A monitoring and early warning method for luggage production

Abstract

The application relates to the technical field of intelligent sensors, and discloses a monitoring and early warning method for luggage production, which comprises collecting acoustic emission signals generated by multilayer materials of luggage in a connecting process, separating low-frequency pressure characteristic sequences representing acoustic impedance changes of the medium and high-frequency transient characteristic sequences representing micro-fractures from the acoustic emission signals, determining a physical judgment threshold dynamically corrected according to real-time acoustic impedance evolution of the materials based on amplitude fluctuation of the low-frequency pressure characteristic sequences, and comparing energy integral values of the high-frequency transient characteristic sequences in a specific frequency band with the physical judgment threshold to generate an early warning signal for microstructure failure. The application effectively eliminates signal distortion caused by nonlinear fluctuation of material pressure density by establishing an acoustic impedance compensation mechanism, and improves the accuracy of the judgment result of the implicit fiber fracture.

Description

Technical Field

[0001] This invention belongs to the field of intelligent sensor technology, and in particular relates to a monitoring and early warning method for bag production. Background Technology

[0002] Current bag manufacturing processes involve the physical connection of multiple heterogeneous materials such as leather, fabric, and skeleton. With the increasing automation of production, online monitoring of the evolution of the internal physical structure of materials during processing has become a key link in ensuring product quality. The industry's common technology mainly uses visual sensors to monitor the quality of stitches or collects feedback from the electronic control parameters of the drive motor to assess the resistance of the needle running, thereby determining the abnormal state of the processing. However, the multi-layered materials of bags are subjected to transient mechanical stress impacts during high-speed needle punching or high-frequency welding, which can easily cause hidden physical damage such as matrix cracking or fiber layer peeling. Since such damage occurs inside heterogeneous materials and has a short evolution cycle, visual inspection methods are limited by the physical characteristics of optical reflection and cannot penetrate the surface to detect the structural continuity of the material. The motor feedback signals are mostly low-frequency mechanical quantities, and their response frequency band is far lower than the high-frequency elastic wave energy released by fiber breakage, which cannot effectively analyze the physical state of the material during processing.

[0003] The industry has attempted to introduce acoustic emission sensing technology, using high-frequency piezoelectric sensors to capture the elastic waves released when materials fracture. However, in actual industrial production environments, when heterogeneous materials in bags are subjected to high-speed instantaneous compression by processing parts, their local physical density and interlayer contact stiffness are in a state of severe nonlinear dynamic deformation. The transient compaction state causes an order-of-magnitude jump in the acoustic impedance of the material. The existing static judgment threshold cannot adapt to the real-time drift of the acoustic impedance of the medium caused by processing stress, resulting in severe characteristic distortion of the high-frequency response signal captured by the sensor under the static judgment benchmark. This implicit constraint leads to a high false alarm rate in the monitoring system in actual production lines, making it difficult to accurately determine the physical integrity of composite materials.

[0004] Therefore, based on the evolution of transient acoustic characteristics of the material processing interface within a millisecond period, an intelligent sensing and early warning method with physical background cancellation capability and adaptive compensation capability for judgment threshold is constructed to achieve non-destructive and quantitative measurement of hidden physical damage to bag composite materials, which is the technical problem to be solved by this invention. Summary of the Invention

[0005] This invention proposes a monitoring and early warning method for bag production, comprising the following steps: Step S101: Using an array of acoustic emission sensors arranged around the bag processing mold, broadband acoustic emission signals induced by transient mechanical stress are collected during the physical connection process of sewing or welding of multi-layer heterogeneous materials of bags. Step S102: Extract the low-frequency pressure feature sequence, which characterizes the real-time density evolution and acoustic impedance change of the multilayer heterogeneous material of the bag under pressure, and the high-frequency transient feature sequence, which characterizes the release of elastic potential energy by microscopic damage inside the multilayer heterogeneous material of the bag, from the broadband acoustic emission signal using a frequency division decoupling algorithm. Step S103: Based on the real-time amplitude fluctuation law of the low-frequency pressure characteristic sequence, establish a nonlinear compensation model for the attenuation of sound wave transmission by the acoustic impedance of the medium in the physical connection process, and calculate the physical judgment threshold that changes in real time with the degree of compaction of the multi-layer heterogeneous material of the bag based on the nonlinear compensation model, so that the physical judgment threshold can automatically offset the acoustic characteristic distortion interference caused by the extrusion of the processing parts. Step S104: Calculate the transient energy integral value of the high-frequency transient characteristic sequence in the preset physical characteristic frequency band from 150kHz to 300kHz. The transient energy integral value is used to quantitatively characterize the microscopic damage intensity of internal fiber layer fracture or interface peeling in the multilayer heterogeneous material of the bag. Step S105: Compare the transient energy integral value with the physical judgment threshold. When the transient energy integral value exceeds the physical judgment threshold, determine that the multi-layer heterogeneous material of the bag is in a state of microstructural failure and output a warning signal.

