A multi-tapping station automatic tapping equipment and control system

Abstract

The present invention relates to an automatic tapping device with multiple tapping stations and a control system: comprising a position change seat, a blanking and sorting device, a switching cleaning device, and at least two tapping devices; the control system comprises an intelligent interactive subsystem, and the intelligent interactive subsystem comprises a model generation module, an action matching module, and an instruction generation module. The tapping device can be provided with different tapping drills to achieve different tapping effects, so that the device can adapt to the processing of various parts. Moreover, by providing at least two tapping stations, switching can be performed between tapping actions to avoid overheating of the tapping device and at the same time, chip cleaning operations can be performed on the tapping position in a timely manner. By switching the operation of the cleaning device, the parts to be processed are driven to rotate during the chip cleaning process, ensuring the smooth completion of the chip cleaning action, and at the same time avoiding the phenomenon of the rotating structure being subjected to a large impact force and becoming misaligned due to the reuse of the rotating station and the tapping station.

Description

Technical Field

[0001] The present invention relates to the technical field of parts processing, and more particularly to an automatic tapping device with multiple tapping stations. Background Art

[0002] With the continuous improvement of the level of automation in the manufacturing industry, the efficiency and quality requirements for precision tapping equipment in the field of metal parts processing are increasing day by day. Traditional tapping equipment mostly adopts a single-station operation mode, completing the tapping and chip cleaning processes through a rotating clamping mechanism. Specifically, the following technical defects are common in the existing technology:

[0003] Continuous operation at a single station causes a significant temperature rise in the tool, especially when processing high-hardness materials. The drill bit is prone to thermal deformation due to the lack of cooling intervals, which directly affects the tapping accuracy and tool life. Although some solutions have attempted to install a cooling mechanism, it increases the complexity of the equipment and makes it difficult to eradicate the problem of heat accumulation. In this process, the problem of heat accumulation needs to be solved by stopping the machine to dissipate heat. There is also the problem of chip cleaning, but this process requires the cooperation of multiple components. The current solution is to perform tapping and chip cleaning at one station, and switch between tapping and chip cleaning. However, this way: the chip cleaning function is too highly coupled with the tapping station. In conventional designs, the chip cleaning process relies on the rotation of the clamping mechanism, which requires the rotating parts of the clamping mechanism to perform high-frequency rotation while bearing the axial impact force of tapping, accelerating bearing wear and causing positioning deviation. According to statistics, the positioning accuracy of such a structure can drop by more than 0.05mm after 200 hours of continuous operation.

[0004] On the other hand, tapping systems also suffer from poor adaptability to multi-specification machining. A single tapping module cannot handle the machining needs of different thread sizes, and tool changes require downtime and adjustment, severely impacting equipment utilization. While some literature has proposed modular tool assembly solutions, these solutions do not address the issue of positioning and calibration during tool changes. Summary of the Invention

[0005] In view of this, the object of the present invention is to provide a load resource data management system based on quality verification rules with intelligent switching, precise chip cleaning and online detection functions.

[0006] In order to solve the above technical problems, the technical solution of the present invention is: a multi-tapping station automatic tapping equipment: including a position change seat, a blanking and sorting device, a switching and cleaning device, and at least two tapping devices;

[0007] The transposition seat includes a station turntable and a rotation drive member, and the rotation drive member is used to drive the station turntable to rotate;

[0008] The transposition seat is provided with a loading station, a tapping station and a blanking station, the tapping station is provided corresponding to the tapping device, and the blanking and sorting device is provided at the blanking station;

[0009] The tapping device includes a tapping robot arm and a tapping module, wherein the tapping robot arm is used to drive the tapping module to move at the tapping station, and the tapping module is used to perform the tapping action;

[0010] The switching cleaning device is arranged between the tapping devices, and includes a reversing auxiliary part and a cleaning execution module. When the workpiece to be processed moves between the two tapping stations driven by the station turntable, the reversing auxiliary part drives the workpiece to be processed to rotate on the station turntable, and the cleaning execution module is used to perform the cleaning action;

[0011] The transposition seat is provided with a plurality of fixed modules, and the fixed modules are used to fix the parts to be processed at the tapping station;

[0012] The unloading and sorting device includes an image recognition module and a sorting module. The image recognition module is configured with an image recognition strategy. The image recognition strategy is used to identify the image of the processed parts to generate quality data, and generate sorting instructions based on the quality data and send them to the sorting module. The sorting module drives the processed parts to the corresponding unloading area according to the sorting instructions.

[0013] Furthermore: the transposition seat is provided with a fixed groove on the tapping station. When the part to be processed rotates to the tapping station, it falls into the fixed groove to limit the rotation of the part to be processed. The tapping device is provided with a suction head unit. When the suction head unit is working, it sucks the part to be processed to make it leave the fixed groove.

[0014] Furthermore: the image recognition module includes an image acquisition unit and a focus adjustment unit, the image acquisition unit is arranged on the focus adjustment unit, and the focus adjustment unit is used to adjust the distance between the image acquisition unit and the processed part.

