.

[0008] Preferably, wherein S200 includes: S201. Virtually decompose the three-dimensional feature model W0 of the clamping assembly, and map the jaws, connecting pins and drive arms to the assembly areas of different assembly stations. S202. Extract the spatial fit relationships of key structural surfaces in each assembly area, including transition fit, clearance fit, or interference fit types, and calculate the minimum allowable assembly deviation value for the corresponding fit structure. S203. Based on the assembly direction and posture of each structural component of the clamping assembly, and using the kinematic chain analysis method, calculate the minimum tolerance range Tn allowed for each assembly station during its execution. S204. Sort and aggregate the minimum tolerance range Tn according to the work station number n to generate the accuracy coordination threshold sequence T={T1,T2,…,Tn} for subsequent accuracy matching judgment.

[0009] Preferably, wherein S300 includes: S301. Record the signal pulses after the assembly action is completed, and generate a workstation action completion indicator; S302: Collect the spatial pose information of the gripper, connecting pin and drive arm after the workstation is executed, calculate the pose error based on the preset assembly reference pose, and output the pose confirmation signal. S303. Real-time monitoring of minute displacement changes of the gripper in the Z direction, obtaining the gripper pressing depth and comparing it with the target assembly depth to generate a gripper position feedback signal. S304. Combine the workstation action completion identifier, posture confirmation signal and gripper position feedback signal according to the timestamp to form the assembly status signal Si, where i is the current workstation number.

[0010] Preferably, the S400 includes: S401. Compare the assembly status signal Si of the current workstation with the corresponding accuracy coordination threshold Ti item by item, and calculate the posture error, position deviation and action completion time difference respectively. S402. Construct a weighted evaluation model based on various error values, and use a normalized weighted algorithm to calculate the action matching coefficient R between the current workstation and adjacent workstations; S403. Compare the calculated action matching coefficient R with the preset threshold R0; S404. When the matching coefficient R is lower than the preset threshold R0, a control instruction containing the target workstation number and adjustment type is generated. The instruction to delay execution or advance execution is sent according to the workstation position to realize the dynamic adjustment of the workstation action rhythm.

[0011] Preferably, the S500 includes: S501: Parse the target workstation number and adjustment type in the control command and identify it as either "delayed execution" or "early execution" strategy mode; S502. In the "delayed execution" mode, a dynamic waiting time Δt1 is introduced to the target station to delay the trigger signal for starting the assembly action, ensuring that it is aligned with the completion cycle of the preceding station. S503. In the "early execution" mode, the workstation start conditions are reconstructed, and the start logic is converted from static trigger signal to predictive trigger logic based on the pre-completion signal of the preceding workstation action. S504. Write the adjusted action timing control strategy into the action control table, which includes the workstation number, action trigger delay value, pre-trigger condition and synchronization identifier.

[0012] Preferably, the S600 includes: S601. Call the adjusted motion control table, read the trigger type, trigger delay value or predicted trigger condition corresponding to each workstation, and determine the action start sequence of each workstation. S602. The assembly actions of the gripper, connecting pin and drive arm are triggered sequentially according to the start-up sequence. The assembly actions include positioning, insertion, pressing and locking operations. S603. During the execution of actions at each assembly station, the assembly posture and gripper pressing depth data are collected in real time and compared synchronously with the accuracy coordination threshold Ti of the corresponding station. S604. When the assembly actions of all workstations are completed and the collected data meets the accuracy coordination threshold range, it is determined that the clamping components have been assembled in a coordinated manner, and the assembly completion status W1 is generated.

[0013] Preferably, the S700 includes: S701. Call the assembly completion status W1 of the current component and the three-dimensional feature model W0 of the clamping component obtained from the initial modeling, and extract the spatial position and attitude data of the key structural surfaces. S702. Map and compare the corresponding structural surfaces of W1 and W0 one by one, and calculate the position deviation value, angle error value and gripper pressing depth difference value respectively to form the assembly error vector E. S703. Compare the assembly error vector E with the preset accuracy judgment threshold set ΔT item by item. If all error values ​​do not exceed the corresponding threshold, the assembly is deemed qualified. S704. If any item in the error comparison result exceeds the corresponding threshold, an incorrect assembly prompt signal will be output, and the current workpiece number will be marked to trigger the manual re-inspection process.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention achieves dynamic coordination of clamp assembly under micro-structure and high-precision requirements by constructing a multi-station collaborative control method driven by a three-dimensional feature model. Compared with existing assembly methods that rely solely on fixed cycle time and static station control, this invention can perceive the assembly status of each station in real time, dynamically calculate the matching relationship between stations based on action completion signals, spatial posture and position feedback information, and adaptively adjust the sequence and rhythm of station actions. This significantly reduces the incidence of misoperation, cycle time misalignment and assembly interference between stations, and improves the overall assembly stability and precision consistency.