[0006] Preferably, step S103 includes the following sub-steps: step S1031, obtaining the envelope amplitude intensity of the low-frequency pressure characteristic sequence; step S1032, mapping the envelope amplitude intensity to an attenuation correction amount using a material impedance compensation coefficient; step S1033, superimposing the attenuation correction amount onto a static reference threshold to obtain a physical judgment threshold, so as to counteract the energy shielding interference caused by the pressure density evolution of the multilayer heterogeneous material of the bag to the high-frequency transient characteristic sequence.

[0007] Preferably, step S102 specifically includes: using a low-pass filtering algorithm with a cutoff frequency of 20kHz to smooth the broadband acoustic emission signal and extract the low-frequency pressure feature sequence reflecting the processing pressure fluctuation; at the same time, using a band-pass filtering algorithm of 150kHz to 400kHz to remove the broadband mechanical noise generated by mechanical operation in order to obtain the high-frequency transient feature sequence.

[0008] Preferably, step S104 specifically includes: within a sliding time window of 10μs to 50μs, squaring the transient amplitude of the high-frequency transient characteristic sequence and calculating the time accumulation to obtain the transient energy integral value characterizing the strain energy release intensity inside the multilayer heterogeneous material of the bag, which is used to determine the initiation of hidden microcracks.

[0009] Preferably, step S105 further includes: calculating the growth gradient of the transient energy integral value within a preset connection period; identifying the evolution stage of the multilayer heterogeneous material of the bag from stress accumulation to fiber breakage based on the growth gradient; and dynamically adjusting the audible and visual warning frequency or data sampling rate of the warning signal according to the degree of danger of the evolution stage.

[0010] Preferably, step S101 includes: capturing elastic wave signals by multiple acoustic emission sensors arranged around the processing position; transmitting the signals captured by each acoustic emission sensor to an impedance matching device for impedance equalization through a shielded cable; and using a signal conditioning unit to pre-amplify and anti-aliasing filter the equalized signal to convert it into a digital broadband acoustic emission signal.

[0011] Preferably, after the warning signal is output in step S105, the following operations are also included: controlling the connected equipment to reduce the mold closing rate or to stop the machine in an emergency; at the same time, using a marking inkjet device to spray physical marks on the damage coordinate positions on the surface of the multi-layer heterogeneous materials of the bag, and writing the damage data corresponding to the warning signal into the electronic quality traceability file of the batch of products.

[0012] Preferably, the method further includes an environmental noise reduction step: acquiring environmental background noise signals under no-load processing conditions; calculating the spectral distribution characteristics of the environmental background noise signals; and using the spectral distribution characteristics to perform real-time spectral subtraction processing on the broadband acoustic emission signals acquired in step S101 to suppress power frequency interference and high-frequency pulse noise in the production line environment.

[0013] Preferably, the method also includes a parameter closed-loop optimization step: performing a correlation analysis between the physical judgment threshold of each processing cycle and the destructive tensile strength data of the finished product; and automatically updating the material impedance compensation coefficient in step S1032 based on the deviation trend obtained from the correlation analysis, so as to achieve adaptive parameter optimization for different batches of bag materials.

[0014] Compared with existing technologies, the monitoring and early warning method for bag production of the present invention has the following advantages: 1. In the monitoring and early warning of bag production, a dual-source sensing link of main acoustic emission signal and reference acoustic signal is constructed. The time offset parameter of the two in the short-time energy envelope is calculated based on the dynamic time warping algorithm. This solves the phase distortion problem caused by multipath dispersion effect in the transmission of mechanical waves in multilayer heterogeneous composite materials. It realizes nonlinear phase reconstruction and physical background offset of mechanical background noise of processing equipment, and removes strong mechanical impact interference in high-speed sewing or welding process. This enables the extracted intrinsic acoustic wave signal to truly and purely characterize the transient damage state of the internal microstructure of bag materials.