[0015] A multi-tapping station automatic tapping control system is also provided, which is configured in the above-mentioned multi-tapping station automatic tapping equipment, and also includes an intelligent interactive subsystem, the intelligent interactive subsystem includes a model generation module, an action matching module and an instruction generation module, the model generation module is used to generate a part simulation model according to the part information to be processed input by the user, and process the part simulation model through the processing target information to obtain a processing difference model, the action matching module is configured with a tapping configuration parameter library, the tapping configuration parameter library stores a number of tapping constraint features, the tapping constraint features are indexed by the tapping device parameters, the The action matching module obtains the corresponding tapping constraint features by substituting the tapping device parameters of different tapping devices, and generates tapping morphological constraints based on the tapping constraint features. The different tapping morphological constraints are matched with the processing difference model to generate a tapping trajectory sequence. The tapping trajectory sequence includes several tapping sub-trajectories. The instruction generation module is configured with a static cost algorithm. The static cost algorithm is used to calculate the static cost of each tapping sub-trajectory, and use the total static cost value as a constraint to generate a tapping trajectory set based on adjacent tapping sub-trajectories. The corresponding tapping action instructions are generated based on the tapping trajectory set to control the operation of the corresponding tapping device.

[0016] Furthermore: the intelligent interactive subsystem also includes a cleaning association module, the cleaning association module includes a cleaning instruction set and an action association table, the cleaning instruction set includes a number of cleaning instructions, the cleaning instructions are indexed by action association tags, the action association table stores a number of action association tags, the action association tags are indexed by associated action number information, the associated action number information is a front action number and a back action number, the cleaning association module obtains the corresponding front action number and back action number according to the adjacent tapping action instructions to call the cleaning instruction, and the cleaning instruction is used to control the operation of the cleaning execution module.

[0017] Furthermore: it also includes a dynamic detection module, the dynamic detection module includes a device detection unit and a workpiece detection unit, the device detection unit is used to detect the state feedback data of the tapping device when it is working, and the workpiece detection unit is used to detect the dynamic feedback data of the part to be processed when it is being processed. The intelligent interactive subsystem also includes a dynamic correction module, and the dynamic correction module is configured with a dynamic correction condition. When the dynamic correction condition is triggered, the dynamic correction module generates a dynamic correction instruction based on the state feedback data and the dynamic feedback data, and divides the target tapping sub-trajectory from the current tapping trajectory set to the adjacent target tapping trajectory set according to the dynamic correction instruction.

[0018] Furthermore: the intelligent interaction subsystem also includes a macro correction module, the macro correction module is configured with a quality matching strategy, the quality matching strategy generates macro correction parameters according to quality data, and corrects the static cost algorithm according to the macro correction parameters.

[0019] Furthermore: the dynamic correction condition includes a device temperature sub-condition, a component temperature sub-condition and a component chip sub-condition.

[0020] Furthermore: the blanking and sorting device is configured with an abnormal feature database, which stores a number of quality abnormality items, and each quality abnormality item is indexed by a quality graphic feature; the image recognition strategy generates a shooting instruction based on the processing difference model, and controls the image recognition module to shoot the part image according to the shooting instruction, and matches the quality graphic feature and the part image to obtain the corresponding quality difference item, and forms the quality data based on the quality difference item.

[0021] Furthermore: the macro correction module is configured with a macro correction database, the macro correction database is configured with a number of abnormal correlation sub-items and corresponding abnormal correlation values, the abnormal correlation sub-items and abnormal correlation values, the quality matching strategy includes calculating the total abnormal correlation value under each abnormal correlation item, the total abnormal correlation value is the weighted result of the abnormal correlation values ​​corresponding to the abnormal correlation sub-items contained in the abnormal correlation item, when the total abnormal correlation value is greater than the preset abnormal trigger value, the macro correction parameter is generated according to the total abnormal correlation value.

[0022] The technical effects of the present invention are mainly reflected in the following aspects: through such a setting, the tapping device can be set with different tapping drills to achieve different tapping effects, so that the equipment can adapt to the processing of various parts, and through the setting of at least two tapping stations, it is possible to switch between tapping actions to avoid overheating of the tapping device and at the same time to perform chip cleaning operations on the tapping position in time. By switching the operation of the cleaning device, during the chip cleaning process, the parts to be processed are driven to rotate, ensuring the smooth completion of the chip cleaning action, and at the same time avoiding the phenomenon of the rotating structure being subjected to a large impact force and becoming misaligned due to the reuse of the rotating station and the tapping station. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 : Schematic diagram of the shaft side structure of a multi-tapping station automatic tapping equipment of the present invention Figure 1 ;

[0024] Figure 2 : Schematic diagram of the shaft side structure of a multi-tapping station automatic tapping equipment of the present invention Figure 2 ;

[0025] Figure 3 : A side view of an automatic tapping device with multiple tapping stations according to the present invention;

[0026] Figure 4 : A top view of an automatic tapping device with multiple tapping stations according to the present invention;

[0027] Figure 5: The present invention is a multi-tapping station automatic tapping control system architecture diagram.

[0028] : Figure numerals: 100, transposition seat; 110, work station turntable; 120, rotating drive member; 130, fixed module; 131, workpiece detection unit; 200, blanking and sorting device; 210, image recognition module; 211, image acquisition unit; 212, focal length adjustment unit; 220, sorting module; 300, tapping device; 310, tapping robot arm; 320, tapping module; 321, device detection unit; 322, tapping drive motor; 323, tapping linkage structure; 324, tapping tool head; 325, tool setting instrument; 326, suction head unit; 400, switching cleaning device; 410, reversing auxiliary part; 420, cleaning execution module; 421, cleaning drive mechanism; 422, air outlet drive member; 423, air nozzle; 510, model generation module; 520, action matching module; 530, instruction generation module. DETAILED DESCRIPTION

[0029] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings to make the technical solutions of the present invention easier to understand and grasp.