[0015] 2. This invention achieves closed-loop verification of the entire assembly process by introducing a comparison mechanism between the assembly error vector and the accuracy threshold set. By mapping and comparing the actual assembly state with the initial design model, deviations in multiple dimensions such as position, posture, and pressing depth can be accurately identified. Once an error exceeding the limit is detected, a misassembly warning signal is immediately output and manual intervention is triggered to prevent defective products from flowing into subsequent processes, thereby enhancing the traceability and controllability of assembly quality. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] For examples, please refer to Figure 1 As shown, the automatic clamping assembly station collaborative control method described in this embodiment includes: S100: Collect geometric dimension feature data of the clamping assembly to be assembled, including the minute structural parameters of the clamping jaws, connecting pins and drive arms, and construct a three-dimensional feature model W0 of the clamping assembly.

[0020] S101: Acquire image data of the clamping jaws, connecting pins, and the parts of the drive arm to be assembled.

[0021] First, the various parts of the clamping assembly to be assembled are placed within the visual acquisition area of ​​the assembly station. A high-precision camera with an optical magnification of at least 5x is used to acquire multi-angle image data of the clamping jaws, connecting pins, and drive arms. The image acquisition employs ring lighting to eliminate image interference caused by metal reflections from the parts to be assembled. The image resolution is at least 5 micrometers per pixel to meet the requirement of clearly extracting the edges of minute structures. During image acquisition, by setting the exposure time, focal length parameters, and depth of field, the image focus is concentrated on the surface to be assembled, ensuring that the image data accurately reflects the actual geometric shape of the parts to be assembled.

[0022] S102: Perform edge extraction and feature point fitting on the image data to identify the geometric feature parameters of each part to be assembled, including length, cylindricity, coaxiality and tilt angle.

[0023] After the acquired image data is standardized and preprocessed, the edge contours of the clamping jaws, connecting pins and drive arm are extracted using an edge detection algorithm based on the Canny operator. Then, the extracted edges are fitted with curves using the least squares method to obtain the position coordinates of key geometric feature points.

[0024] Based on this, the following geometric parameters are calculated respectively: Length parameter: calculated by fitting the Euclidean distance between the two endpoints of the edge line; Cylindricity error: The average deviation is calculated by fitting the set of points on the cylindrical surface to the set of minimum distances between the point set and the ideal cylindrical surface. Coaxiality error: Compare the eccentricity between the axis of the connecting pin and the axis of the reference hole of the jaw; Tilt angle: Calculated using the angle between the center line of the gripper and the vertical assembly reference surface, in degrees, with an accuracy of 0.1 degrees.

[0025] All extracted geometric feature parameters are stored in structured data form, providing an input basis for 3D modeling.

[0026] S103: Based on the extracted geometric feature parameters, generate a three-dimensional feature model W0 of the clamping component.

[0027] The three-dimensional feature model W0 is a spatial structure model that reflects the assembly state of the clamping assembly. Its construction process includes two stages: three-dimensional reconstruction and assembly benchmark calibration.

[0028] In the 3D reconstruction stage, based on the geometric parameters such as length, cylindricity, coaxiality, and tilt angle obtained in step S102, the gripper, connecting pin, and drive arm are each reconstructed into standard geometric bodies using a parameter-driven modeling method. The reconstruction uses Boolean operations to combine the substructures, forming a complete 3D geometric model of the clamping assembly.

[0029] During the assembly datum calibration stage, the spatial orientation of each sub-component in the 3D model is adjusted according to the assembly datum surface preset in the assembly process, so that its relative position in the 3D space is consistent with the actual assembly conditions, ensuring the model's reference value in subsequent assembly simulation and error analysis.

[0030] S200. Based on the three-dimensional feature model W0 of the clamping assembly, determine the minimum motion tolerance range Tn required for each assembly station, where n is the station number, and obtain the precision coordination threshold sequence T={T1,T2,…,Tn} for each station.

