A riveting state control method and system of a rivet setting device and a storage medium

Abstract

The application relates to the technical field of mechanical arm control, and discloses a riveting state control method and system of a riveting device and a storage medium. The method comprises the following steps: when feeding, a mechanical arm drives a riveting gun to rotate into a riveting nut, the pressure and fluctuation are monitored based on a pressure sensor, the rotating process value is calculated, and an exception is prompted; when aligning, the mechanical arm moves to a riveting station, the riveting gun is driven to rotate back and forth, and the rotating frequency and the angle difference are dynamically adjusted according to the pressure parameter; when riveting, the riveting tension is adjusted according to the rotating process value, the nut is formed, and the mechanical arm and the riveting gun are controlled to move reversely to realize the exit of a rivet rod. The parameters can also be optimized through torque / displacement monitoring and spectrum analysis, retry or alarm is performed when an exception occurs, and the component damage is judged by comparing two spectrums. The application realizes accurate control of the whole riveting process, improves the quality and efficiency, and reduces the risk of rework and equipment damage.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of mechanical arm control, and in particular to a riveting state control method and system of a riveting device and a storage medium. BACKGROUND

[0002] With the development of the automobile industry towards light weight and high strength, the automobile riveting technology has evolved through several stages. In the early stage of automobile assembly, riveting was mostly dependent on manual operation, and simple tools were used to complete the connection, which was low in efficiency and poor in precision. With the wide application of steel plates, aluminum alloys and other materials, traditional riveting cannot meet the structural strength requirements, and semi-automatic special riveting mechanical arm equipment has emerged. In recent years, the key parts such as the vehicle body frame and chassis have suddenly increased the requirement for connection reliability, which has promoted the riveting process from ordinary riveting to precise methods such as riveting.

[0003] The large-scale production of the automobile manufacturing industry has given rise to the construction of the riveting automation manufacturing system. In the current automated production line, the riveting equipment is mostly operated in cooperation with industrial robots and automatic rivet feeding systems to realize the continuous operation of rivet feeding, positioning and riveting. The automatic system controls the riveting force and stroke through a preset program, which greatly improves the assembly efficiency and adapts to the multi-station batch production scene of automobile body-in-white and the like.

[0004] The existing automobile riveting quality control mostly adopts a post-detection mode, such as checking the rivet and workpiece combination state of the assembled rivet through X-ray flaw detection, ultrasonic detection and other means. Such a mode can only find defects after riveting is completed, and cannot monitor the riveting process of the mechanical arm in real time. When the mechanical arm with the parts appears abnormal deformation or poor fit, it is difficult to intervene in time, which may lead to rivet loosening, insufficient structural strength and other hidden dangers, affecting the safety performance of the whole vehicle, and the cost of post-rework is high, which is difficult to adapt to the high-precision manufacturing demand. SUMMARY

[0005] In order to improve the stability of the riveting process of the riveting mechanical arm, the present application provides a riveting state control method and system of a riveting device and a storage medium.

[0006] In a first aspect, the present application provides a riveting state control method of a riveting device, which adopts the following technical solution:

[0007] A riveting state control method of a riveting device, comprising the following steps:

[0008] Based on the arrival command of the riveting nut at the loading station, the robotic arm is controlled to move to the corresponding loading station. The robotic arm moves towards the loading station and simultaneously rotates the riveting gun on the robotic arm along the preset screwing direction. After reaching the set stroke, the end of the rivet rod is screwed into the riveting nut. During the screwing of the rivet rod into the riveting nut, the screwing pressure value on the rivet rod is acquired in real time based on the pressure sensor, and the pressure fluctuation value of the screwing pressure value is calculated. A pressure comparison value is calculated based on the screwing pressure value and the preset reference pressure value. A fluctuation comparison value is calculated based on the pressure fluctuation value and the preset reference fluctuation value. The screwing process value is calculated based on the pressure comparison value and the fluctuation comparison value. If the screwing process value is outside the preset screwing value range, a loading abnormality prompt is issued.

[0009] The robotic arm is controlled to move in the opposite direction from the loading station, and then to the position corresponding to the riveting station. The robotic arm is then controlled to move toward the riveting station to align with the riveting plate. The riveting gun is driven to rotate firstly along a preset first direction and then rotate secondly along a preset second direction and then rotate back and forth for a preset number of times. The number of rotations is adjusted according to the pressure comparison value; the larger the pressure comparison value, the more rotations are required, and the smaller the pressure comparison value, the fewer rotations are required. The angle difference between the first angle and the second angle is adjusted according to the pressure fluctuation value; the larger the pressure fluctuation value, the larger the angle difference, and the smaller the pressure fluctuation value, the smaller the angle difference.

[0010] The rivet rod is controlled to retract the rivet gun with a preset riveting pull force. The riveting pull force is adjusted according to the screw-in process value. The larger the screw-in process value, the larger the riveting pull force, and the smaller the screw-in process value, the smaller the riveting pull force. This makes the rivet nut form and press the plate tightly. The rivet gun is controlled to rotate in the opposite direction of screwing in, and the robotic arm is controlled to move in the opposite direction of the riveting station, so that the rivet rod is withdrawn from the rivet nut.

[0011] By adopting the above technical solutions, abnormalities are intercepted in advance during the material feeding stage, preventing unqualified materials from entering subsequent stages and reducing defects from the source; the alignment stage dynamically corrects deviations by adaptively adjusting parameters to correct potential misalignments before riveting, avoiding riveting defects caused by poor alignment; the riveting stage precisely matches the pulling force, optimizing the pulling force based on the previous state to ensure qualified riveting in one go and reduce rework caused by improper pulling force; the exit stage coordinates actions, controlling the reverse rotation of the rivet gun and the reverse movement of the robotic arm synchronously to avoid the rivet rod rubbing against the forming nut during exit, protecting the integrity of the formed structure.

[0012] Optionally, the step of driving the rivet gun to first rotate along a preset first direction by a preset first angle, and then rotate along a preset second direction by a preset second angle, and reciprocate for a preset number of times, further includes the following sub-steps:

[0013] The rotational pressure value of the rivet rod is obtained during the process of the rivet gun rotating at a first angle along a preset first direction, and the rotational pressure curve is obtained based on the rotational pressure value;

[0014] Calculate the shape similarity value between the rotational pressure curve and the preset reference pressure curve;

[0015] If the shape similarity value is less than the preset reference similarity value, the size of the first angle is adjusted according to the negative correlation of the shape similarity value. The larger the shape similarity value, the smaller the first angle, and the smaller the shape similarity value, the larger the first angle. Otherwise, the trend of the shape similarity value is calculated.

[0016] If the trend of change is increasing, the control will drive the rivet gun to first rotate along the preset second direction by a preset first angle, and then rotate along the preset first direction by a preset second angle, repeating the rotation a set number of times; otherwise, the size of the second angle will be adjusted according to the positive correlation of the shape similarity value. The larger the shape similarity value, the larger the second angle, and the smaller the shape similarity value, the smaller the second angle.

[0017] By adopting the above technical solution, a low similarity value indicates abnormal resistance in the first direction of the hole and difficulty in aligning the hole position. In this case, the first angle is increased by negative correlation to expand the search range. When the similarity value meets the standard, its trend is judged. If the trend increases, it indicates that the difficulty of the first direction of the hole has increased. The rotation sequence is immediately switched to the second direction first, and the more suitable second direction hole is passed first. If the trend is stable, the second angle is adjusted by positive correlation to adapt to the current hole state.

[0018] Optionally, the step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps:

[0019] The displacement of the rivet rod and the corresponding time data are obtained during the riveting process based on the displacement sensor.

[0020] The riveting displacement diagram and average riveting speed are calculated based on the riveting displacement and time data.

[0021] The similarity value of the displacement diagram is calculated based on the riveting displacement diagram and the preset reference displacement diagram;

[0022] The displacement map comparison value is calculated based on the displacement map similarity value and the preset reference similarity value;

[0023] The riveting speed comparison value is calculated based on the average riveting speed and the preset reference average speed;

[0024] The riveting process values ​​are calculated based on the comparison values ​​of the displacement diagram and the riveting speed.

[0025] If the riveting process value is within the preset riveting reference range, the rotation speed of the riveting gun in the opposite direction of the screwing direction and the movement speed of the robotic arm in the opposite direction of the riveting station are adjusted according to the positive correlation of the riveting process value. The larger the riveting process value, the faster the rotation speed and movement speed, and the smaller the riveting process value, the slower the rotation speed and movement speed.

[0026] Otherwise, issue an abnormal riveting warning.

[0027] By adopting the above technical solution and monitoring displacement and velocity in two dimensions, a real-time judgment and exit parameter adjustment mechanism for riveting quality is constructed to accurately ensure the qualification of riveting.

[0028] Optionally, the step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps:

[0029] The displacement of the rivet rod and the corresponding time data are obtained during the riveting process based on the displacement sensor.