[0030] Reference Figure 1-5 , a multi-tapping station automatic tapping equipment: including a transposition seat 100, a blanking and sorting device 200, a switching cleaning device 400, and at least two tapping devices 300;

[0031] Reference Figure 1-5 The transposition seat 100 includes a station turntable 110 and a rotation driving member 120, and the rotation driving member 120 is used to drive the station turntable 110 to rotate;

[0032] The transposition seat 100 is provided with a loading station, a tapping station and a blanking station. The tapping station is provided corresponding to the tapping device 300, and the blanking and sorting device 200 is provided at the blanking station.

[0033] The tapping device 300 includes a tapping robot arm 310 and a tapping module 320. The tapping robot arm 310 is used to drive the tapping module 320 to move at the tapping station, and the tapping module 320 is used to perform the tapping action. The tapping robot arm 310 can be set as a three-axis robot arm driven by three driving devices to ensure the range of movement. The tapping module 320 includes a tapping drive motor 322, a tapping linkage structure 323 and a tapping cutter head 324. The tapping drive motor 322 drives the tapping linkage structure 323 to move. The tapping linkage structure 323 includes a quick chuck for clamping the tapping cutter head 324. The tapping head 324 structure set in different tapping devices 300 can be different to adapt to different scenarios. An oil cooling device is also provided on the side of the tapping head 324 to assist in cooling the tapping head 324. Preferably, the tapping device 300 also includes a tool setter 325 for judging whether the installation position of the tapping head 324 is correct and whether there is any offset. The tool setter 325 can be set as an infrared sensor. During installation, the tapping head 324 is moved to the corresponding position, and the tool setter 325 collects data once as initial data. Then, the tool setter 325 judges whether the tool head is offset according to the actual distance of the tool head.

[0034] The switching cleaning device 400 is arranged between the tapping devices 300. The switching cleaning device 400 includes a reversing auxiliary part 410 and a cleaning execution module 420. When the part to be processed moves between the two tapping stations under the drive of the station turntable 110, the reversing auxiliary part 410 drives the part to be processed to rotate on the station turntable 110, and the cleaning execution module 420 is used to perform the cleaning action; It should be noted that the switching cleaning device 400 is arranged between the tapping devices 300. The figure shows two tapping devices 300. If the number of tapping devices 300 is set to three, the switching cleaning device 400 0 quantity should be set to two, and the reversing auxiliary part 410 can be set as a friction track under the transposition seat 100. When the transposition seat 100 rotates, it drives the parts to move, and the parts will rotate when passing through the friction track, so that the switching cleaning device 400 can clean the parts to be processed in all directions. The cleaning execution module 420 is specifically set to a cleaning drive mechanism 421, an air outlet drive part 422, and an air nozzle 423. The cleaning drive mechanism 421 is used to drive the air nozzle 423 to follow the movement of the air outlet drive part 422, the air outlet drive part 422 is used to generate an air source, and the air nozzle 423 is used to perform chip cleaning actions when working.

[0035] The transposition seat 100 is provided with a plurality of fixed modules 130, which are used to fix the workpiece to be processed in the tapping station. The transposition seat 100 is provided with a fixed groove on the tapping station. When the workpiece to be processed rotates to the tapping station, it falls into the fixed groove to restrict the rotation of the workpiece to be processed. The tapping device 300 is provided with a suction head unit 326. When the suction head unit 326 is in operation, it sucks the workpiece to be processed and releases it from the fixed groove. Preferably, a clamping mechanism is provided in the fixed groove to clamp different types of workpieces. When processing is completed, the clamping mechanism is released, and the workpiece is sucked out of the fixed groove and leaves the fixed groove. The transposition seat 100 is then operated to remove the corresponding tapping station.

[0036] The unloading and sorting device 200 includes an image recognition module 210 and a sorting module 220. The image recognition module 210 is configured with an image recognition strategy for identifying images of processed parts to generate quality data, and generating sorting instructions based on the quality data and sending them to the sorting module 220. The sorting module 220 drives the processed parts to the corresponding unloading area according to the sorting instructions. The sorting module 220 includes a support frame, on which a transfer drive is provided. The transfer drive drives the magnetic clamp to move through the transfer structure to move the product to different output ports according to the quality data. The image recognition module 210 includes an image acquisition unit 211 and a focus adjustment unit 212. The image acquisition unit 211 is provided on the focus adjustment unit 212, and the focus adjustment unit 212 is used to adjust the distance between the image acquisition unit 211 and the processed parts.

[0037] Reference Figure 4 The working principle is as follows: loading is carried out through the loading station, and the rotating drive part 120 drives the rotating disk to rotate to the first tapping station. After completing the first part of the tapping action, it is driven to the second tapping station. During the process, the tapping device 300 is cooled and the cleaning device 400 is switched to clean the parts to be processed. If there are multiple actions, it will continue to go back and forth between the first tapping station and the second tapping station until all actions are completed, and then go to the unloading station for quality identification and unloading.

[0038] Reference Figure 5 , intelligent interaction subsystem, the intelligent interaction subsystem includes a model generation module 510, an action matching module 520 and an instruction generation module 530,

[0039] The model generation module is used to generate a part simulation model based on the part information to be processed input by the user, and process the part simulation model through the processing target information to obtain a processing difference model. The purpose of the model generation module is to customize the model so that the processing action is constructed according to the model. Since the ideal processed information is known, as long as the input information is constructed, the position to be processed can be obtained by model difference. The input information can be obtained by performing a three-dimensional scan on the part to be processed, and then the relative position relationship is unified. The processing target information is imported into the model to obtain the corresponding processing difference model. Specifically, the scanned part model is simulated as follows: ,in, is the model coupling operator, For the part simulation model, For the part model obtained by scanning, are simulation parameters, *K shape / E, where is the tapping stress, is the material thickness, E is the elastic modulus, K shape is the shape correction coefficient. The generation formula of the processing difference model is ,in, Processing information for a goal, It is the processing difference model.