[0015] 2. By utilizing the obtained macroscopic thickness parameters of the bag material, the density parameters of the matrix material, and the acoustic medium attenuation physical model, the real-time acoustic attenuation damping factor is calculated. Based on this, the initial benchmark threshold is exponentially compensated and amplified to form a physical fracture threshold that dynamically floats with the material's stress and compaction state. This breaks through the logical bottleneck of traditional monitoring systems that rely on constant thresholds to measure the dynamic stress state of materials, eliminates the submersion or distortion of characteristic signals caused by the nonlinear evolution of the density of composite materials during processing, and ensures the accuracy and objectivity of the determination of hidden fiber fracture and interface peeling.

[0016] 3. By extracting high-frequency response sequences and low-frequency compaction envelope sequences, and establishing a physical quantization mapping relationship between the two based on time integration, a feedback mechanism is constructed that directly constrains material damage determination by the mechanical kinetic energy input of the equipment. This enables the calculated net physical damage index to automatically offset acoustic impedance interference caused by processing speed fluctuations or material batch thickness tolerances. Without changing the sensor hardware accuracy, the overall signal-to-noise ratio of the test system is improved by utilizing the logical coupling of multi-dimensional physical parameters, thereby enhancing the sensor system's ability to quantitatively test the internal physical continuity of heterogeneous materials under extreme noise environments. Attached Figure Description

[0017] Figure 1 This is a timing flowchart of the monitoring and early warning method for the physical connection process of multi-layer materials in bags according to the present invention; Figure 2 This is a signal analysis and response logic diagram for monitoring and early warning of physical damage to bag materials according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] It should be noted that all directional and positional terms used in this invention, such as: up, down, left, right, front, back, vertical, horizontal, inner, outer, top, bottom, transverse, longitudinal, center, etc., are only used to explain the relative positional relationship and connection between components in a specific state (as shown in the accompanying drawings). They are only for the convenience of describing this invention and do not require that this invention be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention. In addition, the descriptions of "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated.

[0020] In the description of this invention, unless otherwise explicitly specified and limited, the terms installation, connection, and linking should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections; they can refer to direct connections or indirect connections through an intermediate medium; they can refer to the internal connection of two components. For those skilled in the art, the specific meaning of the above terms in this invention can be understood according to the specific circumstances.

[0021] In the description of this specification, references to the terms "an embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example, and the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0022] A monitoring and early warning method for bag production includes the following steps: Step S101: Using an array of acoustic emission sensors arranged around the bag processing mold, broadband acoustic emission signals induced by transient mechanical stress are collected during the physical connection process of sewing or welding of multi-layer heterogeneous materials of bags. Step S102: Extract the low-frequency pressure feature sequence, which characterizes the real-time density evolution and acoustic impedance change of the multilayer heterogeneous material of the bag under pressure, and the high-frequency transient feature sequence, which characterizes the release of elastic potential energy by microscopic damage inside the multilayer heterogeneous material of the bag, from the broadband acoustic emission signal using a frequency division decoupling algorithm. Step S103: Based on the real-time amplitude fluctuation law of the low-frequency pressure characteristic sequence, establish a nonlinear compensation model for the attenuation of sound wave transmission by the acoustic impedance of the medium in the physical connection process, and calculate the physical judgment threshold that changes in real time with the degree of compaction of the multi-layer heterogeneous material of the bag based on the nonlinear compensation model, so that the physical judgment threshold can automatically offset the acoustic characteristic distortion interference caused by the extrusion of the processing parts. Step S104: Calculate the transient energy integral value of the high-frequency transient characteristic sequence in the preset physical characteristic frequency band from 150kHz to 300kHz. The transient energy integral value is used to quantitatively characterize the microscopic damage intensity of internal fiber layer fracture or interface peeling in the multilayer heterogeneous material of the bag. Step S105: Compare the transient energy integral value with the physical judgment threshold. When the transient energy integral value exceeds the physical judgment threshold, determine that the multi-layer heterogeneous material of the bag is in a state of microstructural failure and output a warning signal.

[0023] Preferably, step S103 includes the following sub-steps: step S1031, obtaining the envelope amplitude intensity of the low-frequency pressure characteristic sequence; step S1032, mapping the envelope amplitude intensity to an attenuation correction amount using a material impedance compensation coefficient; step S1033, superimposing the attenuation correction amount onto a static reference threshold to obtain a physical judgment threshold, so as to counteract the energy shielding interference caused by the pressure density evolution of the multilayer heterogeneous material of the bag to the high-frequency transient characteristic sequence.

