A method and system for monitoring and early warning of the wear state of a piercing die

Abstract

The present application relates to the technical field of punching die monitoring, and discloses a punching die wear state monitoring and early warning method and system, which comprises the following steps: collecting multi-stage monitoring data of the punching die in a continuous punching cycle; performing time sequence correlation analysis on the multi-stage monitoring data to extract working condition deviation features, die vibration energy response features and blade edge acoustic emission high-frequency activity features; determining whether there is working condition interference in the current punching cycle according to the working condition deviation features; when it is determined that there is no working condition interference, determining whether the punching die has structural stiffness degradation according to the die vibration energy response features; when it is determined that there is structural stiffness degradation, determining whether the blade edge region of the punching die has wear according to the blade edge acoustic emission high-frequency activity features; generating a wear warning level of the punching die based on the determination results, and performing a warning control operation according to the wear warning level. The present application realizes fine identification of the blade edge wear state and improves the wear diagnosis reliability.

Description

Technical Field

[0001] This invention relates to the field of punching die monitoring technology, and more specifically, to a method and system for monitoring and early warning of wear status of punching dies. Background Technology

[0002] Punching dies are crucial process equipment used in stamping to punch, trim, and form sheet metal. Their working condition directly affects the dimensional accuracy of stamped parts, hole wall quality, and overall machine production efficiency. Under continuous stamping conditions, punching dies are subjected to cyclic impact loads, high-frequency friction, and localized stress concentration over extended periods, easily leading to wear failure phenomena such as blunting of the cutting edge, microcrack propagation, localized chipping, surface depressions, and structural loosening. Accelerated die wear not only results in increased burrs, larger hole diameter deviations, and decreased product consistency in stamped parts, but can also cause die damage, abnormal equipment downtime, and even batch scrapping, thereby increasing manufacturing costs and impacting production cycle time.

[0003] Existing patent CN119549547A discloses a wear monitoring system and method for stamping dies in new energy heavy-duty trucks. It proposes using multiple detection methods, including acoustic emission sensors, ultrasonic sensors, spectral confocal sensors, temperature sensors, and pressure sensors, to monitor the operating status of the stamping die in real time, locating the wear position and determining the degree of wear. However, in actual stamping processes, changes in material thickness, feeding conditions, stamping speed, and stamping pressure can all cause changes in the monitoring signals. This results in the monitoring signals simultaneously containing information about changes in operating conditions and die wear information, making it difficult to effectively distinguish between signal anomalies caused by changes in operating conditions and signal changes caused by die wear. This leads to low accuracy in judging the die wear state and a high likelihood of misjudgments or omissions.

[0004] Therefore, it is necessary to design a method and system for monitoring and early warning of wear status of punching dies to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a method and system for monitoring and early warning of wear status of punching dies, aiming to solve the problem that changes in working conditions during the stamping process can also cause changes in the monitoring signal, making it difficult to effectively distinguish between signal abnormalities caused by changes in working conditions and signal changes caused by die wear. This results in low accuracy in judging the wear status of dies and is prone to misjudgment or omission.

[0006] In one aspect, the present invention proposes a method for monitoring and early warning of wear condition of punching dies, comprising: Collect multi-stage monitoring data of the punching die during a continuous stamping cycle. The multi-stage monitoring data includes the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. Time-series correlation analysis was performed on the multi-stage monitoring data to extract working condition deviation characteristics, mold vibration energy response characteristics, and cutting edge acoustic emission high-frequency activity characteristics. Based on the working condition deviation characteristics, determine whether there is working condition interference in the current stamping cycle; when it is determined that there is no working condition interference, determine whether the structural stiffness of the punching die has degraded based on the mold vibration energy response characteristics; when it is determined that the structural stiffness has degraded, determine whether the cutting edge area of ​​the punching die has worn based on the cutting edge acoustic emission high frequency activity characteristics. Based on the judgment result, a wear warning level for the punching die is generated, and a warning control operation is performed according to the wear warning level.

[0007] Furthermore, the collection of the multi-stage monitoring data includes: Acquire the slider displacement signal and / or crankshaft angle signal of the stamping equipment; divide the continuous stamping cycle into an unloaded downward stage, a blanking penetration stage, and a material ejection and reset stage based on the slider displacement signal and / or crankshaft angle signal; in the blanking penetration stage and the material ejection and reset stage, synchronously trigger the vibration sensor and acoustic emission sensor installed on the force-bearing part of the punching die to collect the vibration signal and acoustic emission signal.

[0008] Furthermore, when extracting the operating condition offset features, the following are included: Extract the actual stamping tonnage curve and the actual slide speed curve within the current stamping cycle as the real-time operating parameters; compare the actual stamping tonnage curve with the reference tonnage curve under the initial healthy state to obtain the tonnage offset; calculate the deviation between the actual slide speed curve and the reference speed curve under the initial healthy state to obtain the speed offset. The operating condition offset features include the tonnage offset and the speed offset.