[0030] A riveting displacement diagram is calculated based on the riveting displacement and time data; a riveting spectrum diagram is then calculated based on the riveting displacement diagram using a spectrum conversion algorithm.

[0031] The riveting frequency variation amplitude and riveting energy variation amplitude values ​​were calculated based on the riveting spectrum diagram.

[0032] The riveting frequency comparison value is calculated based on the riveting frequency change amplitude value and the preset riveting reference frequency change value.

[0033] The riveting energy comparison value is calculated based on the riveting energy change amplitude value and the preset riveting reference energy change value;

[0034] The amplitude of the riveting spectrum variation is calculated based on the comparison values ​​of riveting frequency and riveting energy.

[0035] If the riveting spectrum change amplitude is less than the preset riveting reference change amplitude, the rotation speed of the rivet gun in the opposite direction of the screwing direction and the movement speed of the robotic arm in the opposite direction of the riveting station are adjusted according to the negative correlation of the riveting spectrum change amplitude. The larger the riveting spectrum change amplitude, the slower the rotation speed and movement speed, and the smaller the riveting spectrum change amplitude, the faster the rotation speed and movement speed.

[0036] By adopting the above technical solution, the micro-dynamic characteristics of the riveting process can be mined through spectrum analysis, enabling more accurate quality judgment and exit parameter adjustment, thereby further improving the reliability of riveting.

[0037] Optionally, the method further includes the following steps during the process of screwing the rivet into the rivet nut:

[0038] Data on riveting torque and time during the process of screwing the rivet into the rivet nut are obtained based on a torque sensor.

[0039] The torque variation diagram and the average speed of torque variation were calculated based on the riveting torque and time data.

[0040] The torque diagram similarity value is calculated based on the torque variation diagram and the preset reference torque diagram;

[0041] The torsion diagram comparison value is calculated based on the torsion diagram similarity value and the preset reference similarity value;

[0042] The torque-speed comparison value is calculated based on the average speed of torque change and the preset reference average speed.

[0043] The torque process value is calculated based on the torque diagram comparison value and the torque change comparison value.

[0044] If the torque process value is within the preset torque reference range, the rotation speed of the rivet rod screwing into the rivet nut is adjusted according to the positive correlation of the torque process value. The larger the torque process value, the faster the rotation speed, and the smaller the torque process value, the slower the rotation speed.

[0045] Otherwise, issue a torque abnormality warning.

[0046] By adopting the above technical solution, the torque change during the screwing-in process of the rivet is monitored in real time by a torque sensor, so as to achieve dynamic adaptation of the screwing-in speed and abnormal early warning, thereby improving the feeding and screwing quality.

[0047] Optionally, the method further includes the following steps during the process of screwing the rivet into the rivet nut:

[0048] Data on riveting torque and time during the process of screwing the rivet into the rivet nut are obtained based on a torque sensor.

[0049] The riveting torque-displacement diagram is calculated based on the riveting torque and time data; the riveting torque spectrum diagram is then calculated using a spectrum conversion algorithm based on the riveting torque-displacement diagram.

[0050] The torque frequency variation amplitude and torque energy variation amplitude are calculated based on the riveting torque spectrum diagram;

[0051] The torque frequency comparison value is calculated based on the torque frequency change amplitude value and the preset torque reference frequency change value.

[0052] The torque energy comparison value is calculated based on the torque energy change amplitude value and the preset torque reference energy change value.

[0053] The torque spectrum variation amplitude of the riveting torque spectrum diagram is calculated based on the torque frequency comparison value and the torque energy comparison value.

[0054] If the torque spectrum change amplitude is less than the preset torque reference change amplitude, the rotation speed of the rivet screw into the rivet nut is adjusted according to the negative correlation of the torque spectrum change amplitude. The larger the torque spectrum change amplitude, the slower the rotation speed and the moving speed; the smaller the torque spectrum change amplitude, the faster the rotation speed and the moving speed.

[0055] Otherwise, issue an abnormal riveting warning.

[0056] By adopting the above technical solution and using spectrum analysis to deeply analyze the dynamic characteristics of torque during rivet screwing, the screwing speed can be precisely controlled and abnormal warnings can be provided, further ensuring the quality of thread engagement.

[0057] Optionally, the method further includes the following steps:

[0058] If, during the process of screwing the rivet rod into the rivet nut, the torque spectrum change amplitude value calculated based on the riveting torque data obtained by the torque sensor exceeds the preset torque reference change amplitude value, it is recorded as the first screwing abnormality. Then, the robotic arm is controlled to drive the rivet rod out of the rivet nut and the operation of screwing the rivet rod into the rivet nut is re-executed for a second screwing.

[0059] If the torque spectrum change amplitude value calculated during the second screwing process still exceeds the preset torque reference change amplitude value, an alarm will be triggered and the riveting operation will be stopped.

[0060] By adopting the above technical solution, if the first screwing is abnormal, the accidental factor can be eliminated by screwing it in a second time, avoiding efficiency loss caused by misjudgment due to a single fluctuation; if the second screwing is still abnormal, it is determined to be a persistent problem, and an alarm will be triggered to stop the machine immediately, preventing further damage to the rivet rod, nut or robotic arm caused by forced screwing.

[0061] Optionally, the method further includes the following steps:

[0062] During the process of screwing in the rivet nut for the second time, the riveting torque spectrum diagram of the second screwing in is obtained, and the frequency energy change difference between the riveting torque spectrum diagram of the first screwing in and the riveting torque spectrum diagram of the second screwing in is calculated.

[0063] The difference value is obtained by subtracting the energy values ​​of corresponding frequency bands from two spectrum histograms and then weighting and summing the energy values ​​of non-overlapping frequency bands according to their two-dimensional coordinates, which include frequency and energy values.

[0064] If the difference in energy change at this frequency exceeds a preset difference threshold, a warning will be issued indicating damage to the rivet or rivet nut.

[0065] By adopting the above technical solution, the energy difference between the corresponding frequency bands of the two spectrum diagrams is calculated, and the non-overlapping frequency bands are weighted and accumulated according to the two-dimensional coordinates of frequency and energy to quantify the frequency energy change difference; this difference can reflect the degree of change in the dynamic characteristics of torque during the two rotation processes.

[0066] Secondly, this application provides a riveting status control system for a riveting device, which adopts the following technical solution:

[0067] A riveting state control system for a riveting device includes a processor, wherein the processor executes the steps of the riveting state control method for the riveting device as described in any one of the preceding claims.

[0068] Thirdly, this application provides a storage medium, which adopts the following technical solution:

[0069] A storage medium storing a program, which, when executed by a processor, implements the steps of the riveting state control method for the riveting device described in any one of the preceding claims.

[0070] In summary, this application includes at least one of the following beneficial technical effects: By accurately sensing and optimizing parameters in each stage of feeding, alignment, riveting, and withdrawal, quality control is transformed from post-production inspection to in-process prevention, significantly reducing the riveting defect rate and rework costs. Through comprehensive analysis of multi-dimensional data such as pressure fluctuations and torque spectrum changes, adaptive adjustment of key parameters such as screw-in speed, riveting pull force, and withdrawal speed is achieved, ensuring both connection strength and forming quality while also considering production efficiency. Attached Figure Description

[0071] Figure 1 This is a step diagram illustrating the riveting status control method for a riveting device.

[0072] Figure 2 The diagram shows the sub-steps of driving the rivet gun to first rotate along a preset first direction by a preset first angle, and then rotate along a preset second direction by a preset second angle, repeating the rotation a preset number of times.

[0073] Figure 3 This is a diagram showing the sub-steps of controlling the rivet rod to retract the rivet gun with a preset riveting tension. Detailed Implementation

[0074] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0075] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0076] This application discloses a method for controlling the riveting state of a riveting device, referring to... Figure 1 It includes the following steps:

[0077] Upon receiving the instruction that the rivet nut has reached the loading station, the precise displacement control of the robotic arm is triggered. The robotic arm is driven to move along a preset path to a position perfectly aligned with the loading station, ensuring the initial alignment accuracy of the subsequent screwing action. Two operations are executed simultaneously: on the one hand, the robotic arm is controlled to slowly move towards the loading station, gradually approaching the rivet nut; on the other hand, the rivet gun at the end of the robotic arm is driven to rotate stably along a preset screwing direction, generally clockwise. Through the coordinated movement and rotation, the end of the rivet rod is precisely aligned with the threaded hole of the nut, and when the robotic arm reaches the preset stroke, the initial screwing of the rivet rod and the rivet nut is completed.