[0024] Preferably, step S102 specifically includes: using a low-pass filtering algorithm with a cutoff frequency of 20kHz to smooth the broadband acoustic emission signal and extract the low-frequency pressure feature sequence reflecting the processing pressure fluctuation; at the same time, using a band-pass filtering algorithm of 150kHz to 400kHz to remove the broadband mechanical noise generated by mechanical operation in order to obtain the high-frequency transient feature sequence.

[0025] Preferably, step S104 specifically includes: within a sliding time window of 10μs to 50μs, squaring the transient amplitude of the high-frequency transient characteristic sequence and calculating the time accumulation to obtain the transient energy integral value characterizing the strain energy release intensity inside the multilayer heterogeneous material of the bag, which is used to determine the initiation of hidden microcracks.

[0026] Preferably, step S105 further includes: calculating the growth gradient of the transient energy integral value within a preset connection period; identifying the evolution stage of the multilayer heterogeneous material of the bag from stress accumulation to fiber breakage based on the growth gradient; and dynamically adjusting the audible and visual warning frequency or data sampling rate of the warning signal according to the degree of danger of the evolution stage.

[0027] Preferably, in step S103, the physical judgment threshold is calculated. Follow these rules: ,in, is the physical judgment threshold; is the preset static reference threshold; is the impedance gain coefficient; This represents the real-time amplitude of the low-frequency pressure characteristic sequence; This is the preset material characteristic pressure constant.

[0028] Preferably, step S101 includes: capturing elastic wave signals by multiple acoustic emission sensors arranged around the processing position; transmitting the signals captured by each acoustic emission sensor to an impedance matching device for impedance equalization through a shielded cable; and using a signal conditioning unit to pre-amplify and anti-aliasing filter the equalized signal to convert it into a digital broadband acoustic emission signal.

[0029] Preferably, after the warning signal is output in step S105, the following operations are also included: controlling the connected equipment to reduce the mold closing rate or to stop the machine in an emergency; at the same time, using a marking inkjet device to spray physical marks on the damage coordinate positions on the surface of the multi-layer heterogeneous materials of the bag, and writing the damage data corresponding to the warning signal into the electronic quality traceability file of the batch of products.

[0030] Preferably, the method further includes an environmental noise reduction step: acquiring environmental background noise signals under no-load processing conditions; calculating the spectral distribution characteristics of the environmental background noise signals; and using the spectral distribution characteristics to perform real-time spectral subtraction processing on the broadband acoustic emission signals acquired in step S101 to suppress power frequency interference and high-frequency pulse noise in the production line environment.

[0031] Preferably, the method also includes a parameter closed-loop optimization step: performing a correlation analysis between the physical judgment threshold of each processing cycle and the destructive tensile strength data of the finished product; and automatically updating the material impedance compensation coefficient in step S1032 based on the deviation trend obtained from the correlation analysis, so as to achieve adaptive parameter optimization for different batches of bag materials.