[0009] Furthermore, when determining whether there is operational interference in the current stamping cycle based on the aforementioned operational deviation characteristics, the following steps are included: When the tonnage offset is greater than the tonnage fluctuation threshold or the speed offset is greater than the speed fluctuation threshold, it is determined that there is working condition interference in the current stamping cycle; when the tonnage offset is less than or equal to the tonnage fluctuation threshold and the speed offset is less than or equal to the speed fluctuation threshold, it is determined that there is no working condition interference.

[0010] Furthermore, when extracting the vibration energy response characteristics of the mold, the following are included: Based on the vibration signals of the punching penetration stage and the material removal and reset stage, the low-frequency structural resonance frequency band and the high-frequency impact frequency band are extracted; the first energy proportion in the low-frequency structural resonance frequency band and the second energy proportion in the high-frequency impact frequency band are determined; The vibration energy response characteristics of the mold include the first energy ratio and the second energy ratio.

[0011] Furthermore, when determining whether the punching die has experienced structural stiffness degradation based on the vibration energy response characteristics of the die, the following steps are included: Obtain a first energy baseline value and a second energy baseline value under the initial healthy state; calculate the first energy increase of the first energy ratio and the first energy baseline value, and the second energy increase of the second energy ratio and the second energy baseline value; When the first energy increase is greater than or equal to the second energy increase, and the first energy percentage is greater than the energy warning line, it is determined that the punching die has experienced structural stiffness degradation; when the first energy increase is less than the second energy increase or the first energy percentage is less than or equal to the energy warning line, it is determined that the punching die has not experienced structural stiffness degradation.

[0012] Furthermore, when extracting the high-frequency activity features of the blade's acoustic emission, the following are included: Based on the acoustic emission signal during the punching penetration stage, the ringing count rate and peak amplitude during the acoustic emission burst are extracted; the activity level of microcrack propagation on the cutting edge is determined based on the ringing count rate, and the severity of material spalling on the cutting edge is determined based on the peak amplitude. The high-frequency activity characteristics of the blade's acoustic emission include the degree of activity and the intensity.

[0013] Furthermore, when determining whether wear has occurred in the cutting edge area of ​​the punching die based on the high-frequency activity characteristics of the cutting edge acoustic emission, the following steps are included: The activity level is compared with the activity threshold, and the intensity level is compared with the intensity threshold; When the activity level is greater than the activity threshold and the intensity level is greater than the intensity threshold, it is determined that the cutting edge area has experienced severe wear. When the activity level is greater than the activity threshold and the intensity is less than or equal to the intensity threshold, or when the activity level is less than or equal to the activity threshold and the intensity is greater than the intensity threshold, it is determined that the cutting edge area has experienced slight wear. When the activity level is less than or equal to the activity threshold and the intensity level is less than or equal to the intensity threshold, the blade edge region is determined to be in a normal stage.

[0014] Furthermore, when generating the wear warning level of the punching die based on the judgment result, it includes: The wear warning levels include a first wear warning level and a second wear warning level. When the cutting edge area experiences severe wear, a first wear warning level is generated; when the cutting edge area experiences slight wear, a second wear warning level is generated.

[0015] Compared with existing technologies, the advantages of this invention are as follows: By collecting vibration signals, acoustic emission signals, and real-time operating parameters of the stamping equipment during continuous stamping cycles of the punching die, and performing time-series correlation analysis on the multi-source monitoring data, joint monitoring of stamping conditions, die structural status, and cutting edge wear status is achieved. By extracting operating condition deviation features and prioritizing the determination of whether there is operating condition interference in the current stamping cycle before wear diagnosis, it is possible to distinguish between signal changes caused by abnormal feeding, material fluctuations, lubrication changes, or equipment operating fluctuations and signal changes caused by die wear itself, avoiding misjudging operating condition changes as die wear, thereby reducing the false alarm rate and improving the reliability of online monitoring results. By analyzing the vibration signals during the punching penetration stage and the material removal and reset stage, energy response features corresponding to the low-frequency structural resonance frequency band and the high-frequency impact frequency band are extracted, and the first energy ratio and the second energy ratio are used to determine whether the die has undergone structural stiffness degradation. This allows for the detection of structural performance change trends before the die shows obvious failure, achieving early identification of abnormal die conditions. By extracting the ringing count rate and peak amplitude from the acoustic emission signal during the punching penetration stage, and characterizing the degree of microcrack propagation activity and the severity of material spalling at the cutting edge, the high-frequency acoustic emission activity characteristics generated during early micro-damage and continuous wear in the cutting edge region can be captured, thereby achieving refined identification of the cutting edge wear state and improving the sensitivity of wear diagnosis.

[0016] On the other hand, this application also provides a punching die wear condition monitoring and early warning system, used to apply the above-mentioned punching die wear condition monitoring and early warning method, including: The acquisition unit is configured to acquire multi-stage monitoring data of the punching die during a continuous stamping cycle. The multi-stage monitoring data includes the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. The analysis unit is configured to perform time-series correlation analysis on the multi-stage monitoring data and extract working condition deviation features, mold vibration energy response features, and cutting edge acoustic emission high-frequency activity features. The judgment unit is configured to determine whether there is working condition interference in the current stamping cycle based on the working condition offset characteristics; when it is determined that there is no working condition interference, it determines whether the structural stiffness of the punching die has degraded based on the mold vibration energy response characteristics; when it is determined that the structural stiffness has degraded, it determines whether the cutting edge area of ​​the punching die has worn based on the cutting edge acoustic emission high frequency activity characteristics. The early warning unit is configured to generate a wear warning level for the punching die based on the judgment result, and to perform early warning control operations according to the wear warning level.