[0078] Throughout the entire process of screwing the rivet into the rivet nut, high-frequency pressure data is collected via a pressure sensor built into the rivet gun to obtain the real-time screwing pressure value on the rivet, such as the resistance pressure during thread engagement. Based on the collected pressure data, pressure fluctuation and pressure comparison values ​​are further calculated. The pressure fluctuation value, i.e., the change in screwing pressure per unit time, is the difference between the pressure fluctuation value and the preset reference fluctuation value, used to reflect the stability of the engagement process, such as whether there is thread jamming or obstruction by foreign objects. The pressure comparison value is calculated by comparing the real-time screwing pressure value with the preset reference pressure value, which is set according to the standard engagement pressure range based on the nut model and material, to determine whether the current screwing pressure meets the standard.

[0079] By combining pressure comparison values ​​and fluctuation comparison values, and using a preset algorithm, such as a normalized weighted summation algorithm, the screwing-in process value is calculated. This value comprehensively reflects the overall fit between the rivet and the nut. If the screwing-in process value exceeds the preset screwing-in range, such as excessive pressure indicating abnormal thread interference or excessive pressure fluctuation indicating unstable engagement, an abnormal feeding warning is immediately triggered, such as an audible and visual alarm and a pop-up window in the central control system. At the same time, the robotic arm operation is paused to prevent unqualified materials, such as mismatched nut models or damaged threads, from entering subsequent stages, thus reducing riveting defects caused by feeding problems at the source.

[0080] After loading is completed, the system controls the robotic arm to move in the opposite direction of the loading station, leaving the loading area. Then, according to the preset riveting station coordinates, the robotic arm is driven to move to the position corresponding to the riveted sheet metal. To ensure precise alignment between the rivet rod and the riveting hole on the sheet metal, the robotic arm is controlled to slowly move towards the riveting station, while the riveting gun is driven to rotate back and forth in a pattern of first rotating in a first direction (e.g., counterclockwise) by a preset first angle (e.g., 30°), and then rotating in a second direction (e.g., clockwise) by a preset second angle (e.g., 20°), until the preset number of rotations is reached.

[0081] In this process, instead of using fixed parameters, the rotation parameters are adaptively adjusted based on the pressure comparison and pressure fluctuation values ​​obtained during the feeding stage.

[0082] Rotation frequency adjustment: The pressure comparison value is positively correlated with the number of rotations. If the pressure comparison value is large, it means that the screwing pressure deviates significantly from the standard during feeding, and there may be a slight misalignment between the rivet and the nut. In this case, increase the number of reciprocating rotations, such as from 5 to 8 times, to gradually correct the deviation between the rivet and the hole position on the plate through more rotations. If the pressure comparison value is small, the screwing pressure is close to the standard, and the initial alignment is good, then reduce the number of rotations to avoid excessive movement and wasting time, while balancing alignment accuracy and efficiency.

[0083] Angle difference adjustment: The pressure fluctuation value is positively correlated with the difference between the first and second angles. If the pressure fluctuation value is large, it indicates that the screwing process is unstable during feeding, and there may be slight deformation of the rivet or slight misalignment of the hole. In this case, the angle difference should be increased, such as from 10° to 25°, to increase the probability of hole alignment by covering a larger range of rotation. If the pressure fluctuation value is small, the screwing process is stable and the risk of hole deviation is low. In this case, the angle difference should be reduced to achieve fine alignment and avoid new deviations caused by excessive angle.

[0084] Upon entering the riveting stage, the rivet rod is first controlled to retract into the rivet gun with a preset initial riveting tension. Simultaneously, based on the screw-in process value calculated during the feeding stage, the riveting tension is optimized in real time. The screw-in process value is positively correlated with the riveting tension: if the screw-in process value is large, it indicates that the feeding and screwing state is close to the ideal value, and the rivet rod and nut have good compatibility. In this case, the riveting tension is appropriately increased, such as from 5kN to 7kN, to ensure that the rivet nut can deform fully, tightly press the plate, and improve the connection strength. If the screw-in process value is small, it indicates that there is a slight abnormality in the feeding and screwing, such as a large thread gap. In this case, the riveting tension is reduced to avoid rivet rod breakage or plate damage due to excessive tension, ensuring the safety of the riveting process.

[0085] After the rivet nut is formed and clamped to the sheet metal, the system enters the exit phase. At this time, coordinated action control is executed: on the one hand, the rivet gun is controlled to rotate in the opposite direction of the screw-in direction (such as counterclockwise) to loosen the engagement between the rivet rod and the formed nut; on the other hand, the robotic arm is simultaneously controlled to move in the opposite direction of the riveting station, driving the rivet rod to smoothly exit the rivet nut. Through the synchronous design of reverse rotation and reverse movement, it is possible to avoid the rivet rod rubbing against the formed nut during exit, such as unilateral force causing the nut to be warped or scratched, effectively protecting the integrity of the formed structure and ensuring the stability of the connection after riveting.

[0086] The above solutions enable pressure monitoring and anomaly alerts during the material feeding stage, intercepting substandard materials at the source and reducing subsequent rework; dynamic parameter adjustment during the alignment stage proactively corrects potential misalignments, avoiding riveting defects caused by poor alignment; adaptive force optimization during the riveting stage ensures one-time qualified forming, reducing quality problems caused by improper force; and coordinated actions during the exit stage protect the formed structure and guarantee the final connection quality.

[0087] Reference Figure 2 The step of driving the rivet gun to rotate firstly along a preset first direction by a preset first angle, and then along a preset second direction by a preset second angle, repeating this rotation a preset number of times, also includes the following sub-steps:

[0088] When the rivet gun begins to rotate at a preset angle in the first direction, such as 30° counterclockwise, the system simultaneously activates the pressure sensor built into the rivet rod to collect the rotational pressure value at a high frequency, such as 50 times per second. This rotational pressure value directly reflects the contact resistance between the rivet rod and the hole in the sheet metal. If the rivet rod is aligned with the hole, the rotational pressure will remain in a stable low-resistance range. If there is a misalignment, such as the end of the rivet rod touching the surface of the sheet metal, a sudden increase or fluctuation in pressure will occur.

[0089] The collected rotational pressure values ​​are integrated in chronological order to generate a rotational pressure curve. The horizontal axis of the curve represents rotation time, corresponding to the progress of rotation angle, while the vertical axis represents the real-time rotational pressure value. This curve allows for a direct observation of the dynamic changes in resistance during rotation in the first direction, such as the presence of pressure peaks and fluctuation frequencies, providing data support for subsequent compatibility assessments.

[0090] After generating the rotational pressure curve, the system calls the preset reference pressure curve, which is generated based on a large amount of pressure data from qualified alignment scenarios. Its characteristics are: stable pressure, no obvious peaks, and fluctuation amplitude less than the preset threshold. The system calculates the shape similarity value between the two curves using curve shape similarity algorithms, such as dynamic time warping algorithm and Pearson correlation coefficient method.

[0091] The shape similarity value typically ranges from 0 to 1. The closer it is to 1, the better the actual rotation pressure curve matches the ideal reference curve, indicating that the hole resistance is normal and the hole alignment probability is high when rotating in the first direction. If the similarity value is less than the preset reference similarity value, such as 0.7, it indicates that there is an abnormality in the actual rotation pressure. For example, a sudden increase in pressure indicates that the rivet rod is hitting the plate, and high-frequency fluctuations indicate that the hole is blocked. The hole adaptability in the first direction is poor, and the parameters need to be adjusted in time.

[0092] Based on the magnitude and trend of shape similarity values, the system adaptively adjusts the rotation parameters in three scenarios to ensure precise alignment under different insertion conditions:

[0093] Scenario 1: The shape similarity value is lower than the reference similarity value; expand the first angle to increase the search range:

[0094] When the calculated shape similarity value is less than the preset reference similarity value, the system determines that the hole-entry resistance in the first direction is abnormal and the hole alignment is difficult. At this time, the system adjusts the parameters according to the rule of "adjusting the first angle with a negative correlation to the shape similarity value": if the shape similarity value is smaller, such as 0.5, it indicates that the hole-entry resistance is more severe, such as severe rivet misalignment, then the first angle is increased more, such as from 30° to 45°, to increase the probability of the rivet finding the hole position in the plate by expanding the rotation angle range; if the shape similarity value is close to the reference similarity value, such as 0.68, then the first angle is slightly increased, such as from 30° to 35°, to expand the search range while avoiding excessive rotation that would lead to efficiency loss.

[0095] After adjustment, the rivet gun will rotate again along the first direction at the new first angle. The system will collect pressure data and analyze similar values ​​again until the similar values ​​meet the standard or other adjustment logic is triggered to ensure that the hole resistance returns to normal.

[0096] Scenario 2: The shape similarity value meets the standard and the trend of change is increasing; switch the rotation order, prioritizing the second direction for entering the hole:

[0097] If the shape similarity value is greater than or equal to the reference similarity value, such as 0.72, the system further calculates the changing trend of the shape similarity value through time series analysis. For example, by comparing the similarity value of the previous 5 rotation cycles, it determines whether it is continuously rising, falling, or stable. If the changing trend is increasing, such as the similarity value rising from 0.72 to 0.85 and then to 0.9, it indicates that the difficulty of drilling in the first direction is gradually increasing. This may be due to slight offset of the hole position in the sheet metal, slight wear at the end of the rivet, etc., which will cause the subsequent rotation resistance to continue to increase.