[0040] The motion matching module is equipped with a tapping configuration parameter library, which stores several tapping constraint features indexed by tapping device parameters. These include: all static parameters associated with the tapping device (e.g., robotic arm range of motion, tool head type, motor performance, torque limit, etc.), as well as dynamic constraints derived from these parameters (e.g., maximum feed rate, machining angle range, cutting force threshold). All parameters are associated using the tapping device's unique identifier (e.g., device ID). Mechanical parameters include: the number of robotic arm axes (e.g., three axes), travel range (maximum displacement in the X, Y, and Z axes), and repeatability (±0.01 mm). Power parameters include: the rated power (e.g., 500 W), maximum speed (3000 rpm), and torque curve (e.g., maximum torque 5 N·m) of the tapping drive motor. Tool parameters include: tool head type (e.g., M6 tap), tool tip angle (118°), and cooling requirements (oil / air cooling). Sensor parameters: tool setter detection accuracy (±0.005mm), infrared sensor sampling frequency (100Hz). This allows the output of tapping constraint features based on these device parameters. Constraints determine how each tapping action is split. Constraints extracted from the tapping configuration parameter library are used to limit the physical boundaries and dynamic performance of the tapping action, ensuring machining safety and accuracy. For example, tapping constraint features can constrain maximum feed rate and torque limits, tool life, and cooling efficiency.

[0041] The action matching module obtains the corresponding tapping constraint features by substituting the tapping device parameters of different tapping devices, and generates tapping morphological constraints based on the tapping constraint features. Different tapping morphological constraints are matched with the machining difference model to generate a tapping trajectory sequence. On the basis of the machining difference model (the area to be processed), the machining path rules generated by the tapping constraint features are combined to describe the geometric shape (such as thread depth, helix angle) and dynamic characteristics (such as feed rate, acceleration) of the tapping action. Generation steps: First, segment the model: decompose the machining difference model into multiple sub-areas (such as threaded hole area 1 and threaded hole area 2). Then perform parameter matching: match the constraint features of the tapping device for each sub-area. For example: the rotation speed needs to be reduced for deep hole processing (to avoid poor chip removal). Morphological rule generation: For example, helix angle calculation: ,in, is the helix angle constraint, is the feed rate, n is the rotational speed, and d is the thread diameter.

[0042] The tapping trajectory sequence contains several tapping sub-trajectories, and the instruction generation module is configured with a static cost algorithm, which is used to calculate the static cost of each tapping sub-trajectory, and use the total static cost value as a constraint to generate a tapping trajectory set based on adjacent tapping sub-trajectories, and generate corresponding tapping action instructions based on the tapping trajectory set to control the operation of the corresponding tapping device. The tapping morphological constraints are converted into specific motion paths, which are composed of multiple tapping sub-trajectories (such as feed, cut, and retract) in sequence. The generation method is as follows: Trajectory planning: A* algorithm: Search for the optimal path within the range of motion of the robot arm to avoid singular points. Spline interpolation: Generate a smooth acceleration curve to reduce mechanical vibration. Sub-trajectory division: Feed trajectory: The tool head moves quickly from a safe position to the processing starting point. Cutting trajectory: Thread processing is performed according to the helix angle β and layer depth. Retract trajectory: Reverse rotation to exit and return to a safe position. The motion equation of a single sub-trajectory: ,in is the displacement change over time, is the rotation angle of the cutter head, t is the time, is the movement speed, is the starting position, is the acceleration change, is the starting angle, is the angle change. The cost function is designed as follows: ,in, The sub-track time is the time required to move to the next tapping device. is the energy consumption cost, which reflects the power consumption cost of the tapping equipment and cooling equipment required for this work. Tool wear coefficient reflects the loss value of the tool. The longer the single work, the greater the loss cost of the tool. At the cost of accuracy, The corresponding configuration weights can be adjusted based on actual cost preferences to reflect the impact of the current tool operation on accuracy. Trajectory Set Generation: Traverses all possible sub-trajectory combinations. Dynamic Programming (DP) selects the sequence with the lowest total cost. Constrained Optimization: Uses Lagrange multipliers to handle multi-constraint problems (e.g., total time ≤ T_max). Example: Assume two candidate trajectory sets: Trajectory Set A: Total cost = 1.2. Trajectory Set B: Total cost = 1.5. The algorithm selects Trajectory Set A as the optimal solution.