[0017] It is understandable that the above-mentioned methods and systems for monitoring and early warning of wear conditions of punching dies have the same beneficial effects, and will not be elaborated further here. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of a method for monitoring and early warning of wear status of punching dies provided in an embodiment of the present invention; Figure 2 A logic flowchart of the punching die wear monitoring process provided in an embodiment of the present invention; Figure 3 This is a functional block diagram of the punching die wear condition monitoring and early warning system provided in an embodiment of the present invention. Detailed Implementation

[0019] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] In some embodiments of this application, see Figure 1-2 As shown, a method for monitoring and early warning of wear condition of punching dies is proposed, including: S100: Collects multi-stage monitoring data of punching dies during continuous stamping cycles. The multi-stage monitoring data includes vibration signals and acoustic emission signals of the punching dies, as well as real-time operating parameters of the stamping equipment. S200: Perform time-series correlation analysis on multi-stage monitoring data to extract working condition deviation characteristics, mold vibration energy response characteristics, and high-frequency activity characteristics of cutting edge acoustic emission; S300: Determine whether there is working condition interference in the current stamping cycle based on the working condition deviation characteristics; when it is determined that there is no working condition interference, determine whether the structural stiffness of the punching die has degraded based on the mold vibration energy response characteristics; when it is determined that the structural stiffness has degraded, determine whether the cutting edge area of ​​the punching die has worn based on the high frequency activity characteristics of the cutting edge acoustic emission. S400: Generates a wear warning level for the punching die based on the judgment result, and performs a warning control operation according to the wear warning level.

[0021] Specifically, the punching die wear monitoring and early warning method of the present invention is applicable to continuous stamping production scenarios, and is especially suitable for punching processes that require online identification and early warning of die wear. This method collects multi-stage monitoring data of the punching die during a continuous stamping cycle, and jointly analyzes changes in stamping conditions, die structural response, and cutting edge wear status, thereby achieving step-by-step judgment and early warning control of the die wear status.

[0022] In step S100, multi-stage monitoring data of the punching die during the continuous stamping cycle is collected. The multi-stage monitoring data includes at least the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. Preferably, the real-time operating parameters of the stamping equipment include the actual stamping tonnage curve, the actual slide speed curve, the slide displacement curve, the crankshaft angle curve, the feeding status signal, and the equipment control command signal.

[0023] When collecting multi-stage monitoring data, priority should be given to acquiring the slide displacement signal and / or crankshaft angle signal of the stamping equipment. For mechanical stamping equipment, the crankshaft angle signal is preferred as the basis for stage division; for servo presses, the slide displacement signal is preferred as the basis for stage division; when the equipment has both slide displacement signal and crankshaft angle signal, the two can be used together to improve the stability of stamping cycle identification. Specifically, based on the slide displacement signal and / or crankshaft angle signal, the continuous stamping cycle can be divided into an unloaded downward stage, a blanking penetration stage, and a material ejection and reset stage.

[0024] Specifically, the stage division is not simply based on equal time intervals, but rather on identification of different positions or action points during the stamping process. For example, when the slider is at the top dead center and begins to descend but has not yet contacted the sheet metal, it is determined to be the no-load descent stage; when the slider descends to contact the sheet metal and completes the stamping, it is determined to be the stamping penetration stage; when the slider completes the stamping and continues its return stroke, driving the unloading mechanism, it is determined to be the unloading reset stage. To make the stage division more accurate, data from several normal stamping cycles are collected during the equipment trial operation phase. The changes in crankshaft angle, slider displacement, sudden changes in tonnage, and sudden changes in acoustic emission are statistically calibrated to determine the boundary points of each stage.

[0025] In the punching penetration stage and the material ejection and reset stage, vibration sensors and acoustic emission sensors installed on the stress-bearing parts of the punching die are triggered simultaneously to collect vibration signals and acoustic emission signals. The reason for choosing these two stages for simultaneous acquisition is that the punching penetration stage can reflect the instantaneous impact, friction, and local crack initiation characteristics when the cutting edge enters the material, while the material ejection and reset stage can reflect the characteristics of die unloading, springback, material stripping, and local loosening. Therefore, it is more conducive to capturing abnormal responses caused by die wear.

[0026] The no-load descent phase is primarily used to establish baseline information and stage positioning information for the current stamping cycle. During this phase, real-time operating parameters such as slider displacement signals, crankshaft angle signals, actual slider speed curves, and actual stamping tonnage curves are preferably collected to determine the operating status and characteristics of the current stamping cycle. Since the punch has not yet contacted the sheet metal and the die has not undergone actual punching action during this phase, it is preferable not to use the vibration and acoustic emission signals generated during this phase as wear assessment criteria, but rather to record them as baseline data for equipment background noise and environmental vibration.