[0098] To prevent further escalation of the difficulty in entering the hole, the system immediately triggers a rotation sequence switch: instead of rotating the riveting gun first along the primary direction, it first rotates along a preset secondary direction by a preset primary angle, such as 30° clockwise, and then rotates along the primary direction by a preset secondary angle of 20°, maintaining the same number of reciprocating rotations. By prioritizing the use of the more suitable secondary direction for entering the hole, which offers lower resistance and better similarity, the difficulty of entering the hole is quickly reduced, ensuring alignment efficiency and accuracy.

[0099] Scenario 3: The shape similarity value meets the standard and the trend of change is stable; adjust the second angle to adapt to the current state:

[0100] If the shape similarity value is greater than or equal to the reference similarity value and the trend remains stable, such as a stable similarity value of around 0.8 without significant increase or decrease, the system determines that the hole in the first direction is in good condition, but further optimization of the alignment effect is needed by adjusting the second angle. At this time, the system adjusts the second angle according to the rule of positive correlation between the shape similarity value and the shape similarity value: if the shape similarity value is larger, such as 0.9, it means that the hole condition is closer to the ideal value and the deviation between the rivet and the hole position is extremely small, so the second angle is increased more, such as from 20° to 30°, to enhance the ability to correct small deviations by increasing the rotation angle in the second direction; if the shape similarity value is close to the reference similarity value, such as 0.75, the second angle is slightly increased, such as from 20° to 25°, to ensure the correction effect while avoiding excessive rotation that would affect the alignment rhythm.

[0101] Compared to traditional fixed-angle rotation, it can sense subtle changes in the resistance to the hole in real time. By dynamically adjusting the first angle, switching the rotation sequence, and optimizing the second angle, it effectively solves problems such as obstruction of the hole in the first direction or increased difficulty in hole insertion, and greatly improves the alignment success rate between the rivet and the hole position of the plate. At the same time, through fine parameter adjustment, it avoids the efficiency loss caused by blindly increasing the rotation range or number of rotations, and takes into account both alignment accuracy and production efficiency.

[0102] Reference Figure 3 The step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps:

[0103] When the rivet rod begins to retract the rivet gun with a preset tension, the system simultaneously activates the high-precision displacement sensors built into the rivet gun, such as laser displacement sensors and grating displacement sensors, to perform high-frequency synchronous acquisition of the riveting displacement and corresponding time data throughout the entire riveting process. The acquisition frequency can reach over 100Hz. The riveting displacement refers to the real-time distance the rivet rod moves from its initial position to its retracted target position, directly reflecting the forming progress of the rivet nut. If the displacement reaches a preset threshold, it indicates that the nut has completed plastic deformation and clamped the sheet metal. The time data is used to calculate the rate of displacement change.

[0104] The collected displacement-time data are integrated sequentially to form a riveting displacement graph. This graph, with time on the horizontal axis and riveting displacement on the vertical axis, visually presents the trend of displacement change over time during the riveting process. Ideally, the displacement graph should show a curve that first rises rapidly and then stabilizes, corresponding to the rapid forming of the nut and the subsequent stable tightening. Simultaneously, the system automatically calculates the average riveting speed based on the formula of total displacement / total riveting time. This speed reflects the overall smoothness of the riveting process. If the average speed deviates from the standard range, it may indicate tension fluctuations or abnormal material resistance.

[0105] To accurately determine the quality of riveting, riveting process values ​​are generated through multi-step calculations based on the collected displacement and velocity data. The specific calculation logic is as follows:

[0106] Step 1: Calculate the similarity value of the displacement map to determine the compliance of the forming trajectory:

[0107] A preset reference displacement map is invoked. This map is generated based on a large number of qualified riveting scenarios and is characterized by: a stable slope with no obvious jamming inflection points during the displacement rise phase, and displacement fluctuations of less than ±0.1mm during the steady phase. A curve similarity algorithm, such as the cosine similarity algorithm or the dynamic time warping algorithm, is used to calculate the displacement similarity value between the actual riveting displacement map and the reference displacement map. The similarity value ranges from 0 to 1. The closer it is to 1, the better the actual displacement trajectory matches the ideal trajectory, indicating a smooth nut forming process without problems such as local jamming or excessive deformation. If the similarity value is low, such as less than 0.7, it indicates an abnormal forming trajectory, such as a sudden increase or decrease in displacement, which may be caused by uneven sheet thickness or unbalanced force on the rivet.

[0108] Step 2: Generate displacement map comparison values ​​to quantify the degree of trajectory deviation.

[0109] The system compares the displacement map similarity value with a preset reference similarity value, such as 0.8, and calculates the difference based on different nut models and plate materials to obtain the displacement map comparison value. The calculation formula can be set as: Displacement map comparison value = (Actual displacement map similarity value - Reference similarity value) / Reference similarity value. If the comparison value is positive and larger, it indicates that the actual trajectory is better than the reference trajectory; if it is negative and the absolute value is larger, it indicates that the trajectory deviation is more serious and requires close attention.

[0110] Step 3: Calculate the riveting speed comparison value and evaluate the smoothness of the process.

[0111] The system compares the actual average riveting speed with a preset reference average speed to generate a riveting speed comparison value. The reference average speed, for example, is 5 mm / s, set based on the material's plastic deformation characteristics. The calculation formula is: Riveting speed comparison value = (Actual average speed - Reference average speed) / Reference average speed. If the comparison value is within ±0.1, it indicates that the riveting speed is stable and conforms to the material's deformation rhythm. If the comparison value exceeds this range, such as being greater than 0.2 or less than -0.2, it indicates that the speed is too fast, which may lead to insufficient nut forming; or that the speed is too slow, which may lead to excessive stress on the material.

[0112] Step 4: Integrate and generate riveting process values ​​to comprehensively evaluate riveting quality:

[0113] The system employs a normalized weighted summation algorithm to integrate the displacement map comparison value and the riveting speed comparison value into a riveting process value. The weights can be set according to actual needs, such as 60% for the displacement map comparison value and 40% for the speed comparison value. Example formula: Riveting process value = 0.6 × Displacement map comparison value + 0.4 × Riveting speed comparison value. This value directly reflects the overall level of riveting quality. If it falls within the preset riveting reference range (e.g., -0.1 to 0.2), it indicates that the riveting is qualified; if it exceeds the range, it is judged as an abnormal riveting process.

[0114] Based on the results of the riveting process values, the system precisely adjusts the parameters of the subsequent exit action in two scenarios to ensure that the exit process is both efficient and protects the formed structure:

[0115] Scenario 1: The riveting process value is qualified; adjust the exit speed accordingly.

[0116] When the riveting process value is within the preset reference range, the system determines that the riveting formation meets the standard. At this time, according to the rule of adjusting the exit speed positively correlated with the riveting process value, two parameters are adjusted simultaneously:

[0117] Riveting gun reverse rotation speed: The higher the value during the riveting process, such as 0.2, the better the riveting forming quality, the more the displacement trajectory follows the ideal curve, the more stable the speed, and the more stable the connection between the rivet nut and the plate. The rotation speed of the rivet gun in the reverse direction of screwing can be increased, such as from 100 r / min to 150 r / min, to speed up the thread loosening efficiency. If the value during the riveting process is too small, such as -0.05, it means that the forming quality is close to the lower limit of acceptance. If the displacement fluctuation is slightly large, the rotation speed needs to be slowed down, such as reduced to 80 r / min, to avoid slight loosening at the connection between the nut and the plate due to excessive speed.

[0118] The reverse movement speed of the robotic arm is consistent with the adjustment logic of the riveting gun rotation speed. The larger the value of the riveting process, the faster the robotic arm moves in the reverse direction of the riveting station. For example, increasing it from 20mm / s to 30mm / s shortens the exit time. When the value of the process is small, the movement speed is slowed down. For example, reducing it to 15mm / s prevents the rivet rod from scraping against the formed nut due to excessive movement, which could damage the nut structure that is not fully stable. This is especially important for materials with high plasticity such as aluminum alloys.

[0119] Scenario 2: Abnormal values ​​during the riveting process trigger an immediate warning and interception:

[0120] If the riveting process value exceeds the preset reference range, such as greater than 0.3 or less than -0.2, the system immediately determines that the riveting quality is unqualified and executes dual abnormality handling actions:

[0121] Activate riveting anomaly warning: Through pop-up windows and audible and visual alarms in the central control system, such as flashing red warning lights and buzzer prompts, the anomaly type is fed back to the operator in real time to help quickly locate the root cause of the problem. For example, a warning based on a low similarity value of the displacement diagram indicates an abnormal nut forming trajectory, and a warning based on an abnormal speed comparison value indicates excessive fluctuation in riveting speed.