[0043] The intelligent interaction subsystem also includes a cleaning association module, which includes a cleaning instruction set and an action association table. The cleaning instruction set includes several cleaning instructions, indexed by action association tags. The action association table stores several action association tags, each indexed by associated action number information, which includes the preceding and following action numbers. The cleaning association module retrieves cleaning instructions based on the corresponding preceding and following action numbers of adjacent tapping action instructions. These cleaning instructions are used to control the operation of the cleaning execution module. The cleaning instruction set is a structured database that stores all executable cleaning action instructions (such as high-pressure air blowing, oil mist spray, nozzle movement path, etc.). Each instruction is uniquely indexed by an action association tag. Instruction types: Basic cleaning instructions: Short air blowing (e.g., 0.5 seconds), Long air blowing (e.g., 2 seconds), and reciprocating sweeping. Compound cleaning instructions: Air blowing + oil mist cooling (for hot debris), multi-angle cleaning (nozzle oscillation). Parameter Configuration: Air Pressure (MPa): Adjust based on chip type (e.g., 0.3MPa for aluminum chips, 0.5MPa for steel chips). Duration (seconds): Dynamically calculated based on the chip volume generated by the tapping action. Nozzle Path: Defines the nozzle's trajectory (e.g., straight, spiral, or fan-shaped). Definition: The Action Association Table is a relational database that stores the mapping between action association tags and action number information, enabling rapid matching of tapping action sequences with cleaning instructions. Data Structure: Primary Key: Action number information (previous action number + next action number). Value: Action association tag (uniquely identifies the cleaning instruction). Algorithm Construction: Tapping Action Classification: Tapping actions are numbered by type (e.g., T1 = rough tapping, T2 = fine tapping, T3 = chamfering). Association Rule Definition: If the previous and next action types are different (e.g., T1 → T2), a tool change and cleaning instruction is triggered (labeled T1 → T2). If the previous and next actions are of the same type (e.g., T2→T2), a repeat cleaning instruction is triggered (marked T2→T2). Table storage optimization: A hash table is used to implement O(1) time complexity queries. Action association tag definition: An action association tag is a string that uniquely identifies the combination of the previous and next actions, and is used to locate the corresponding cleaning instruction in the cleaning instruction set. Coding rules: Tag generation: Connect the previous action number and the next action number with a separator (e.g., "T1→T2"). Dynamic expansion: Automatically generate new tags when adding new tapping action types (e.g., if T4 is added, the tag T3→T4 is automatically added to the association table). Example: If the tapping action sequence is rough tapping (T1) → fine tapping (T2), it is marked as T1→T2. If fine tapping is performed twice in a row (T2→T2), it is marked as T2→T2, triggering a deep cleaning instruction.Association Action Number Information ‌‌ Definition‌: Association action number information consists of the ‌Previous Action Number‌ and the ‌Next Action Number‌, indicating the sequential relationship between two adjacent tapping actions and serving as the basis for querying the action association table. ‌Numbering Rule‌: ‌Previous Action Number‌: Identifies the type of tapping action currently completed (e.g., T1). ‌Next Action Number‌: Identifies the type of tapping action to be executed (e.g., T2). ‌Application Scenario‌: If the tapping sequence is ‌T1→T3→T2‌, the association action number information is as follows: First Group: Prev = T1, Next = T3 → Mark T1→T3. Second Group: Prev = T3, Next = T2 → Mark T3→T2. Scenario Description‌: ‌Tapping Sequence‌: Rough Tapping (T1) → Fine Tapping (T2) → Chamfering (T3). ‌Cleanliness Requirements‌: T1→T2: After roughing, aluminum chips must be thoroughly removed to prevent impact on fine finishing. T2→T3: After fine finishing, oil mist cooling is required to prevent burrs during chamfering. Execution Process: After T1 completes, the module detects the next action as T2 and generates the marker T1→T2. The command found is high-pressure air blowing (0.5 MPa, 2 seconds), with the nozzle cleaning in a spiral path. After T2 completes, the module detects the next action as T3 and generates the marker T2→T3. The command found is oil mist cooling (8 ml / s) + fixed-point air blowing (0.3 seconds).

[0044] It also includes a dynamic detection module, which includes a device detection unit 321 and a workpiece detection unit 131. The device detection unit 321 is used to detect the state feedback data of the tapping device 300 when it is working, and the workpiece detection unit 131 is used to detect the dynamic feedback data of the part to be processed when it is being processed. The intelligent interactive subsystem also includes a dynamic correction module, which is configured with a dynamic correction condition. When the dynamic correction condition is triggered, the dynamic correction module generates a dynamic correction instruction based on the state feedback data and the dynamic feedback data, and divides the target tapping sub-trajectory from the current tapping trajectory set to the adjacent target tapping trajectory set according to the dynamic correction instruction. The dynamic correction condition includes the device temperature sub-condition, the part temperature sub-condition and the part chip sub-condition. The dynamic detection module is composed of the device detection unit 321 and the workpiece detection unit 131, which collects the processing state data of the tapping device 300 and the workpiece in real time to provide input for dynamic correction. Device detection unit 321, function:

[0045] Monitor the physical status (vibration, temperature, current, etc.) of the tapping device 300 during operation to ensure processing stability. ‌Core Components‌:

[0046] Vibration Sensor: Detects the vibration amplitude of the robot arm's end (unit: mm / s²), with the threshold set to the ISO 10816 standard (e.g., ≤4.5 mm / s²). Temperature Sensor: Monitors the motor winding temperature (unit: °C), with the over-temperature threshold set to 80% of the rated value (e.g., if the motor is rated at 80°C, the alarm threshold is 64°C). Current Sensor: Collects the tapping drive motor's 322-V current (unit: A), using current fluctuations to determine load abnormalities (e.g., a sudden 20% increase in current is considered a stall risk). Encoder: Provides feedback on the robot arm's position accuracy (unit: μm). Errors exceeding the limit (e.g., ±5 μm) trigger calibration. Vibration Spectrum Analysis: Extracts characteristic frequencies and identifies resonance points using FFT (Fast Fourier Transform). Motor Health Assessment: Estimates tool wear based on the relationship between current and torque. Workpiece Detection Unit 131: Function: Monitors dynamic deformation, force, and positional offset during workpiece machining to ensure machining accuracy. Core Components: Force Sensor: Measures the axial force of the tapping head (unit: N). Exceeding the limit (e.g., steel machining force > 500 N) triggers protection. Laser Displacement Sensor: Detects workpiece position offset (unit: μm) in real time to compensate for mechanical errors. Strain Gauge: Attached to the workpiece surface, it detects local deformation (unit: με). Processing is suspended if deformation exceeds the limit (e.g., > 200 με). Dynamic Correction Module: Based on real-time detection data, the module determines whether to trigger correction conditions and replans the tapping trajectory to ensure machining safety and accuracy. Trigger Logic: Threshold Trigger: Any detected parameter exceeds the preset safety range (e.g., vibration > 4.5 mm / s²). Trend Trigger: Abnormal continuous rate of change of a parameter (e.g., temperature rise > 5°C per minute). Composite Trigger: Multiple parameter correlation anomalies (e.g., high vibration combined with high current). is the static cost of the i-th trajectory set, is the deviation penalty term from the original trajectory, and is the weight of the corresponding deviation penalty term. That is, the more abnormalities there are in the corresponding trajectory set, the higher the weight of the deviation penalty term. When the deviation of the trajectory set is small at other costs, it can be corrected, so that the execution instructions of the first group can be switched to the second group.

[0047] ‌Implementation method‌:‌ Pre-stored trajectory library‌: define multiple sets of tapping trajectory sets in advance (such as high-speed roughing set, low-speed finishing set).

[0048] Dynamic matching: Select the optimal trajectory set based on the correction conditions. Example:

[0049] Current trajectory set: High-speed roughing (feed rate 0.3mm / rev, speed 3000rpm).

[0050] ‌Vibration exceeding the limit detected‌ → Trigger correction → ‌Target trajectory set‌: Low-speed finishing (feed rate 0.1mm / rev, speed 1500rpm).

[0051] The intelligent interactive subsystem also includes a macro-correction module, which is equipped with a quality matching strategy. The quality matching strategy generates macro-correction parameters based on quality data and modifies the static cost algorithm based on the macro-correction parameters. The macro-correction module is equipped with a macro-correction database, which is equipped with a number of abnormality-related sub-items and corresponding abnormality-related values. The abnormality-related sub-items and abnormality-related values, the quality matching strategy includes calculating the total abnormality-related value under each abnormality-related item. The total abnormality-related value is the weighted result of the abnormality-related values ​​corresponding to the abnormality-related sub-items contained in the abnormality-related item. When the total abnormality-related value is greater than a preset abnormality trigger value, the macro-correction parameter is generated based on the total abnormality-related value. Macro-correction database, ‌‌Definition‌‌: The macro-correction database is a structured database that stores all associated sub-items related to processing abnormalities and their quantitative parameters (abnormality-related values), used to systematically analyze the root causes of quality problems and derive correction parameters. Abnormal Association Sub-items: Defines categories of independent factors that may lead to quality abnormalities, such as tool wear (sub-item ID: W01), temperature abnormalities (sub-item ID: T02), vibration exceeding the standard (sub-item ID: V03), workpiece deflection (sub-item ID: P04), part flash (sub-item ID: P02), and part breakage (sub-item ID: P15). Abnormal Association Value: A quantitative parameter for each sub-item, indicating the degree of impact of the abnormality on quality. Value range: 0–1 (0 = no impact, 1 = severe failure). Value Assignment Rule: Based on historical data statistics (e.g., tool wear value = accumulated usage time / life cycle). Real-time inspection data is normalized (e.g., vibration exceeding the standard value = current vibration amplitude / threshold). Definition: Analyzes quality data (e.g., machining dimensional error, surface roughness) to calculate the total abnormal association value, which is then used to generate correction parameters to optimize the static cost algorithm. Execution Steps: Data Mapping: Associates quality data with abnormal association sub-items. For example: hole diameter out of tolerance → associated tool wear (W01) and workpiece offset (P04). ‌Abnormal associated total value calculation‌:

[0052] : The total value of the j-th abnormality-related item (such as the "size abnormality" item).

[0053] : Sub-item weight (defined by the expert system, such as W01 weight = 0.6, P04 weight = 0.4).

[0054] : Sub-item associated value.

[0055] Trigger judgment: If > (such as 0.7), then generate macro correction parameters.

[0056] For example, if the hole diameter is out of tolerance, calculate the total correlation value: W01 correlation value = 0.8 (tool wear 80%), weight 0.6 → contribution value = 0.48; P04 correlation value = 0.5 (offset 50%), weight 0.4 → contribution value = 0.2. Total value =0.48+0.20=0. If the threshold =0.65, a correction is triggered. ‌About Macro Correction Parameters‌: Parameters used to adjust the weights or constraints of the static cost algorithm to specifically suppress high-frequency anomalies. ‌Generation Rule‌: ‌Linear Correction‌: Algorithm parameters are adjusted proportionally based on the total value of anomaly correlation. - ); where α is the correction coefficient (e.g. α=0.5), is the adjustment amount of the kth weight in the static cost algorithm. ‌Non-linear correction‌: For serious anomalies (such as >0.9), directly disable high-risk trajectory sets. ‌Parameter Type‌: ‌Weight Correction‌: Adjust the time, energy consumption, and wear weights in the static cost algorithm. ‌Constraint Correction‌: Tighten the vibration or temperature threshold (for example, reduce the vibration threshold from 4.5mm / s² to 3.5mm / s²). For example‌: If the tool is worn ( =0.68) triggers the correction, the original weight of the static cost algorithm is =0.4, =0.3 after correction: = +Δw=0.3+0.5(0.68-0.65)=0.315 Tool wear weight is increased, and the algorithm is more inclined to choose low-wear trajectories. Static cost algorithm correction‌