[0031] S201: Virtually decompose the three-dimensional feature model W0 of the clamping assembly, and map the grippers, connecting pins and drive arms to the assembly areas of different assembly stations.

[0032] The three-dimensional feature model W0 of the clamping assembly includes three types of structural components to be assembled: grippers, connecting pins, and drive arms. During the virtual decomposition process, using the structural hierarchy analysis method in the three-dimensional modeling environment, W0 is divided into three independent assembly units according to the assembly sequence. Each assembly unit is mapped to a spatial region of the actual assembly station based on its structural features and assembly path, and is marked as assembly regions R1, R2, and R3. Each assembly region only contains the structural surfaces and motion paths related to the corresponding station, used for determining subsequent assembly mating relationships.

[0033] S202: Extract the spatial fit relationships of key structural surfaces in each assembly area, including transition fit, clearance fit, or interference fit types, and calculate the minimum allowable assembly deviation value for the corresponding fit structure.

[0034] For each assembly area Rn, the fit type of the structural surfaces is identified. The fit type is determined based on the tolerance fit standard. If there is a gap between the structural surfaces, it is determined to be a clearance fit; if the structural surface dimensions overlap but there is no interference, it is a transition fit; if there is interference, it is an interference fit.

[0035] Based on this, the minimum allowable assembly deviation value ΔLn for each mating structure is calculated as follows: For clearance fits, ΔLn equals the maximum permissible clearance minus the actual measured minimum clearance value; For transition fits, ΔLn is the minimum value of the tolerance grade range for intermediate fits; For interference fits, ΔLn is the actual interference amount minus the maximum allowable press-fit interference error.

[0036] S203: Based on the assembly direction and posture of each structural component of the clamping assembly, and using the kinematic chain analysis method, calculate the minimum tolerance range Tn allowed for each assembly station during its execution.

[0037] Based on the fit deviation, the spatial motion chain of each assembly structure in the actual assembly path is further analyzed. The kinematic chain analysis method is used to establish a model of the relative pose change between structural components, considering the superposition effect of rotation, translation and pressing actions on the fit tolerance.

[0038] The minimum tolerance range Tn of each station is defined as the maximum positional error range that each assembly action can tolerate without causing jamming, misalignment or assembly failure. The calculation method is to project and synthesize all the fit deviations ΔLn acting on the assembly station according to the direction of movement to obtain the actual action tolerance zone, with the unit being micrometers.

[0039] S204: Sort and aggregate the minimum tolerance range Tn according to the work station number n to generate a precision coordination threshold sequence T={T1,T2,…,Tn} for subsequent precision matching judgment.

[0040] The Tn obtained in step S203 is arranged in ascending order according to the workstation number to form a precision coordination threshold sequence T={T1,T2,…,Tn} for collaborative judgment. This threshold sequence is stored in the assembly control dataset in the form of a numerical matrix and is used for precision matching and comparison with the actual assembly status signal in subsequent steps to achieve dynamic assembly timing adjustment and motion compensation control.

[0041] S300: Collects real-time assembly status signals S for each assembly station under the current production cycle, including station action completion indicators, posture confirmation signals, and gripper position feedback signals.

[0042] S301: Record the signal pulse after the assembly action is completed and generate a workstation action completion mark.

[0043] An action status trigger is set at the end of each assembly station. When the robotic arm or actuator completes the predetermined assembly action of the station (such as pressing, insertion, or rotational positioning), the trigger generates an electrical signal pulse with a rising or falling edge. This pulse is recorded by the acquisition circuit and converted into a digital signal. When the pulse width exceeds 5 milliseconds and remains stable and continuous, the current assembly action is determined to be completed, and a corresponding station action completion identifier is generated. The identifier format is a Boolean value "1" or "0", representing action completion and incompleteness, respectively.

[0044] S302: Collects spatial pose information of the gripper, connecting pin and drive arm after the workstation is executed, calculates the pose error based on the preset assembly reference pose, and outputs a pose confirmation signal.

[0045] After assembly is completed, a 3D structured light vision device, fixedly installed above the workstation, scans the assembled components to obtain spatial pose information of the grippers, connecting pins, and drive arms, including the X, Y, and Z axial position coordinates and rotation angles around these axes. This actual pose information is compared with the preset assembly reference pose. If the 3D deviation is within the set pose error tolerance range (e.g., displacement less than ±0.02 mm, angle error less than ±0.5 degrees), a valid pose confirmation signal is generated. This pose error tolerance range is preset by the assembly process tolerance model and is consistent with the 3D modeling data.