[0122] Pause and retract riveting actions: The system automatically controls the rivet rod to stop retracting and freezes the subsequent movements of the robotic arm and rivet gun to prevent unqualified riveted structures from entering the next stage. Forcibly retracting the riveting could cause the nut to fall off or damage the sheet metal. Simultaneously, the system automatically stores abnormal data (such as displacement diagrams, speed curves, and riveting process values) in the database for subsequent quality traceability and process optimization, such as analyzing whether the abnormality is due to improper tension parameter settings or material batch issues.

[0123] Compared to traditional riveting control that relies solely on tension monitoring, this system can accurately identify abnormalities in the nut forming trajectory, such as localized insufficient tightening, through displacement diagram analysis. Speed ​​monitoring can also detect issues with the smoothness of the riveting process, such as uneven material stress caused by sudden speed changes. This dual-dimensional approach significantly improves the accuracy of riveting quality assessment. Positive correlation adjustment of exit parameters avoids fixed-speed modes, preventing both inefficiency due to excessively slow speeds and damage to the formed structure due to excessively fast speeds. It is particularly suitable for the diverse needs of riveting scenarios involving multiple materials and specifications, such as automotive body-in-white. Anomaly warning and action pause mechanisms enable immediate interception of defective products, reducing subsequent rework costs, such as eliminating the need to disassemble already assembled body structures. Furthermore, it provides data support for process optimization, helping to continuously improve the riveting pass rate.

[0124] The step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps:

[0125] After riveting begins, the displacement sensor continuously collects the riveting displacement of the rivet rod and the corresponding time data, generating a riveting displacement graph reflecting macroscopic changes, with time on the horizontal axis and displacement on the vertical axis. To further explore the microscopic features in the data, the system introduces a spectrum conversion algorithm, such as Fast Fourier Transform (FFT), to transform the displacement graph from a time-displacement time-domain signal into a frequency-energy frequency-domain signal, generating a riveting spectrum graph.

[0126] In this spectrum graph, the horizontal axis represents frequency, reflecting the periodicity of displacement fluctuations. For example, high-frequency components represent rapid, minute vibrations, while low-frequency components represent slow deformation. The vertical axis represents energy, reflecting the intensity of the corresponding frequency fluctuation; higher energy indicates more significant fluctuations at that frequency. The spectrum graph allows for a direct visual identification of the micro-dynamics during the riveting process. For instance, in a normal riveting spectrum graph, low-frequency energy is concentrated and stable, while high-frequency energy is weak. If there is rivet jamming, a sudden increase in high-frequency energy will occur.

[0127] The system extracts two characteristic parameters from the riveting spectrum and quantifies the microscopic stability of the riveting process by comparing them with reference values:

[0128] Riveting frequency variation amplitude: Calculate the distribution range and fluctuation degree of the main frequency components (such as 0-10Hz) in the spectrum diagram to reflect the periodic stability of the riveting action. If the frequency variation amplitude is small, it indicates that the rivet rod moves at a uniform rhythm without sudden jamming.

[0129] Riveting energy variation amplitude: The coefficient of variation of energy values ​​in each frequency band is statistically analyzed to reflect the uniformity of riveting force transmission. If the energy variation amplitude is small, it indicates that the plate and nut are subjected to stable force and there is no local stress concentration.

[0130] The system compares the two amplitude values ​​mentioned above with preset riveting reference frequency change values ​​and riveting reference energy change values, respectively. The riveting reference energy change value is calibrated based on the spectral characteristics of qualified riveting samples, and calculates the riveting frequency comparison value (the ratio of the actual frequency amplitude to the reference value) and the riveting energy comparison value (the ratio of the actual energy amplitude to the reference value). By weighted summation, such as a 40% weighting for the frequency comparison value and a 60% weighting for the energy comparison value, the final riveting spectral change amplitude value is obtained. The smaller this value, the closer the micro-dynamics of the riveting process is to the ideal state, and the better the forming quality, such as the absence of minor vibrations and uniform stress; conversely, a larger value indicates the presence of micro-anomalies, such as slight rivet wobbling or uneven local deformation of the material.

[0131] Based on the comparison between the amplitude of the riveting spectrum change and the preset reference amplitude, the exit parameters are dynamically adjusted:

[0132] If the amplitude of the spectrum change is less than the reference value, it indicates that the microscopic condition of the riveting is qualified. Then, the speed is adjusted according to the logic of negative correlation: the larger the amplitude of the spectrum change, the closer it is to the reference value, indicating that there are slight microscopic fluctuations in the riveting process, such as slightly higher high-frequency energy and slightly weaker microscopic stability of the formed structure. At this time, the reverse rotation speed of the rivet gun and the reverse movement speed of the robotic arm are slowed down, such as reducing the rotation speed from 120 r / min to 90 r / min and the movement speed from 25 mm / s to 18 mm / s, to avoid microscopic defects caused by disturbances in the withdrawal action, such as tiny gaps at the joint between the nut and the sheet metal; the smaller the amplitude of the spectrum change, the lower it is than the reference value, indicating that the riveting process is extremely stable, such as concentrated low-frequency energy and no high-frequency interference, and stable formed structure. The withdrawal speed can be accelerated, such as increasing the rotation speed to 150 r / min and the movement speed to 30 mm / s, to improve production efficiency while ensuring quality.

[0133] The amplitude of spectral changes can identify anomalies that are difficult to show in displacement diagrams, such as microscopic jamming of rivets and nuts, and uneven stress transmission inside the sheet metal, providing early warning of potential defects. Speed ​​adjustment based on microscopic stability avoids exit damage caused by macroscopic compliance but microscopic fragility, such as the propagation of microcracks in aluminum alloy sheets. It is especially suitable for parts with stringent requirements for connection strength, such as automobile chassis and body frames, ensuring the overall structural safety through microscopic-level control.

[0134] The method also includes the following steps during the process of screwing the rivet into the rivet nut:

[0135] When the rivet rod begins to screw into the rivet nut, the high-precision torque sensor built into the rivet gun is activated simultaneously. This torque sensor, such as a strain gauge torque sensor, collects high-frequency data on the riveting torque (resistance torque during thread engagement) and corresponding time data during the screwing process, such as 100 times per second. The torque data directly reflects the thread fit: if the nut and rivet rod thread specifications match and there are no obstructions, the torque will steadily increase with the screwing depth, such as an increase of 5-8 N·m for each turn; if there is thread misalignment, such as a misaligned nut hole, the torque will suddenly increase, exceeding the standard value by more than 20% instantaneously; if the nut has cracks or worn threads, the torque may suddenly drop, falling below the standard value by more than 15%.

[0136] The system integrates the collected torque-time data to generate a torque variation graph, with time on the horizontal axis and torque value on the vertical axis, visually presenting the dynamic trend of torque during the thread engagement process. Ideally, it should be a smooth upward curve without obvious peaks or drops. Simultaneously, the average velocity of torque change is calculated by dividing the total torque change by the total thread engagement time, reflecting the uniformity of torque increase. If the average velocity fluctuates by more than ±10%, it indicates an abnormality in the thread engagement.

[0137] To scientifically determine the spin-in state, the system generates torque process values ​​through multi-step calculations:

[0138] Step 1: Calculate the similarity value of the torsion diagram:

[0139] A preset reference torque map is invoked, generated based on a qualified thread engagement scenario. Its characteristics include: torque smoothly increases from an initial value of 0 to a preset peak value, the curve shows no significant fluctuations, and the slope changes uniformly. A curve similarity algorithm, such as the correlation coefficient method, is used to calculate the similarity value between the actual torque change map and the reference map. The torque map similarity value ranges from 0 to 1. The closer the similarity value is to 1, the smoother the thread engagement, such as without jamming or misalignment; a similarity value below 0.7, i.e., 0.7 is the preset threshold, indicates a significant anomaly.

[0140] Step 2: Generate torque map comparison values:

[0141] Compare the torque map similarity value with a preset reference similarity value (e.g., 0.85) to calculate the torque map comparison value: (actual similarity value - reference similarity value) / reference similarity value. A positive value indicates that the engagement condition is better than the standard, while a negative value indicates that there is a deviation. The larger the absolute value of the negative value, the more serious the deviation.

[0142] Step 3: Calculate the torque-speed comparison value:

[0143] Compare the actual average speed of torque change with the reference average speed (e.g., 2 N·m / s) to generate a torque-speed comparison value: (actual average speed - reference average speed) / reference average speed. This value reflects the compliance of the torque increase rhythm. If it exceeds ±0.15, it indicates that the speed is too fast or too slow.

[0144] Step 4: Integrate torque process values:

[0145] The torque process value is obtained by integrating the weighted algorithm (e.g., the torque comparison value accounts for 70% and the speed comparison value accounts for 30%), and the overall quality of thread engagement is comprehensively evaluated.