[0027] Furthermore, the real-time operating parameters collected during the unloaded descent phase are used to extract working condition offset features, calculate tonnage offset and speed offset, and determine whether there is working condition interference in the current stamping cycle. The vibration and acoustic emission signals collected during the blanking penetration phase are mainly used to extract the vibration energy response characteristics of the die and the high-frequency activity characteristics of the cutting edge acoustic emission. The vibration signals collected during the material removal and reset phase are mainly used to characterize the die unloading springback, material removal impact, and structural loosening state, to help determine whether the die has experienced structural stiffness degradation. Through the above methods, the data collected at different stages correspond to different analysis tasks, together constituting multi-stage monitoring data.

[0028] In steps S200 and S300, time-series correlation analysis is performed on the multi-stage monitoring data to extract working condition deviation characteristics, mold vibration energy response characteristics, and cutting edge acoustic emission high-frequency activity characteristics. Time-series correlation analysis refers to aligning vibration signals, acoustic emission signals, and real-time operating parameters collected within the same stamping cycle along a time axis, and then extracting data characteristics corresponding to different stamping stages to avoid signal aliasing from different stages interfering with the diagnostic results.

[0029] Specifically, when extracting operating condition offset features, the actual stamping tonnage curve and the actual slide speed curve within the current stamping cycle are first extracted as real-time operating parameters. Then, the actual stamping tonnage curve is compared with the reference tonnage curve under the initial healthy state to obtain the tonnage offset; the deviation between the actual slide speed curve and the reference speed curve under the initial healthy state is calculated to obtain the speed offset. Operating condition offset features include tonnage offset and speed offset.

[0030] Specifically, when comparing the actual stamping tonnage curve with the reference tonnage curve, the two curves are synchronized in time based on the crankshaft angle signal or the slider displacement signal, ensuring that the actual stamping tonnage curve corresponds to the reference tonnage curve at the same stamping position point. Then, the synchronized tonnage data is normalized to eliminate dimensional differences caused by different batches and sampling scales. The tonnage values ​​at corresponding sampling points are then subtracted to obtain a tonnage deviation sequence, and the absolute value of the tonnage deviation sequence is taken as the average value, which is used as the tonnage offset. Using the same method, the actual slider speed curve and the reference speed curve are synchronized in time and normalized, then subtracted point by point to obtain a speed deviation sequence, and the absolute value of the speed deviation sequence is taken as the average value, which is used as the speed offset. Through these methods, the tonnage offset and speed offset can accurately characterize the degree of deviation of the current stamping cycle from its healthy operating condition.

[0031] The baseline tonnage and speed curves under initial healthy conditions are preferably collected when the mold is first put into use and confirmed to be in normal condition. Specifically, after the mold is assembled and adjusted and confirmed by trial stamping, data from several normal stamping cycles can be collected continuously, and the tonnage and speed curves can be averaged to form the baseline curves. Alternatively, after removing outliers from multiple healthy cycle data, the median curve or weighted average curve can be used as the baseline curve. The purpose of this setting is to ensure that the baseline curves can reflect the actual operating behavior of the mold under healthy conditions as closely as possible.

[0032] The tonnage fluctuation threshold and speed fluctuation threshold are preferably set jointly using healthy sample data and trial production data. Specifically, tonnage and speed curves can be collected for multiple stamping cycles under normal mold conditions, their normal fluctuation range can be statistically analyzed, and a safety margin can be reserved to form the tonnage fluctuation threshold and speed fluctuation threshold. Preferably, the threshold can be determined by a combination of the mean and standard deviation of the healthy samples, for example, by adding a certain multiple of the standard deviation to the mean of the healthy samples as the upper limit of the threshold. This setting can cover the natural fluctuations under normal operating conditions and avoid misjudging instantaneous disturbances as abnormal operating conditions.

[0033] In a preferred embodiment, if the tonnage deviation of a batch of consecutive normal samples is within a small range, the tonnage fluctuation threshold can be set as the safe upper limit of the deviation level of that batch of samples; similarly, the slider speed deviation can be set according to the average deviation level and fluctuation amplitude of normal samples. In practical applications, when the tonnage deviation of the current stamping cycle is greater than the tonnage fluctuation threshold, or the speed deviation is greater than the speed fluctuation threshold, it is determined that there is working condition interference in the current stamping cycle; when neither exceeds the corresponding threshold, it is determined that there is no working condition interference.

[0034] In a preferred embodiment, when there is operational interference in the current stamping cycle, this cycle is marked as an operational interference cycle, and its corresponding tonnage curve, speed curve, vibration signal, and acoustic emission signal are stored separately as reference samples for subsequent operational condition identification and model correction. Subsequently, the next stamping cycle begins, and monitoring data is collected again, and the above judgment process is repeated. If operational interference occurs in multiple consecutive stamping cycles, preferably three stamping cycles, an abnormal operational condition prompt message is preferably output to remind the operator to check the feeding status, lubrication status, material status, or stamping equipment operating status, thereby avoiding the impact of operational condition fluctuations on the die wear judgment results.