[0046] S303: Real-time monitoring of minute displacement changes of the gripper in the Z direction, obtaining the gripper pressing depth and comparing it with the target assembly depth to generate a gripper position feedback signal.

[0047] A miniature photoelectric encoder with a resolution better than 1 micrometer is integrated at the end of the gripper in the assembly station to monitor the displacement trajectory of the gripper in the Z-axis (vertical direction) in real time. The final position Za after pressing is recorded and compared with the preset target assembly depth Z0. If the difference between Za and Z0 is within ±0.01 mm, a position qualification feedback signal is generated in the format "gripper pressing normal". If it exceeds this error threshold, a position anomaly feedback signal is generated, indicating a possible assembly defect. This difference is the pressing depth error ΔZ, which is calculated as follows: .

[0048] S304: Combine the workstation action completion identifier, posture confirmation signal and gripper position feedback signal according to the timestamp to form the assembly status signal Si, where i is the current workstation number.

[0049] The three types of status signals generated in steps S301 to S303 are integrated and synchronized in time. The data is then structured according to the current sampling timestamp to form a complete assembly status signal record. Each assembly status signal Si consists of three parts: an action completion identifier, a pose confirmation signal, and a gripper position feedback signal. The assembly status signal Si serves as the dynamic state representation of the current workstation i and will subsequently be used for comparison and analysis with a preset accuracy coordination threshold Tᵢ to determine whether the collaborative control conditions are met, thereby enabling cycle time optimization and anomaly identification.

[0050] S400: Dynamically compare the status signal S with the accuracy coordination threshold T, calculate the action matching coefficient R between workstations, and if R is lower than the preset threshold R0, send a workstation delay or early execution command.

[0051] S401: Compare the assembly status signal Si of the current workstation with the corresponding accuracy coordination threshold Ti item by item, and calculate the posture error, position deviation and action completion time difference respectively.

[0052] The status signal Si of each assembly station i includes an attitude confirmation signal, a gripper position feedback signal, and an action completion indicator signal; at the same time, its corresponding accuracy coordination threshold Ti is a tolerance upper limit preset based on the three-dimensional assembly model, which includes three indicators: attitude error tolerance, position error tolerance, and time error tolerance.

[0053] This step compares each data point in Si with the corresponding tolerance value in Ti item by item, and obtains: Attitude error value Δθi: is the three-axis angular difference between the current gripper or drive arm attitude and the reference attitude, in degrees; Position deviation value ΔLi: The difference between the final pressing position of the gripper in the Z direction and the theoretical assembly depth, in millimeters; Action completion time difference ΔTi: It is the time difference between the action completion time of the current station and the completion time of the adjacent previous station, with the unit of millisecond.

[0054] S402: Construct a weighted evaluation model based on various error values, and use the normalized weighted algorithm to calculate the action matching coefficient R between the current station and the adjacent station.

[0055] Establish an action matching evaluation model, and use the following normalized weighted calculation method to construct the matching coefficient R, whose value range is limited between 0 and 1. 1 represents a perfect match, and 0 represents complete disharmony.

[0056] First, normalize each error Δθi, ΔLi, ΔTi and their corresponding maximum tolerance values respectively to obtain the unitized errors e1, e2, e3. The calculation methods are as follows: Unit attitude error e1 = Δθi / θ_max; Unit position deviation e2 = ΔLi / L_max; Unit time difference e3 = ΔTi / T_max; Where θ_max, L_max, T_max are the maximum tolerance thresholds of attitude, position and time, respectively sourced from the structural definition of the precision coordination threshold Ti.

[0057] Then, according to the different impacts of the three errors on the overall assembly quality, set the weighting coefficients w1, w2, w3 (for example: attitude weight is 0.4, position weight is 0.4, time weight is 0.2). The final calculation formula for the matching coefficient R is: .

[0058] S403: Compare the calculated action matching coefficient R with the preset threshold R0.