[0146] Based on the determination of the torque process value, the system optimizes the screw-in speed according to different scenarios:

[0147] Scenario 1: Torque value is within acceptable range (e.g., -0.1 to 0.2):

[0148] Adjust the screwing speed according to the positive correlation adjustment logic: The larger the torque process value, such as 0.2, the smoother the thread engagement, the higher the similarity value and the more stable the speed. The speed of the rivet gun can be increased from the initial 50r / min to 70r / min to speed up the feeding efficiency. If the process value is small, such as -0.05, it means that the engagement state is close to the lower limit of the qualification. If the torque fluctuation is slightly large, the speed needs to be reduced to 40r / min to avoid aggravating thread wear due to excessive speed, especially for nuts made of soft materials such as aluminum alloys.

[0149] Scenario 2: Torque value exceeds acceptable range:

[0150] Immediately trigger torque anomaly warning: alert the operator with an audible and visual alarm and display the anomaly type on the central control system, such as sudden torque increase: possible thread misalignment; sudden torque decrease: possible nut defect; at the same time, suspend the screwing action to prevent forced screwing from causing the rivet thread to break or the nut to crack, thereby reducing equipment damage and parts waste.

[0151] Compared to screw-in control that relies solely on stroke judgment, it can identify potential problems such as thread misalignment and foreign object jamming in advance through torque characteristics, avoiding damage to parts caused by blind screwing in; the adaptive adjustment of screwing speed maximizes feeding efficiency while ensuring thread meshing accuracy, such as speeding up in smooth scenarios and slowing down in risky scenarios; the abnormal early warning mechanism enables early detection and handling of problems, reducing the subsequent riveting rework costs caused by unqualified feeding.

[0152] The method also includes the following steps during the process of screwing the rivet into the rivet nut:

[0153] When the rivet rod begins to screw into the rivet nut, the torque sensor collects real-time riveting torque (e.g., once every 0.01 seconds) and corresponding time data at high frequency, and simultaneously records the displacement of the rivet rod as it screws in. This data is then acquired by the robotic arm encoder to generate a riveting torque-displacement diagram. With the displacement as the horizontal axis and the torque value as the vertical axis, the diagram visually presents the torque changes at different screwing depths. For example, during normal engagement, the torque rises steadily with the displacement without drastic fluctuations.

[0154] To uncover the microscopic features in the data, the system introduces a spectrum transformation algorithm, such as the Fast Fourier Transform (FFT), to convert the torque-displacement map from a displacement-torque time-domain signal to a frequency-energy frequency-domain signal, generating a riveting torque spectrum. In this spectrum, the horizontal axis represents frequency, reflecting the periodicity of torque fluctuations; for example, high-frequency components correspond to rapid, minute pauses, while low-frequency components correspond to slow torque changes. The vertical axis represents energy, reflecting the intensity of the corresponding frequency fluctuation; higher energy indicates more significant torque fluctuations at that frequency. The spectrum can identify microscopic anomalies masked by traditional time-domain analysis. For instance, in the spectrum of normal thread engagement, low-frequency energy is concentrated and stable, while high-frequency energy is weak; if thread burrs are present, there will be a sudden increase in high-frequency energy in a specific frequency band.

[0155] The system extracts two core parameters from the riveting torque spectrum and quantifies the microscopic stability of the screw-in process by comparing them with reference values:

[0156] Torque frequency variation amplitude: Calculate the distribution range and dispersion of the main frequency components (such as 0-5Hz) in the spectrum to reflect the periodicity of torque fluctuation. If the frequency variation amplitude is small, it indicates that the torque fluctuation rhythm is stable and there are no sudden interruptions.

[0157] Torque energy variation amplitude: The coefficient of variation of energy values ​​in each frequency band is statistically analyzed to reflect the uniformity of torque transmission. If the energy variation amplitude is small, it indicates that the force is stable during thread meshing and there is no local stress concentration.

[0158] The system compares the two amplitude values ​​mentioned above with preset torque reference frequency change values ​​and torque reference energy change values, respectively. The torque reference energy change value is calibrated based on the spectral characteristics of a large number of qualified engagement samples, and calculates the torque frequency comparison value (the ratio of the actual frequency amplitude to the reference value) and the torque energy comparison value (the ratio of the actual energy amplitude to the reference value). Through a weighted summation algorithm, such as the frequency comparison value accounting for 40% and the energy comparison value accounting for 60%, the torque spectral change amplitude value is finally obtained. The smaller this value, the closer the micro-dynamics of the screwing process is to the ideal state, such as no minor jamming, uniform stress, and better thread engagement quality; conversely, it indicates the existence of micro-anomalies, such as local wear of the thread or slight wobble of the rivet.

[0159] The system dynamically adjusts the input parameters based on the comparison between the torque spectrum variation amplitude and the preset reference variation amplitude.

[0160] If the amplitude of the spectrum change is less than the reference value, it indicates that the microscopic state of the screw-in is qualified. Then, the speed is adjusted according to the negative correlation adjustment logic: the larger the amplitude of the spectrum change, the closer it is to the reference value, indicating that there are slight microscopic fluctuations in the screw-in process, such as slightly higher high-frequency energy, which may be due to slight unevenness of the thread surface. At this time, the screw-in rotation speed of the rivet and the movement speed of the robotic arm are slowed down, such as reducing the rotation speed from 60r / min to 45r / min and the movement speed from 10mm / s to 8mm / s, to avoid aggravating thread wear due to excessive speed. The smaller the amplitude of the spectrum change, the lower it is than the reference value, indicating that the screw-in process is extremely stable, such as concentrated low-frequency energy and no high-frequency interference. The speed can be increased, such as increasing the rotation speed to 75r / min and the movement speed to 12mm / s, to improve the feeding efficiency while ensuring the quality of the thread.

[0161] If the amplitude of the spectrum change exceeds the reference value, it indicates a significant microscopic anomaly. The system immediately triggers a riveting anomaly warning: it alerts the operator through an audible and visual alarm and marks the abnormal characteristics in the central control system, such as a sudden increase in high-frequency energy (possibly thread burrs) or disordered energy distribution (possibly nut hole misalignment). At the same time, the screwing action is suspended to prevent further screwing from causing thread breakage (e.g., a steel nut with excessive hardness may crack due to jamming) or rivet deformation, thus intercepting potential quality problems at the microscopic level.

[0162] The above-mentioned solution can capture subtle problems that are difficult to detect in traditional torque diagrams, such as intermittent jamming caused by local thread concavity and stress fluctuations caused by uneven material hardness, thus providing early warning of potential thread damage risks. Based on speed adaptation for microscopic stability, it avoids excessive acceleration in macroscopically acceptable but microscopically fragile scenarios. For example, although the macroscopic torque of an aluminum nut is normal, high-frequency fluctuations indicate that its threads are prone to wear and require deceleration protection. It is especially suitable for parts such as automotive engine compartments and chassis where stringent requirements for threaded connection strength are placed. Microscopic-level control ensures engagement quality and lays the foundation for the structural reliability of subsequent riveting.

[0163] During the process of screwing the rivet into the rivet nut, the system has added a graded retry and shutdown mechanism based on torque spectrum anomalies. By distinguishing between occasional interference and persistent faults, it reduces production interruptions caused by misjudgments while ensuring quality. The specific method includes the following steps:

[0164] When the system calculates, through torque spectrum analysis, that the amplitude of the torque spectrum change (a core indicator reflecting the microscopic stability of thread engagement) exceeds the preset torque reference amplitude, it immediately determines it as the first engagement anomaly. At this point, the system does not simply shut down, but instead executes an exit-retry operation:

[0165] Orderly withdrawal control: Control the robotic arm to slowly rotate the rivet rod in the opposite direction of the screw insertion. If the rotation speed is reduced to 20 r / min, at the same time move in the opposite direction at a low speed (such as 5 mm / s) to ensure that the rivet rod is smoothly withdrawn from the rivet nut. This avoids aggravating thread damage due to excessively fast movement during withdrawal, especially for thread structures that have already shown slight abnormalities.

[0166] Secondary screw-in execution: After exiting, the system re-calls the loading station coordinates, drives the robotic arm to move back to the alignment position, and performs the rivet screw-in operation according to the initial parameters (or parameters finely adjusted based on the initial abnormal characteristics, such as reducing the initial speed by 10%). The core purpose of secondary screw-in is to eliminate accidental factors; such as the presence of small foreign objects (such as metal fragments) between the threads during the first screw-in, or alignment deviations caused by instantaneous vibration of the robotic arm. These factors may be cleared by the exit action, or eliminated due to the improved alignment accuracy during secondary screw-in.