[0035] Specifically, the extraction method for the vibration energy response characteristics of the die is as follows: based on the vibration signals during the punching penetration stage and the unloading and reset stage, the low-frequency structural resonance band and the high-frequency impact band are extracted, and the first energy proportion within the low-frequency structural resonance band and the second energy proportion within the high-frequency impact band are determined respectively. The low-frequency structural resonance band mainly reflects the overall structural stiffness, assembly state, and load-bearing response characteristics of the die, while the high-frequency impact band mainly reflects the local impact, friction, and transient damage response characteristics of the cutting edge.

[0036] The division of the low-frequency structural resonance band and the high-frequency impact band is primarily based on the characteristics of the mold structure and equipment test data. Specifically, vibration signals from several stamping cycles can be collected under healthy conditions, and frequency domain analysis can be performed to observe the frequency bands where the mold structure response is more concentrated and the frequency bands where the stamping impact is more significant. Then, combined with the natural frequency of the mold structure, the stamping frequency, and the on-site noise conditions, the range of the low-frequency structural resonance band and the high-frequency impact band can be determined.

[0037] Furthermore, to improve the consistency and repeatability of the low-frequency structural resonance band and high-frequency impact band division results, a combination of the main peak of the healthy sample spectrum and the main peak of the punching impact is preferred to determine the band boundaries. Specifically, when the mold is in its initial healthy state, vibration signals of no less than 50 punching cycles are continuously collected, and the vibration signals of each punching cycle are frequency-domain transformed to obtain the corresponding spectrum data. Subsequently, all spectrum data are statistically analyzed to determine the main peak of the structural response with the highest frequency and the largest energy contribution, and the frequency corresponding to this main peak is taken as the structural resonance center frequency. Taking the structural resonance center frequency as the center, the frequency boundaries corresponding to the cumulative energy reaching a preset proportion within its adjacent frequency range are statistically analyzed and used as the upper and lower limits of the low-frequency structural resonance band. In a preferred embodiment, the frequency range corresponding to the cumulative energy reaching 90% of the total energy of the main peak of the structural response can be determined as the low-frequency structural resonance band, thereby ensuring that most of the structural vibration energy is included within the band range. At the same time, the transient impact signal generated during the punching penetration stage in the healthy state is analyzed in the frequency domain to statistically analyze the concentrated distribution area of ​​the punching impact energy and determine the main peak frequency of the punching impact. Furthermore, the high-frequency impact band boundary is determined based on the peak frequency of the punching impact, according to the principle that the cumulative impact energy reaches a preset proportion. In a preferred embodiment, the frequency range corresponding to when the cumulative impact energy reaches 85% of the total energy of the peak impact can be determined as the high-frequency impact band. In practical applications, when the mold specifications, material type, or installation structure changes, health status sample data can be re-collected, and the low-frequency structural resonance band and high-frequency impact band can be re-determined in the above manner. This allows the band division results to adapt to the structural characteristics of different types of punching dies, improving the accuracy and applicability of subsequent structural stiffness degradation identification.

[0038] The calculation of the first and second energy proportions involves first performing a frequency domain transformation on the vibration signal, then integrating the signal energy within the corresponding frequency band, and finally dividing each by the total energy to obtain the proportion. Preferably, statistical processing is performed on vibration data from multiple consecutive stamping cycles within the same stage to reduce the impact of accidental impacts and random noise on the results.

[0039] Specifically, when determining whether the structural stiffness of the punching die has degraded, a first energy reference value and a second energy reference value are obtained under the initial healthy state. The first and second energy reference values ​​are preferably obtained through statistical analysis of healthy samples; that is, when the die is in a normal state, the first energy percentage and the second energy percentage of multiple stamping cycles are collected, and their average or median is calculated. Subsequently, the first energy increase between the first energy percentage of the current cycle and the first energy reference value, and the second energy increase between the second energy percentage of the current cycle and the second energy reference value are calculated.

[0040] Specifically, the energy warning line is also primarily determined based on healthy samples and trial production data. Specifically, the upper limit fluctuation range of the first energy percentage under normal mold conditions can be statistically analyzed, and a warning value slightly higher than the normal fluctuation range can be set based on this. The purpose of this energy warning line is to avoid misjudging normal structural vibration fluctuations under healthy conditions as structural stiffness degradation.

[0041] Specifically, the energy warning line is preferably determined based on the statistical results of healthy samples. Specifically, under the condition that the mold is in its initial healthy state and there is no structural stiffness degradation, a predetermined number of normal stamping cycle data are continuously collected, preferably no less than 50 stamping cycles. For each stamping cycle, a first energy percentage is extracted at the same stage, and after removing abnormal samples, the average value μ and standard deviation σ of the first energy percentage are calculated. Subsequently, the energy warning line is set as the sum of μ and a predetermined multiple of the standard deviation, preferably μ + 3σ, where the predetermined multiple is preferably an integer multiple between 2 and 4, and more preferably 3 times the standard deviation. Using the above method, the energy warning line corresponds to the upper limit of normal fluctuation of the first energy percentage under healthy conditions, thereby distinguishing normal structural vibration fluctuations from structural stiffness degradation characteristics and avoiding misjudgment.