[0059] The action matching coefficient R characterizes the timing and precision matching degree of the assembly actions between the current station and the adjacent station. Compare the calculated R value with the preset threshold R0. The preset threshold R0 represents the lowest acceptable coordination degree in collaborative control, and its value is set according to the experimental conditions, preferably set to 0.85. When R ≥ R0, it is considered that the current station beat and precision state are within the acceptable range; when R < R0, it indicates that there is a significant mismatch between the current beat and the action precision, and a dynamic adjustment instruction needs to be triggered.

[0060] S404: When the matching coefficient R is lower than the preset threshold R0, generate a control instruction containing the target station number and the adjustment type, and send a delay execution or early execution instruction according to the station position to achieve dynamic adjustment of the station action beat.

[0061] If the comparison result determines that the current matching coefficient R is lower than the preset threshold R0, then control instructions will be automatically generated based on the workstation number and the direction of the matching deviation.

[0062] If the current workstation's action is performed ahead of schedule, the advance amount is determined based on the time difference ΔTi, and a "delay execution" instruction is sent to that workstation. The instruction includes the waiting time extension value and the next cycle start delay parameter. If the current workstation's actions are lagging and the error is mainly concentrated in the preceding workstation, an "early execution" command is sent to the preceding workstation to optimize its pre-trigger time. All control commands include the target workstation number and execution type identifier to ensure precise synchronization of collaborative actions between multiple workstations, achieving dynamic cycle reconfiguration and stabilization of assembly rhythm.

[0063] S500 adjusts the action timing control strategy of the corresponding workstation according to the delayed or advanced execution instructions to complete the assembly cycle reconfiguration.

[0064] S501: Analyze the target workstation number and adjustment type in the control command and identify it as either "delayed execution" or "early execution" as two strategy modes.

[0065] The control instruction generated in step S404 includes the target workstation number, adjustment type, and adjustment parameters. The instruction format is a structured array. First, the instruction is parsed, and the "adjustment type" field is extracted. If the field value is "delay," it is identified as a "delayed execution" mode; if the field value is "advance," it is identified as an "early execution" mode. Based on the different adjustment types, subsequent control logic is switched to the corresponding strategy path to ensure the accuracy of the action control logic.

[0066] S502: In "delayed execution" mode, a dynamic waiting time Δt1 is introduced to the target station to delay the trigger signal for starting the assembly action, ensuring alignment with the completion cycle of the preceding station.

[0067] When the "delayed execution" mode is identified, the waiting time parameter Δt1, in milliseconds, is extracted from the control command. Δt1 is calculated jointly by the matching deviation of the preceding workstation and the response lag of the current workstation, with the specific value based on the ratio of the action completion time difference ΔTi to the tolerance T_max. After receiving the start trigger condition, the target workstation will first enter a waiting state, and the timer will delay by Δt1 before triggering the action execution logic, thereby achieving synchronization with the previous workstation's cycle and avoiding assembly conflicts caused by premature actions.

[0068] S503: In the "early execution" mode, the workstation start conditions are reconstructed, and the start logic is changed from static trigger signal to predictive trigger logic based on the pre-completion signal of the preceding workstation action.

[0069] If the system is identified as being in "early execution" mode, the traditional static triggering mechanism (such as fixed position or fixed time triggering) is cancelled, and instead, the system listens for the motion trend data of the preceding workstation. When the posture adjustment completion rate of the preceding workstation exceeds 90% and the position error is less than the specified pre-completion threshold (such as ±0.015 mm), it is determined that the assembly process is about to be completed, and the pre-start logic of the target workstation is activated in advance. By setting the predictive trigger window and trigger tolerance, a flexible transition between workstations in asynchronous states is achieved, improving the overall cycle time efficiency.

[0070] S504: Write the adjusted action timing control strategy into the action control table, which includes the workstation number, action trigger delay value, pre-trigger condition and synchronization identifier.

[0071] After the strategy adjustment, the motion control strategy for the current workstation is recorded in a table in the motion control table, which serves as the core basis for assembly process control. The table includes the following fields: workstation number, corresponding trigger type (delayed or advanced), trigger delay Δt1 value (empty if advanced mode), predicted trigger conditions (posture and position thresholds), and synchronization execution flag (used to determine whether dynamic synchronization logic is enabled). The motion control table is reloaded in each cycle to ensure that each workstation always operates at the optimal rhythm in the dynamic assembly environment, completing the real-time reconstruction of the assembly cycle.

[0072] S600, based on the revised motion control strategy, each assembly station executes the corresponding assembly task, forming a state W1 where the clamping components are assembled in a coordinated manner.