[0057] Definition: Embed the macro correction parameters into the original static cost algorithm to dynamically optimize the machining trajectory selection strategy. Correction formula: Original cost function:

[0058]

[0059] Corrected cost function:

[0060] Fine-tune the threshold to avoid vibration risks in advance. Module linkage process steps: Quality data input: The image recognition module of the blanking and sorting device provides quality data (such as aperture error 0.05mm). Abnormal correlation matching: The macro correction module matches the aperture error to tool wear (W01) and workpiece offset (P04). Total value calculation and triggering: =0.68>0.65→ Trigger correction. ‌Parameter Generation and Injection‌: Increase wear weight = +0.015 Reduce vibration threshold =−0.006 mm / s². ‌Static Cost Algorithm Update‌: The optimized algorithm prioritizes low-wear and low-vibration trajectories.

[0061] The unloading and sorting device is equipped with an abnormality feature database, which stores several quality anomalies, each indexed by a quality graphical feature. The image recognition strategy generates a capture command based on the machining difference model and controls the image recognition module to capture part images according to the capture command. The quality graphical features are then matched with the part images to obtain the corresponding quality difference items, which are then used to generate the quality data. The abnormality feature database is a structured storage system that records all known quality anomalies and their corresponding visual features (such as shape, texture, and color) for rapid identification of part machining defects. "Quality anomaly" defines the defect type in part machining, for example: "Aperture out of tolerance" (ID: D01): The aperture is outside the tolerance range (e.g., the standard is Φ10±0.05mm, but the measured diameter is Φ10.12mm). "Surface crack" (ID: D02): A visible linear crack on the material surface. "Thread burr" (ID: D03): An unremoved metal protrusion on the thread edge. ‌Quality Graphic Features‌ Each abnormality corresponds to an image feature description, including: ‌Geometric Features‌: The shape and size of the defect area (e.g., crack length > 2mm). ‌Texture Features‌: Roughness and contrast extracted by the Gray Level Co-occurrence Matrix (GLCM). ‌Color Features‌: The color difference between the abnormal area and the background (e.g., ΔH > 10° in the HSV color space). It is a mathematical model that determines whether to trigger image capture and analysis by comparing the difference between design parameters and measured data. ‌Input Parameters‌: Design drawing dimensions (e.g., aperture Φ10±0.05mm). Theoretical accuracy of processing equipment (e.g., positioning error ±3μm). ‌Difference Calculation‌: If the deviation of the processing parameters exceeds the theoretical accuracy range, the capture command is triggered. Image Recognition Strategy‌: ‌Function‌: Based on the output of the processing difference model, control the image recognition module to capture part images and identify quality abnormalities through feature matching.

[0062] Execution Steps: Capture command generation. Triggering Time: After machining is complete and the difference exceeds the limit. Shooting Parameters: Light source intensity (adjusted based on the material's reflectivity, e.g., 5000K cold light for aluminum alloy). Camera resolution (≥5 megapixels, accuracy corresponding to 0.01mm / pixel). Multi-angle capture (front view, side view, close-up). Image Feature Extraction: Preprocessing: Noise reduction, edge enhancement, and binarization. Feature Calculation: Hough transform detection of circular aperture contours. Canny algorithm extraction of crack edges. Local Binary Pattern (LBP) analysis of thread texture. Feature Matching: Extracted features are compared with indices in the abnormal feature database for similarity. Module Linkage Process Steps: Process Monitoring: Sensors collect machining dimension data in real time and input it into the machining difference model. Difference Out-of-Limit Detection: The model calculates Δx and, if it exceeds the limit, sends a capture command to the image recognition module. Image Acquisition and Analysis: Capture part images and extract features. Matching with the abnormal feature database generates a quality difference report. ‌Data Feedback‌: Quality data is uploaded to the macro correction module, triggering dynamic adjustment of processing parameters. For example, when batch processing flange parts, thread burrs (D03) appear continuously. ‌Process‌: ‌Difference model trigger‌: After processing, the thread diameter deviation is detected to be 0.1mm (> threshold 0.05mm) → trigger shooting. ‌Image Analysis‌: Extract the LBP texture features of the thread edge and compare them with the D03 index, SSIM=0.72. Determined as D03 abnormality, severity level 2. ‌Quality Data Generation‌: 10% of the parts in this batch are marked as having D03 defects. ‌Macro Correction Response‌: The macro correction module increases the tool compensation weight and increases the deburring process dwell time by 0.2 seconds. ‌Result‌: The D03 incidence rate in subsequent batches dropped to 2%.

[0063] Of course, the above are only typical examples of the present invention. In addition, the present invention may also have many other specific implementation methods. Any technical solutions formed by equivalent replacement or equivalent transformation fall within the scope of protection required by the present invention.