[0042] In a preferred embodiment, if the increase in the first energy percentage relative to the reference value in the current cycle is greater than or equal to the increase in the second energy percentage relative to the reference value, and the first energy percentage is greater than the energy warning line, then it can be determined that the punching die has experienced structural stiffness degradation; if the increase in the first energy percentage is less than the increase in the second energy percentage, or the first energy percentage does not exceed the energy warning line, then it can be determined that the punching die has not experienced structural stiffness degradation. This judgment logic can effectively distinguish between changes in the overall structural response of the die and changes in local impact fluctuations.

[0043] In a preferred embodiment, if the current stamping cycle does not determine that the structural stiffness of the punching die has degraded, it is determined that no abnormality has occurred in the punching die. Similarly, its corresponding tonnage curve, speed curve, vibration signal and acoustic emission signal are stored separately as reference samples for subsequent working condition identification and model correction.

[0044] Specifically, when extracting the high-frequency activity characteristics of acoustic emission from the cutting edge, the ringing count rate and peak amplitude during the acoustic emission burst are extracted based on the acoustic emission signal during the punching penetration stage. The ringing count rate characterizes the activity level of microcrack propagation at the cutting edge, while the peak amplitude characterizes the intensity of material spalling or severe friction at the cutting edge. The high-frequency activity characteristics of acoustic emission from the cutting edge include both activity level and intensity.

[0045] Specifically, the acoustic emission signals acquired during the punching penetration stage are first filtered and preprocessed to remove the influence of environmental noise and equipment background noise. Then, a trigger threshold is determined based on the amplitude distribution of the acoustic emission background signal under healthy conditions. The trigger threshold is preferably set as the average background noise plus a preset safety margin. When the acoustic emission signal amplitude first exceeds the trigger threshold, an acoustic emission burst event is determined to have occurred, and the start time of the burst event is recorded. When the acoustic emission signal remains below the trigger threshold for a preset fall-off time, the burst event is determined to have ended, and the end time is recorded. After identifying the acoustic emission burst event, feature extraction is performed on the signal segment corresponding to the burst event. Specifically, the number of effective oscillations of the acoustic emission signal exceeding the trigger threshold within the signal segment is counted, and the number of effective oscillations is divided by the duration of the burst event to obtain the ringing count rate. At the same time, the maximum signal amplitude in the burst event is extracted as the peak amplitude. The ringing count rate is used to characterize the activity of microcrack propagation in the cutting edge region, and the peak amplitude is used to characterize the energy release intensity generated by material spalling, local chipping, or severe friction at the cutting edge.

[0046] Furthermore, to improve the stability of the acoustic emission emergency identification results, it is preferable to determine the trigger threshold based on healthy state sample data. Specifically, under the condition that the mold is in an initial healthy state and there is no cutting edge wear, acoustic emission signals are continuously collected for no less than 100 stamping cycles, and background noise data of the non-punching area of ​​each stamping cycle is extracted. Then, the average amplitude and fluctuation range of all background noise data are calculated, and the average amplitude is used as the background noise benchmark value. The preset safety margin is preferably determined based on the background noise fluctuation level. In a preferred embodiment, the standard deviation of the background noise amplitude of healthy samples can be statistically analyzed, and the preset safety margin can be set to 2 to 4 times the standard deviation; more preferably, the preset safety margin can be set to 3 times the standard deviation. In this case, the trigger threshold can be expressed as the sum of the average amplitude of the background noise and 3 times the standard deviation, thereby taking into account both the false alarm rate and the false negative rate. When identifying an acoustic emission emergency, when the acoustic emission signal amplitude first exceeds the trigger threshold, it is recorded as the start point of the emergency; thereafter, when the acoustic emission signal continuously falls below the trigger threshold and continues to reach the preset fall-off time, the emergency is determined to have ended. The preset fallback duration is preferably determined based on the average duration of acoustic emission bursts under healthy conditions.

[0047] In a preferred embodiment, the duration distribution of all acoustic emission bursts within at least 100 healthy stamping cycles can be statistically analyzed, and their average duration can be calculated. Then, 10% to 20% of the average duration can be determined as the preset fallback time. Preferably, 15% of the average duration can be taken as the preset fallback time to avoid a burst event being incorrectly divided into multiple independent events due to short-term signal fluctuations.

[0048] The active threshold and the severe threshold are preferably obtained through comparative statistics of healthy samples, lightly worn samples, and heavily worn samples. Specifically, during the experimental or pilot production phase, acoustic emission signals under different wear conditions can be collected, and the distribution range of their ringing count rate and peak amplitude can be statistically analyzed to determine the active threshold and the severe threshold.

[0049] In a preferred embodiment, when the activity level is greater than an activity threshold and the intensity level is greater than a severe threshold, the cutting edge area can be determined to have experienced severe wear; when the activity level is greater than the activity threshold and the intensity level is less than or equal to the severe threshold, or when the activity level is less than or equal to the activity threshold and the intensity level is greater than the severe threshold, the cutting edge area can be determined to have experienced mild wear; when the activity level is less than or equal to the activity threshold and the intensity level is less than or equal to the severe threshold, the cutting edge area can be determined to be in a normal stage. Here, "normal stage" is preferably understood as the cutting edge wear not yet reaching a level requiring warning.