[0073] S601: Call the adjusted motion control table, read the trigger type, trigger delay value or predicted trigger condition corresponding to each workstation, and determine the action start sequence of each workstation.

[0074] First, the motion control table generated in step S504 is read, and the station number, trigger type field (delayed or advanced), trigger delay value, and predicted trigger condition parameters for each station are extracted sequentially. Depending on the trigger type, two methods are used to determine the motion start sequence: If it is a "delayed execution" type, the action will start after a preset time Δt1 after receiving the standard trigger signal; If it is the "early execution" type, the action will be triggered in advance when the predicted triggering conditions are met (such as attitude adjustment completion rate ≥90%, gripper position error ≤0.015 mm).

[0075] The action initiation sequence is indexed by the workstation number, forming a set of weighted action trigger timelines to ensure that the execution order of multiple workstations satisfies both the assembly logic and the coordination strategy.

[0076] S602: The assembly actions of the gripper, connecting pin and drive arm are triggered sequentially according to the start-up sequence. The assembly actions include positioning, insertion, pressing and locking operations.

[0077] Based on the generated action start sequence, the assembly equipment is controlled to sequentially execute the assembly operations of each component. Each assembly action includes the following sub-steps: Spatial positioning and attitude adjustment of the gripper are performed to align it with the assembly reference surface; The control connecting pin is inserted axially into the positioning hole to complete the initial fixation; Start the Z-axis pressing mechanism to precisely press the grippers into the assembly groove and record the pressing depth; The locking unit is triggered to complete the clamping and locking, ensuring the stability of the assembled structure.

[0078] The execution sequence and control logic of all actions are uniformly scheduled by the action control table to prevent action conflicts or assembly rhythm disorder.

[0079] S603: During the execution of actions at each assembly station, the assembly posture and gripper pressing depth data are collected in real time and compared synchronously with the accuracy coordination threshold Ti of the corresponding station.

[0080] During the assembly process, high-precision 3D vision equipment and displacement sensors are used to collect the current spatial orientation of the components and the Z-axis pressing depth of the grippers, respectively. The obtained data is recorded synchronously with timestamps and uploaded to the assembly accuracy analysis unit.

[0081] Each piece of collected data is compared in real time with the corresponding accuracy coordination threshold Ti for that workstation. Ti is a preset set of multi-dimensional tolerance limits, including the maximum attitude angle deviation (e.g., ±0.5 degrees), the maximum position error (e.g., ±0.02 mm), and the minimum pressing depth tolerance (e.g., ±0.01 mm). If all collected data fall within the tolerance range defined by Ti, the assembly action at that workstation is considered qualified; otherwise, the error is recorded and processed in the feedback loop.

[0082] S604: When the assembly actions of all workstations are completed and the collected data meets the accuracy coordination threshold range, it is determined that the clamping components have been assembled in a coordinated manner, and the assembly completion status W1 is generated.

[0083] Once all workstations have completed their assembly actions and their collected data have passed the accuracy coordination threshold Ti, an assembly completion indicator signal can be generated, indicating that the current component has completed collaborative assembly.

[0084] The assembly completion status W1 is a structured data packet containing the component number, workstation completion flag, posture error report, position error report, and final assembly quality judgment flag. This status data is not only used to confirm the assembly of the current component, but also serves as the input basis for the next cycle, participating in subsequent dynamic control judgments to ensure the continuous and stable operation of the assembly line.

[0085] S700: Compare the assembly completion status W1 with the three-dimensional feature model W0 of the clamping component to check for errors. If the assembly accuracy requirements are met, proceed to the next workpiece assembly cycle. If not, output a misassembly prompt signal to trigger manual re-inspection.

[0086] S701: Call the assembly completion status W1 of the current component and the three-dimensional feature model W0 of the clamping component obtained from the initial modeling, and extract the spatial position and attitude data of the key structural surfaces.

[0087] First, extract the measured three-dimensional spatial data from the assembly completion state W1 generated in step S604, including the final position information (X, Y, Z) and spatial attitude information (rotation angles around the X, Y, and Z axes) of the gripper, connecting pin, and drive arm.

[0088] Simultaneously, the three-dimensional feature model W0 of the clamping component constructed in step S103 is invoked to extract the spatial position and attitude data of the same structural component in the design state, ensuring that the two models are consistent in the reference coordinate system.