Claims

1. An automatic tapping equipment with multiple tapping stations, characterized by: It includes a transposition seat, a blanking and sorting device, a switching and cleaning device, and at least two tapping devices; The transposition seat includes a station turntable and a rotation drive member, and the rotation drive member is used to drive the station turntable to rotate; The transposition seat is provided with a loading station, a tapping station and a blanking station, the tapping station is provided corresponding to the tapping device, and the blanking and sorting device is provided at the blanking station; The tapping device includes a tapping robot arm and a tapping module, wherein the tapping robot arm is used to drive the tapping module to move at the tapping station, and the tapping module is used to perform the tapping action; The switching cleaning device is arranged between the tapping devices, and includes a reversing auxiliary part and a cleaning execution module. When the workpiece to be processed moves between the two tapping stations driven by the station turntable, the reversing auxiliary part drives the workpiece to be processed to rotate on the station turntable, and the cleaning execution module is used to perform the cleaning action; The transposition seat is provided with a plurality of fixed modules, and the fixed modules are used to fix the parts to be processed at the tapping station; The blanking and sorting device includes an image recognition module and a sorting module. The image recognition module is configured with an image recognition strategy, which is used to recognize the image of the processed parts to generate quality data, and generate sorting instructions based on the quality data and send them to the sorting module. The sorting module drives the processed parts to the corresponding blanking area according to the sorting instructions; It also includes an intelligent interaction subsystem, which includes a model generation module, an action matching module and an instruction generation module. The model generation module is used to generate a part simulation model based on the information of the part to be processed input by the user, and process the part simulation model through the processing target information to obtain a processing difference model. The action matching module is configured with a tapping configuration parameter library, and the tapping configuration parameter library stores a number of tapping constraint features. The tapping constraint features are indexed by the tapping device parameters. The action matching module obtains the corresponding tapping constraint features by substituting the tapping device parameters of different tapping devices, and generates tapping morphological constraints based on the tapping constraint features. The different tapping morphological constraints are matched with the processing difference model respectively to generate a tapping trajectory sequence. The tapping trajectory sequence includes a number of tapping sub-trajectories. The instruction generation module is configured with a static cost algorithm. The static cost algorithm is used to calculate the static cost of each tapping sub-trajectory, and use the total static cost value as a constraint to generate a tapping trajectory set according to adjacent tapping sub-trajectories, and generate corresponding tapping action instructions based on the tapping trajectory set to control the operation of the corresponding tapping device.

2. The multi-tapping station automatic tapping equipment according to claim 1, characterized in that: The transposition seat is provided with a fixed groove on the tapping station. When the part to be processed rotates to the tapping station, it falls into the fixed groove to limit the rotation of the part to be processed. The tapping device is provided with a suction head unit. When the suction head unit is working, it sucks the part to be processed to make it leave the fixed groove.

3. The multi-tapping station automatic tapping equipment according to claim 1, characterized in that: The image recognition module includes an image acquisition unit and a focus adjustment unit. The image acquisition unit is arranged on the focus adjustment unit. The focus adjustment unit is used to adjust the distance between the image acquisition unit and the processed part.

4. The multi-tapping station automatic tapping equipment according to claim 1, characterized in that: The intelligent interactive subsystem also includes a cleaning association module, which includes a cleaning instruction set and an action association table. The cleaning instruction set includes several cleaning instructions, and the cleaning instructions are indexed by action association tags. The action association table stores several action association tags, and the action association tags are indexed by associated action number information. The associated action number information is a front action number and a back action number. The cleaning association module obtains the corresponding front action number and back action number according to the adjacent tapping action instructions to call the cleaning instruction, and the cleaning instruction is used to control the operation of the cleaning execution module.

5. The multi-tapping station automatic tapping equipment according to claim 4, characterized in that: It also includes a dynamic detection module, which includes a device detection unit and a workpiece detection unit. The device detection unit is used to detect the state feedback data of the tapping device when it is working, and the workpiece detection unit is used to detect the dynamic feedback data of the part to be processed when it is being processed. The intelligent interactive subsystem also includes a dynamic correction module, and the dynamic correction module is configured with a dynamic correction condition. When the dynamic correction condition is triggered, the dynamic correction module generates a dynamic correction instruction based on the state feedback data and the dynamic feedback data, and divides the target tapping sub-trajectory from the current tapping trajectory set to the adjacent target tapping trajectory set according to the dynamic correction instruction.

6. The multi-tapping station automatic tapping equipment according to claim 5, characterized in that: The intelligent interaction subsystem further includes a macro correction module, wherein the macro correction module is configured with a quality matching strategy, wherein the quality matching strategy generates macro correction parameters according to quality data and corrects the static cost algorithm according to the macro correction parameters.

7. The multi-tapping station automatic tapping equipment according to claim 6, characterized in that: The dynamic correction conditions include a device temperature sub-condition, a component temperature sub-condition, and a component chip sub-condition.

8. The multi-tapping station automatic tapping equipment according to claim 7, characterized in that: The blanking and sorting device is equipped with an abnormality feature database, which stores a number of quality abnormality items, each of which is indexed by a quality graphic feature; The image recognition strategy generates a shooting instruction based on the processing difference model, and controls the image recognition module to shoot the part image according to the shooting instruction, matches the quality graphic features and the part image to obtain the corresponding quality difference items, and forms the quality data based on the quality difference items.

9. The multi-tapping station automatic tapping equipment according to claim 8, characterized in that: The macro correction module is configured with a macro correction database, and the macro correction database is configured with several abnormal correlation sub-items and corresponding abnormal correlation values. The abnormal correlation sub-items and abnormal correlation values, the quality matching strategy includes calculating the total abnormal correlation value under each abnormal correlation item, the total abnormal correlation value is the weighted result of the abnormal correlation values ​​corresponding to the abnormal correlation sub-items contained in the abnormal correlation item, when the total abnormal correlation value is greater than the preset abnormal trigger value, the macro correction parameter is generated according to the total abnormal correlation value.

Images