[0050] In step S400, a wear warning level for the punching die is generated based on the judgment result, and a warning control operation is performed according to the wear warning level. Preferably, the wear warning level includes at least a first wear warning level and a second wear warning level, wherein the first wear warning level corresponds to a severe wear state, and the second wear warning level corresponds to a mild wear state. If the cutting edge area is in a normal stage, only the current monitoring result can be recorded, and the warning control is not triggered.

[0051] In a preferred embodiment, when severe wear occurs in the cutting edge area and a first wear warning level is generated, the control equipment performs a shutdown operation and outputs a prompt for mold replacement or cutting edge re-sharpening to avoid batch defects or mold damage caused by continued production; when slight wear occurs in the cutting edge area and a second wear warning level is generated, the control equipment can reduce the stamping speed, increase the lubrication frequency, or prompt manual inspection to delay further wear development and ensure subsequent stamping quality; when the warning conditions are not met, the control equipment maintains normal operation and stores the monitoring data of the current period in the historical database as the basis for subsequent health benchmark updates and trend analysis.

[0052] Based on another preferred embodiment described above, see [link to previous document]. Figure 3 As shown, this embodiment provides a punching die wear condition monitoring and early warning system for applying the above-mentioned punching die wear condition monitoring and early warning method, including: The acquisition unit is configured to acquire multi-stage monitoring data of the punching die during a continuous stamping cycle. The multi-stage monitoring data includes the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. The analysis unit is configured to perform time-series correlation analysis on multi-stage monitoring data to extract working condition deviation characteristics, mold vibration energy response characteristics, and cutting edge acoustic emission high-frequency activity characteristics. The judgment unit is configured to determine whether there is working condition interference in the current stamping cycle based on the working condition offset characteristics; when it is determined that there is no working condition interference, it determines whether the structural stiffness of the punching die has degraded based on the mold vibration energy response characteristics; when it is determined that the structural stiffness has degraded, it determines whether the cutting edge area of ​​the punching die has worn based on the high-frequency activity characteristics of the cutting edge acoustic emission. The early warning unit is configured to generate a wear warning level for the punching die based on the judgment result, and to perform early warning control operations according to the wear warning level.

[0053] In summary, by collecting vibration signals, acoustic emission signals, and real-time operating parameters of the stamping equipment from the punching die during continuous stamping cycles, and performing time-series correlation analysis on the multi-source monitoring data, joint monitoring of stamping conditions, die structural status, and cutting edge wear status was achieved. By extracting operating condition deviation features and prioritizing the determination of whether there is operating condition interference in the current stamping cycle before wear diagnosis, it is possible to distinguish signal changes caused by abnormal feeding, material fluctuations, lubrication changes, or equipment operating fluctuations from signal changes caused by die wear itself. This avoids misjudging operating condition changes as die wear, thereby reducing the false alarm rate and improving the reliability of online monitoring results. By analyzing the vibration signals during the punching penetration stage and the material removal and reset stage, energy response features corresponding to the low-frequency structural resonance band and the high-frequency impact band are extracted. The first energy ratio and the second energy ratio are used to determine whether the die has experienced structural stiffness degradation. This allows for the detection of structural performance change trends before the die shows obvious failure, enabling early identification of abnormal die conditions. By extracting the ringing count rate and peak amplitude from the acoustic emission signal during the punching penetration stage, and characterizing the degree of microcrack propagation activity and the severity of material spalling at the cutting edge, the high-frequency acoustic emission activity characteristics generated during early micro-damage and continuous wear in the cutting edge region can be captured, thereby achieving refined identification of the cutting edge wear state and improving the sensitivity of wear diagnosis.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for monitoring and early warning of wear condition of punching dies, characterized in that, include: Collect multi-stage monitoring data of the punching die during a continuous stamping cycle. The multi-stage monitoring data includes the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. Time-series correlation analysis was performed on the multi-stage monitoring data to extract working condition deviation characteristics, mold vibration energy response characteristics, and cutting edge acoustic emission high-frequency activity characteristics. The working condition deviation characteristics are used to determine whether there is working condition interference in the current stamping cycle; when it is determined that there is no working condition interference, the mold vibration energy response characteristics are used to determine whether the structural stiffness of the punching mold has degraded. When it is determined that the structural stiffness degradation has occurred, the wear of the cutting edge area of ​​the punching die is determined based on the high-frequency activity characteristics of the cutting edge acoustic emission. Based on the judgment result, a wear warning level for the punching die is generated, and a warning control operation is performed according to the wear warning level.

2. The method for monitoring and early warning of wear status of punching dies according to claim 1, characterized in that, When collecting the multi-stage monitoring data, the following are included: Acquire the slider displacement signal and / or crankshaft angle signal of the stamping equipment; divide the continuous stamping cycle into an unloaded downward stage, a blanking penetration stage, and a material ejection and reset stage based on the slider displacement signal and / or crankshaft angle signal; in the blanking penetration stage and the material ejection and reset stage, synchronously trigger the vibration sensor and acoustic emission sensor installed on the force-bearing part of the punching die to collect the vibration signal and acoustic emission signal.