[0167] During the second screwing process, the system simultaneously initiates torque spectrum monitoring and recalculates the torque spectrum change amplitude value:

[0168] If the amplitude value of the second screw-in returns to the qualified range: the first abnormality is determined to be an accidental interference, the system will execute the subsequent riveting process normally, and automatically record data such as abnormal time and torque spectrum characteristics (to optimize the subsequent abnormal judgment threshold), without affecting the continuity of production.

[0169] If the amplitude of the second screw-in still exceeds the reference range: the system immediately determines it as a persistent fault, such as damage to the rivet thread, misalignment of the rivet nut hole, or mismatch between the two specifications, triggering dual emergency measures:

[0170] 1. Multi-level alarm prompts: Through audible and visual alarms (flashing red warning light + high-frequency buzzer), pop-up windows in the central control system, and push notifications of abnormal information from the production management platform, operators are clearly informed that the screwing-in abnormality cannot be repaired. A spectrum comparison chart of the first and second screwing-in is also provided to help quickly locate the problem. If both spectra show a sudden increase in a specific high-frequency energy, it may indicate a burr in the rivet thread.

[0171] 2. Emergency Stop Protection: Automatically controls the robotic arm to stop all movements, maintains its current position, avoids secondary collisions caused by movement, freezes the rivet gun drive system, and prevents more serious consequences due to continuous screwing; such as thread breakage causing metal debris to contaminate the workstation, bending of the rivet rod damaging the robotic arm transmission mechanism, and nut breakage causing subsequent assembly dimensional deviations.

[0172] Using the above methods, for single anomalies caused by accidental factors (such as momentary foreign objects or minor vibrations), a second screw-in can avoid direct shutdown, reducing production interruptions caused by misjudgment. This is especially beneficial in batch production scenarios, reducing downtime by several minutes to tens of minutes. For persistent faults (such as component defects or equipment malfunctions), decisive shutdown is crucial to prevent cascading damage caused by operating with defects, such as escalating from thread wear to robotic arm failure, thus reducing equipment maintenance costs and component scrap rates. The comparison of the spectral data from the two screw-ins provides a key basis for fault diagnosis, making it easier to distinguish problem types compared to single anomalies. For example, accidental interference usually manifests as significant differences between the two spectral analyses, while component defects show similar spectral anomaly characteristics, improving the accuracy of anomaly handling.

[0173] During the secondary screwing in of the rivet nut, the system adds a damage identification sub-step based on the comparison of the torque spectrum of the two screwing operations. By quantitatively analyzing the degree of variation in the spectrum characteristics, it accurately determines whether the rivet rod or rivet nut is damaged due to the first screwing in, avoiding the chain risk caused by the continued participation of damaged parts in the operation. The specific steps include the following:

[0174] After the second screw-in operation is initiated, the system synchronously starts the torque sensor and data acquisition module, recording the riveting torque and corresponding time and displacement data at the same sampling frequency as the first screw-in (e.g., once every 0.01 seconds), generating a riveting torque spectrum for the second screw-in. The time-domain data is then converted into a frequency-domain signal using a Fast Fourier Transform. This spectrum maintains the same frequency range division as the spectrum for the first screw-in, such as 0-10Hz, with each 0.5Hz band ensuring accurate comparison.

[0175] To achieve refined difference analysis, the system converts both spectrum graphs into spectrum bar charts; with frequency bands as the horizontal axis and the energy value (in dB) of each frequency band as the vertical axis, the energy distribution characteristics of different frequency bands are visualized. For example, the energy of a certain frequency band is 50dB when it is first screwed in, and the energy of the same frequency band is 65dB when it is screwed in a second time, which intuitively shows the energy increase.

[0176] The system converts the differences between two spectral histograms into quantifiable differences in frequency energy variation through two steps:

[0177] Step 1: Calculate the energy difference in the corresponding frequency bands:

[0178] For the overlapping frequency bands (e.g., 0-8Hz) in two spectral histograms, calculate the difference in energy values ​​for each band, i.e., the energy value of the second spin-in minus the energy value of the first spin-in, and take the absolute value. The larger the difference, the more significant the energy fluctuation in that frequency band. For example, if the energy in the 2-2.5Hz band is 45dB during the first spin-in and 60dB during the second spin-in, the corresponding difference is 15dB, reflecting that the torque fluctuation in that frequency band is significantly amplified during the second spin-in.

[0179] Step 2: Weighted accumulation of energy in non-overlapping frequency bands:

[0180] For non-overlapping frequency bands in two spectrum diagrams, if the first screw-in exhibits high-frequency energy of 8-10Hz but the second screw-in lacks this energy, or vice versa, the system calculates the accumulated value according to a two-dimensional coordinate weighting rule: using the frequency value (Hz) of the frequency band as the X-axis and the energy value (dB) as the Y-axis, the weighted energy of each non-overlapping frequency band is calculated using a preset weighting formula, such as weighted energy = 0.3 × frequency value + 0.7 × energy value. The weighted energies of all non-overlapping frequency bands are then accumulated, where the weighting parameters are all normalized. Energy changes in high-frequency bands (such as 8-10Hz) are often related to momentary thread jamming; higher energy values ​​indicate more severe jamming, thus requiring higher weighting to highlight potential serious damage, such as thread breakage.

[0181] Final gap value integration: The sum of the energy differences in the corresponding frequency bands is added to the weighted sum of the non-overlapping frequency bands to obtain the frequency energy change gap; the larger this value is, the more significant the difference in the torque dynamic characteristics of the two screwing in is, and the higher the possibility of component damage.

[0182] The system compares the calculated frequency energy change difference with a preset difference threshold, where the difference threshold is calibrated based on the spectral difference characteristics of a large number of component damage samples, such as setting the threshold to 100dB·Hz.

[0183] If the difference is less than the threshold: the difference in the spectrum between the two spin-ins is determined to be within the normal range, which may be due to slight fluctuations caused by accidental factors. If the amplitude of the spectrum change returns to the qualified value after the second spin-in, the system continues to execute the subsequent process; if it is still abnormal, it is treated as a continuous fault, such as alarm shutdown.

[0184] If the difference value exceeds the threshold: Immediately trigger a warning of damage to the rivet rod or rivet nut, specifically manifested as follows:

[0185] 1. Level prompt content: The central control system displays: Spectrum difference exceeds the limit: There may be thread wear / deformation, and the frequency band with the most significant difference is marked. For example, if the energy of the 8-10Hz frequency band suddenly increases by 30dB, it is suspected that the rivet thread has broken teeth, which helps the operator to locate the damaged part.

[0186] 2. Interlocking protection measures: While issuing a prompt, the system automatically pauses all operations after the second screw-in, such as not entering the riveting stage again, to prevent damaged parts (such as burr-bearing rivets) from continuing to work, which could lead to secondary failures such as scratches on the riveting plate or jamming of the robotic arm transmission mechanism.

[0187] 3. Data archiving and traceability: Automatically stores the original torque data, spectrum diagram and difference calculation process of the two screwing in, providing a basis for subsequent quality analysis, such as optimizing the difference threshold of different nut models through long-term data statistics.

[0188] Using the above method, microscopic damage such as thread wear and deformation can be detected during the initial screwing and subsequent screwing. This damage may not be visible to the naked eye, but it is manifested through spectral characteristic anomalies, preventing damaged parts from entering subsequent stages and causing batch quality problems. Frequency band difference analysis can preliminarily determine the type of damage; for example, a sudden increase in high-frequency energy indicates localized tooth breakage caused by momentary jamming, while a shift in low-frequency energy indicates overall thread deformation, reducing troubleshooting time. Suspected components can be identified without disassembling the entire equipment. Compared to the passive mode of shutting down after a significant fault occurs, this mechanism can intervene promptly in the early stages of damage, reducing equipment maintenance costs (such as preventing damaged rivet rods from wearing down the robotic arm bearings) and parts scrap rates (such as timely replacement of damaged nuts to prevent collateral damage to the sheet metal).

[0189] This application also discloses a riveting state control system for a riveting device, including a processor, wherein the processor executes the steps of the riveting state control method for the riveting device as described in any of the above embodiments.

[0190] This application also discloses a storage medium storing a program, which, when executed by a processor, implements the steps of the riveting state control method for the riveting equipment described in any of the above embodiments.