[0089] S702: Map and compare the corresponding structural surfaces of W1 and W0 one by one, calculate the position deviation value, angle error value and gripper pressing depth difference value respectively, and form the assembly error vector E.

[0090] Establish the mapping relationship between the corresponding structural surfaces in W0 and W1, and align the measurement model to the design model using a coordinate registration algorithm (e.g., least squares rigid transformation fitting).

[0091] For each structural surface or feature point, calculate the following error parameters: Position deviation value ΔL: is the difference in three-dimensional coordinates between the center point of the structural surface measured in W1 and the corresponding point in W0, in millimeters; Angular error value Δθ: The attitude deviation of the corresponding structural surface around the three axes, in degrees; Pressing depth difference ΔZ: is the difference between the actual gripper end depth and the modeled pressing reference depth.

[0092] All error values ​​are aggregated into a vector form to construct an assembly error vector E={ΔL1,Δθ1,ΔZ1,...,ΔLg,Δθg,ΔZg}, where g represents the number of comparison structures.

[0093] S703: Compare the assembly error vector E with the preset accuracy judgment threshold set ΔT item by item. If all error values ​​do not exceed the corresponding threshold, the assembly is deemed qualified.

[0094] Each error value in the assembly error vector E is compared one-to-one with the accuracy judgment threshold set ΔT. ΔT is a set of preset assembly accuracy upper limits, including the maximum allowable position deviation (e.g., ±0.03 mm), the maximum allowable attitude angle error (e.g., ±0.5 degrees), and the gripper pressing depth tolerance (e.g., ±0.01 mm).

[0095] If all elements in the error vector E are less than or equal to the corresponding ΔT element, the current assembly process is considered to meet the accuracy requirements, the assembly status is recorded as "qualified", and the status data is fed back to the control port to enter the next workpiece assembly process.

[0096] S704: If any item in the error comparison result exceeds the corresponding threshold, an error prompt signal will be output, the current workpiece number will be marked, and the manual re-inspection process will be triggered.

[0097] When any error value ΔEi exceeds its corresponding threshold ΔTi, the system automatically determines it as an assembly anomaly. It immediately outputs a misassembly warning signal and records the current component number, the anomaly error item, and its corresponding value in the quality anomaly log.

[0098] Simultaneously, a manual re-inspection process is triggered, removing the workpiece from the automated assembly process and transferring it to the manual re-inspection station. The operator then analyzes and intervenes in the assembly defects based on the error prompts, preventing unqualified components from entering downstream processes and ensuring closed-loop control of product assembly quality.

[0099] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A collaborative control method for an automatic clamping assembly station, characterized in that: include: S100. Collect the geometric dimension feature data of the clamping assembly to be assembled, including the micro-structural parameters of the clamping jaws, connecting pins and drive arms, and construct a three-dimensional feature model W0 of the clamping assembly. S200. Based on the three-dimensional feature model W0 of the clamping assembly, determine the minimum motion tolerance range Tn required for each assembly station, where n is the station number, and obtain the precision coordination threshold sequence T={T1,T2,…,Tn} for each station. S300: Collects real-time assembly status signals S for each assembly station under the current production cycle, including station action completion indicators, posture confirmation signals, and gripper position feedback signals. S400: Dynamically compare the status signal S with the accuracy coordination threshold T, calculate the action matching coefficient R between workstations, and if R is lower than the preset threshold R0, send a workstation delay or early execution command. S500: Adjust the action timing control strategy of the corresponding workstation according to the delayed or advanced execution instructions to complete the assembly cycle refactoring. S600. Based on the revised motion control strategy, each assembly station executes the corresponding assembly task to form the clamping component collaborative assembly completion state W1. S700: Compare the assembly completion status W1 with the three-dimensional feature model W0 of the clamping component to check for errors. If the assembly accuracy requirements are met, proceed to the next workpiece assembly cycle. If not, output a misassembly prompt signal to trigger manual re-inspection.

2. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S100 includes: S101. Acquire image data of the clamping jaws, connecting pins, and the parts of the drive arm to be assembled. S102. Perform edge extraction and feature point fitting on the image data to identify the geometric feature parameters of each part to be assembled, including length, cylindricity, coaxiality and tilt angle. S103. Based on the extracted geometric feature parameters, generate a three-dimensional feature model W0 of the clamping component.

3. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S200 includes: S201. Virtually decompose the three-dimensional feature model W0 of the clamping assembly, and map the jaws, connecting pins and drive arms to the assembly areas of different assembly stations. S202. Extract the spatial fit relationships of key structural surfaces in each assembly area, including transition fit, clearance fit, or interference fit types, and calculate the minimum allowable assembly deviation value for the corresponding fit structure. S203. Based on the assembly direction and posture of each structural component of the clamping assembly, and using the kinematic chain analysis method, calculate the minimum tolerance range Tn allowed for each assembly station during its execution. S204. Sort and aggregate the minimum tolerance range Tn according to the work station number n to generate the accuracy coordination threshold sequence T={T1,T2,…,Tn} for subsequent accuracy matching judgment.

4. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S300 includes: S301. Record the signal pulses after the assembly action is completed, and generate a workstation action completion indicator; S302: Collect the spatial pose information of the gripper, connecting pin and drive arm after the workstation is executed, calculate the pose error based on the preset assembly reference pose, and output the pose confirmation signal. S303. Real-time monitoring of minute displacement changes of the gripper in the Z direction, obtaining the gripper pressing depth and comparing it with the target assembly depth to generate a gripper position feedback signal. S304. Combine the workstation action completion identifier, posture confirmation signal and gripper position feedback signal according to the timestamp to form the assembly status signal Si, where i is the current workstation number.

5. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S400 includes: S401. Compare the assembly status signal Si of the current workstation with the corresponding accuracy coordination threshold Ti item by item, and calculate the posture error, position deviation and action completion time difference respectively. S402. Construct a weighted evaluation model based on various error values, and use a normalized weighted algorithm to calculate the action matching coefficient R between the current workstation and adjacent workstations; S403. Compare the calculated action matching coefficient R with the preset threshold R0; S404. When the matching coefficient R is lower than the preset threshold R0, a control instruction containing the target workstation number and adjustment type is generated. The instruction to delay execution or advance execution is sent according to the workstation position to realize the dynamic adjustment of the workstation action rhythm.

6. The method for collaborative control of automatic clamping assembly station according to claim 5, characterized in that: The S500 includes: S501: Parse the target workstation number and adjustment type in the control command and identify it as either "delayed execution" or "early execution" strategy mode; S502. In the "delayed execution" mode, a dynamic waiting time Δt1 is introduced to the target station to delay the trigger signal for starting the assembly action, ensuring that it is aligned with the completion cycle of the preceding station. S503. In the "early execution" mode, the workstation start conditions are reconstructed, and the start logic is converted from a static trigger signal to a predictive trigger logic based on the pre-completion signal of the preceding workstation action. S504. Write the adjusted action timing control strategy into the action control table, which includes the workstation number, action trigger delay value, pre-trigger condition and synchronization identifier.

7. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S600 includes: S601. Call the adjusted motion control table, read the trigger type, trigger delay value or predicted trigger condition corresponding to each workstation, and determine the action start sequence of each workstation. S602. The assembly actions of the gripper, connecting pin and drive arm are triggered sequentially according to the start-up sequence. The assembly actions include positioning, insertion, pressing and locking operations. S603. During the execution of actions at each assembly station, the assembly posture and gripper pressing depth data are collected in real time and compared synchronously with the accuracy coordination threshold Ti of the corresponding station. S604. When the assembly actions of all workstations are completed and the collected data meets the accuracy coordination threshold range, it is determined that the clamping components have been assembled in a coordinated manner, and the assembly completion status W1 is generated.

8. The method for collaborative control of automatic clamping assembly station according to claim 1, characterized in that: The S700 includes: S701. Call the assembly completion status W1 of the current component and the three-dimensional feature model W0 of the clamping component obtained from the initial modeling, and extract the spatial position and attitude data of the key structural surfaces. S702. Map and compare the corresponding structural surfaces of W1 and W0 one by one, and calculate the position deviation value, angle error value and gripper pressing depth difference value respectively to form the assembly error vector E. S703. Compare the assembly error vector E with the preset accuracy judgment threshold set ΔT item by item. If all error values ​​do not exceed the corresponding threshold, the assembly is deemed qualified. S704. If any item in the error comparison result exceeds the corresponding threshold, an incorrect assembly prompt signal will be output, and the current workpiece number will be marked to trigger the manual re-inspection process.

Images