3. The method for monitoring and early warning of wear status of punching dies according to claim 2, characterized in that, When extracting operating condition offset features, the following are included: Extract the actual stamping tonnage curve and the actual slide speed curve within the current stamping cycle as the real-time operating parameters; compare the actual stamping tonnage curve with the reference tonnage curve under the initial healthy state to obtain the tonnage offset; calculate the deviation between the actual slide speed curve and the reference speed curve under the initial healthy state to obtain the speed offset. The operating condition offset features include the tonnage offset and the speed offset.

4. The method for monitoring and early warning of wear condition of punching dies according to claim 3, characterized in that, When determining whether there is operational interference in the current stamping cycle based on the aforementioned operational deviation characteristics, the following are included: When the tonnage offset is greater than the tonnage fluctuation threshold or the speed offset is greater than the speed fluctuation threshold, it is determined that there is working condition interference in the current stamping cycle; when the tonnage offset is less than or equal to the tonnage fluctuation threshold and the speed offset is less than or equal to the speed fluctuation threshold, it is determined that there is no working condition interference.

5. The method for monitoring and early warning of wear condition of punching dies according to claim 2, characterized in that, When extracting the vibration energy response characteristics of a mold, the following are included: Based on the vibration signals of the punching penetration stage and the material removal and reset stage, the low-frequency structural resonance frequency band and the high-frequency impact frequency band are extracted; the first energy proportion in the low-frequency structural resonance frequency band and the second energy proportion in the high-frequency impact frequency band are determined; The vibration energy response characteristics of the mold include the first energy ratio and the second energy ratio.

6. The method for monitoring and early warning of wear condition of punching dies according to claim 5, characterized in that, When determining whether the punching die has experienced structural stiffness degradation based on the vibration energy response characteristics of the die, the following methods are included: Obtain a first energy baseline value and a second energy baseline value under the initial healthy state; calculate the first energy increase of the first energy ratio and the first energy baseline value, and the second energy increase of the second energy ratio and the second energy baseline value; When the first energy increase is greater than or equal to the second energy increase, and the first energy percentage is greater than the energy warning line, it is determined that the punching die has experienced structural stiffness degradation; when the first energy increase is less than the second energy increase or the first energy percentage is less than or equal to the energy warning line, it is determined that the punching die has not experienced structural stiffness degradation.

7. The method for monitoring and early warning of wear condition of punching dies according to claim 2, characterized in that, Extracting high-frequency activity features of acoustic emission from the cutting edge includes: Based on the acoustic emission signal during the punching penetration stage, the ringing count rate and peak amplitude during the acoustic emission burst are extracted; the activity level of microcrack propagation on the cutting edge is determined based on the ringing count rate, and the severity of material spalling on the cutting edge is determined based on the peak amplitude. The high-frequency activity characteristics of the blade's acoustic emission include the degree of activity and the intensity.

8. The method for monitoring and early warning of wear condition of punching dies according to claim 7, characterized in that, When determining whether wear has occurred in the cutting edge area of ​​the punching die based on the high-frequency activity characteristics of the cutting edge acoustic emission, the following methods are included: The activity level is compared with the activity threshold, and the intensity level is compared with the intensity threshold; When the activity level is greater than the activity threshold and the intensity level is greater than the intensity threshold, it is determined that the cutting edge area has experienced severe wear. When the activity level is greater than the activity threshold and the intensity is less than or equal to the intensity threshold, or when the activity level is less than or equal to the activity threshold and the intensity is greater than the intensity threshold, it is determined that the cutting edge area has experienced slight wear. When the activity level is less than or equal to the activity threshold and the intensity level is less than or equal to the intensity threshold, the blade edge region is determined to be in a normal stage.

9. The method for monitoring and early warning of wear condition of punching dies according to claim 8, characterized in that, When generating the wear warning level of the punching die based on the judgment result, it includes: The wear warning levels include a first wear warning level and a second wear warning level; When the cutting edge area experiences severe wear, a first wear warning level is generated; when the cutting edge area experiences slight wear, a second wear warning level is generated.

10. A punching die wear condition monitoring and early warning system, used to apply the punching die wear condition monitoring and early warning method as described in any one of claims 1-9, characterized in that, include: The acquisition unit is configured to acquire multi-stage monitoring data of the punching die during a continuous stamping cycle. The multi-stage monitoring data includes the vibration signal and acoustic emission signal of the punching die, as well as the real-time operating parameters of the stamping equipment. The analysis unit is configured to perform time-series correlation analysis on the multi-stage monitoring data and extract working condition deviation features, mold vibration energy response features, and cutting edge acoustic emission high-frequency activity features. The judgment unit is configured to determine whether there is working condition interference in the current stamping cycle based on the working condition offset characteristics; When it is determined that there is no interference from the working condition, the structural stiffness degradation of the punching die is determined based on the vibration energy response characteristics of the die. When it is determined that the structural stiffness degradation has occurred, the wear of the cutting edge area of ​​the punching die is determined based on the high-frequency activity characteristics of the cutting edge acoustic emission. The early warning unit is configured to generate a wear warning level for the punching die based on the judgment result, and to perform early warning control operations according to the wear warning level.

Images