[0191] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for controlling the riveting state of a riveting device, characterized in that, Includes the following steps: Based on the arrival command of the riveting nut at the loading station, the robotic arm is controlled to move to the corresponding loading station. The robotic arm moves towards the loading station and simultaneously rotates the riveting gun on the robotic arm along the preset screwing direction. After reaching the set stroke, the end of the rivet rod is screwed into the riveting nut. During the screwing of the rivet rod into the riveting nut, the screwing pressure value on the rivet rod is acquired in real time based on the pressure sensor, and the pressure fluctuation value of the screwing pressure value is calculated. A pressure comparison value is calculated based on the screwing pressure value and the preset reference pressure value. A fluctuation comparison value is calculated based on the pressure fluctuation value and the preset reference fluctuation value. The screwing process value is calculated based on the pressure comparison value and the fluctuation comparison value. If the screwing process value is outside the preset screwing value range, a loading abnormality prompt is issued. Control the robotic arm to move in the opposite direction from the loading station, then move to the position corresponding to the riveting station, control the robotic arm to move toward the riveting station to align with the riveting plate, drive the riveting gun to first rotate along the preset first direction at a preset first angle, then rotate along the preset second direction at a preset second angle, and repeat the rotation for a preset number of times. The number of times is adjusted according to the positive correlation of the pressure comparison value. The larger the pressure comparison value, the more times the number of times is set, and the smaller the pressure comparison value, the fewer times the number of times is set. The angle difference between the first angle and the second angle is adjusted according to the positive correlation of the pressure fluctuation value. The larger the pressure fluctuation value, the larger the angle difference, and the smaller the pressure fluctuation value, the smaller the angle difference. The rivet rod is controlled to retract the rivet gun with a preset riveting pull force. The riveting pull force is adjusted according to the screw-in process value. The larger the screw-in process value, the larger the riveting pull force, and the smaller the screw-in process value, the smaller the riveting pull force. This makes the rivet nut form and press the plate tightly. The rivet gun is controlled to rotate in the opposite direction of screwing in, and the robotic arm is controlled to move in the opposite direction of the riveting station, so that the rivet rod is withdrawn from the rivet nut.

2. The riveting state control method for riveting equipment according to claim 1, characterized in that, The step of driving the riveting gun to rotate firstly along a preset first direction by a preset first angle, and then along a preset second direction by a preset second angle, repeating this rotation a preset number of times, also includes the following sub-steps: The rotational pressure value of the rivet rod is obtained during the process of the rivet gun rotating at a first angle along a preset first direction, and the rotational pressure curve is obtained based on the rotational pressure value; Calculate the shape similarity value between the rotational pressure curve and the preset reference pressure curve; If the shape similarity value is less than the preset reference similarity value, the size of the first angle is adjusted according to the negative correlation of the shape similarity value. The larger the shape similarity value, the smaller the first angle, and the smaller the shape similarity value, the larger the first angle. Otherwise, calculate the trend of change in shape similarity values; If the trend of change is increasing, the control will drive the rivet gun to first rotate along the preset second direction by a preset first angle, and then rotate along the preset first direction by a preset second angle, repeating the rotation a set number of times. Otherwise, the size of the second angle is adjusted according to the positive correlation between the shape similarity value and the shape similarity value. The larger the shape similarity value, the larger the second angle, and the smaller the shape similarity value, the smaller the second angle.

3. The riveting state control method for riveting equipment according to claim 2, characterized in that, The step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps: The displacement of the rivet rod and the corresponding time data are obtained during the riveting process based on the displacement sensor. The riveting displacement diagram and average riveting speed are calculated based on the riveting displacement and time data. The similarity value of the displacement diagram is calculated based on the riveting displacement diagram and the preset reference displacement diagram; The displacement map comparison value is calculated based on the displacement map similarity value and the preset reference similarity value; The riveting speed comparison value is calculated based on the average riveting speed and the preset reference average speed; The riveting process values ​​are calculated based on the comparison values ​​of the displacement diagram and the riveting speed. If the riveting process value is within the preset riveting reference range, the rotation speed of the riveting gun in the opposite direction of the screwing direction and the movement speed of the robotic arm in the opposite direction of the riveting station are adjusted according to the positive correlation of the riveting process value. The larger the riveting process value, the faster the rotation speed and movement speed, and the smaller the riveting process value, the slower the rotation speed and movement speed. Otherwise, issue an abnormal riveting warning.

4. The riveting state control method for riveting equipment according to claim 3, characterized in that, The step of controlling the rivet rod to retract the rivet gun with a preset riveting tension also includes the following steps: The displacement of the rivet rod and the corresponding time data are obtained during the riveting process based on the displacement sensor. A riveting displacement diagram is calculated based on the riveting displacement and time data; a riveting spectrum diagram is then calculated based on the riveting displacement diagram using a spectrum conversion algorithm. The riveting frequency variation amplitude and riveting energy variation amplitude values ​​were calculated based on the riveting spectrum diagram. The riveting frequency comparison value is calculated based on the riveting frequency change amplitude value and the preset riveting reference frequency change value. The riveting energy comparison value is calculated based on the riveting energy change amplitude value and the preset riveting reference energy change value; The amplitude of the riveting spectrum variation is calculated based on the comparison values ​​of riveting frequency and riveting energy. If the riveting spectrum change amplitude is less than the preset riveting reference change amplitude, the rotation speed of the rivet gun in the opposite direction of the screwing direction and the movement speed of the robotic arm in the opposite direction of the riveting station are adjusted according to the negative correlation of the riveting spectrum change amplitude. The larger the riveting spectrum change amplitude, the slower the rotation speed and movement speed. The smaller the riveting spectrum change amplitude, the faster the rotation speed and movement speed. Otherwise, issue an abnormal riveting warning.

5. The riveting state control method for riveting equipment according to claim 1, characterized in that, The process of screwing the rivet into the rivet nut also includes the following steps: Data on riveting torque and time during the process of screwing the rivet into the rivet nut are obtained based on a torque sensor. The torque variation diagram and the average speed of torque variation were calculated based on the riveting torque and time data. The torque diagram similarity value is calculated based on the torque variation diagram and the preset reference torque diagram; The torsion diagram comparison value is calculated based on the torsion diagram similarity value and the preset reference similarity value; The torque-speed comparison value is calculated based on the average speed of torque change and the preset reference average speed. The torque process value is calculated based on the torque diagram comparison value and the torque change comparison value. If the torque process value is within the preset torque reference range, the rotation speed of the rivet rod screwing into the rivet nut is adjusted according to the positive correlation of the torque process value. The larger the torque process value, the faster the rotation speed, and the smaller the torque process value, the slower the rotation speed. Otherwise, issue a torque abnormality warning.

6. The riveting state control method for riveting equipment according to claim 1, characterized in that, The process of screwing the rivet into the rivet nut also includes the following steps: Data on riveting torque and time during the process of screwing the rivet into the rivet nut are obtained based on a torque sensor. The riveting torque-displacement diagram is calculated based on the riveting torque and time data; the riveting torque spectrum diagram is then calculated using a spectrum conversion algorithm based on the riveting torque-displacement diagram. The torque frequency variation amplitude and torque energy variation amplitude are calculated based on the riveting torque spectrum diagram; The torque frequency comparison value is calculated based on the torque frequency change amplitude value and the preset torque reference frequency change value. The torque energy comparison value is calculated based on the torque energy change amplitude value and the preset torque reference energy change value. The torque spectrum variation amplitude of the riveting torque spectrum diagram is calculated based on the torque frequency comparison value and the torque energy comparison value. If the torque spectrum change amplitude is less than the preset torque reference change amplitude, the rotation speed of the rivet screw into the rivet nut is adjusted according to the negative correlation of the torque spectrum change amplitude. The larger the torque spectrum change amplitude, the slower the rotation speed and the moving speed; the smaller the torque spectrum change amplitude, the faster the rotation speed and the moving speed. Otherwise, issue an abnormal riveting warning.

7. The riveting state control method for riveting equipment according to claim 1, characterized in that, The method also includes the following steps: If, during the process of screwing the rivet rod into the rivet nut, the torque spectrum change amplitude value calculated based on the riveting torque data obtained by the torque sensor exceeds the preset torque reference change amplitude value, it is recorded as the first screwing abnormality. Then, the robotic arm is controlled to drive the rivet rod out of the rivet nut and the operation of screwing the rivet rod into the rivet nut is re-executed for a second screwing. If the torque spectrum change amplitude value calculated during the second screwing process still exceeds the preset torque reference change amplitude value, an alarm will be triggered and the riveting operation will be stopped.

8. The riveting state control method for riveting equipment according to claim 7, characterized in that, The method also includes the following steps: During the process of screwing in the rivet nut for the second time, the riveting torque spectrum diagram of the second screwing in is obtained, and the frequency energy change difference between the riveting torque spectrum diagram of the first screwing in and the riveting torque spectrum diagram of the second screwing in is calculated. The difference value is obtained by subtracting the energy values ​​of corresponding frequency bands from two spectrum histograms and then weighting and summing the energy values ​​of non-overlapping frequency bands according to their two-dimensional coordinates, which include frequency and energy values. If the difference in energy change at this frequency exceeds a preset difference threshold, a warning will be issued indicating damage to the rivet or rivet nut.

9. A riveting status control system for a riveting device, characterized in that, The device includes a processor that performs the steps of the riveting state control method for a riveting device as described in any one of claims 1-8.

10. A storage medium, characterized in that, The storage medium stores a program that, when executed by a processor, implements the steps of the riveting state control method for the riveting device according to any one of claims 1-8.

Images