[0032] Example 1: In the operation of a mold used for high-speed stitching of multi-layered heterogeneous materials such as leather and nylon skeletons in bag processing, the high-frequency physical penetration of the needle generates broadband mechanical noise. Simultaneously, the multi-layered heterogeneous materials at the processing interface are subjected to transient mechanical stress impacts, causing nonlinear dynamic deformation of the local physical density and interlayer contact stiffness. This transient compaction state leads to a magnitude jump in the acoustic impedance of the material medium, causing conventional methods relying on static benchmarks to measure the release of elastic potential energy to suffer from energy shielding interference and signal distortion when faced with real-time drift in the medium impedance. Using fabric... An array of acoustic emission sensors placed around the mold for bag processing collects broadband acoustic emission signals induced by transient mechanical stress during the physical connection process of stitching multi-layered heterogeneous materials in bags. A low-pass filtering algorithm with a cutoff frequency of 20kHz is used to smooth and extract the low-frequency pressure feature sequence characterizing the real-time density evolution and acoustic impedance change from the broadband acoustic emission signal. At the same time, a band-pass filtering algorithm from 150kHz to 400kHz is used to remove the broadband mechanical noise generated by mechanical operation, and to obtain the high-frequency transient feature sequence characterizing the elastic potential energy released by microscopic damage inside the multi-layered heterogeneous materials of the bag.This signal frequency division and extraction process directly transforms the aliased mechanical impact components into fundamental physical mapping parameters for probing the compressive state of the material. The underlying sensing physical transformation is based on the fact that when a piezoelectric sensor encounters a low-frequency, high-amplitude external excitation—the strong pressure applied by the processing machine—in addition to the microscopic high-frequency resonance response of the internal chip, the sensor base and the piezoelectric ceramic oscillator body will experience a relatively slow polarization charge discharge effect under macroscopic strain due to mechanical compressive stress. This low-frequency baseline drift electrical signal, directly excited by mechanical push-pull, can be smoothly extracted using a 20kHz cutoff frequency to faithfully reconstruct a dynamic profile of material compaction that fluctuates with transient physical processing pressure. This process also removes the high-energy mechanical impact components transmitted by rigid components of the equipment. Based on the principle of multi-dimensional time-series signal morphological similarity measurement, a vibration acceleration sensor rigidly fixed to the end of the processing equipment spindle is introduced as an auxiliary data source link to synchronously acquire the structural vibration reference sound signal associated with the needle penetration action. The control unit then extracts these signals. The short-time energy envelopes of the broadband acoustic emission signal and the structural vibration reference acoustic signal are calculated using a dynamic time warping algorithm. This algorithm calculates the shortest alignment path matrix of the two independent waveform sequences within the same sliding window, outputting a nonlinear time-series offset parameter. Specifically, because the high-frequency micro-stress wave and the low-frequency macro-mechanical wave have order-of-magnitude differences in the original frequency domain and time axis, they cannot directly produce physical interference. The differential subtraction is not a direct mathematical subtraction of the original high-frequency waveform. Instead, the time-series offset parameter extracted by dynamic time warping is applied to the energy envelope layer. This offset parameter is used to precisely align the time window of the low-frequency energy envelope, which includes macro-mechanical background noise, with the calculation time window of the high-frequency sound wave in the time domain. This allows the periodic environmental mechanical disturbance energy component to be subtracted within the same macro-energy calculation benchmark. The processing module performs inverse translation compensation on the structural vibration reference acoustic signal in the time domain based on the nonlinear time-series offset parameter, differentially subtracting it from the broadband acoustic emission signal. This process suppresses the periodic mechanical background noise in the physical waveform.

[0033] The envelope amplitude intensity of the low-frequency pressure characteristic sequence is obtained as the real-time amplitude. Based on the amplitude fluctuation law, a nonlinear compensation model for the attenuation of sound wave transmission by the acoustic impedance of the medium in the physical connection process is established, and the formula is applied. Calculate the physical judgment threshold that changes in real time with the degree of compaction of the multi-layered heterogeneous materials of the bag, where, This is the physical judgment threshold. The preset static baseline threshold, is the impedance gain coefficient, which is a dimensionless constant; This represents the real-time amplitude of the low-frequency pressure characteristic sequence. As a preset material characteristic pressure constant, this formula directly transforms the low-frequency parameter characterizing the mechanical kinetic energy input into a dynamic compensation boundary for the high-frequency judgment threshold, offsetting the energy shielding interference caused by the evolution of the extrusion density of the processed parts on the high-frequency transient characteristic sequence. Within a sliding time window of 10μs to 50μs, the square of the transient amplitude of the high-frequency transient characteristic sequence is calculated and the time accumulation is obtained, yielding the transient energy integral value within the preset physical characteristic frequency band of 150kHz to 300kHz. To address the physical interference of high-frequency pulse overlap caused by the increase in the puncture frequency of the processed parts, the control unit continuously monitors the time interval between two adjacent puncture actions. When the time interval is less than the safety analysis cycle benchmark value, it is adjusted according to the preset proportional coefficient. To reduce the width of the sliding time window and avoid energy aliasing, based on the dynamic law of microcrack propagation in solid materials, the control unit extracts the transient energy integral value within five consecutive sliding time windows. The first-order difference algorithm is used to calculate the incremental difference quotient of the integral value within adjacent time windows to obtain the real-time growth gradient. The control unit compares the real-time growth gradient with the numerical values ​​of a mapping table established in advance through destructive tensile experiments. The logic for obtaining the reference boundary value inside the mapping table is as follows: In the offline material test during the preparatory stage, the load curve of the universal tensile testing machine and the evolution of the acoustic emission characteristic parameters of the sample are compared simultaneously. The extreme value of the energy integral slope induced by the cluster of small abrupt changes when the load curve deviates from the linear proportional limit and enters the initial stage of plastic yield is marked as the first gradient boundary value.

[0034] Simultaneously, the characteristics of the violent elastic wave energy impulse step at the moment of visible macroscopic fiber breakage and instantaneous load drop at the tensile fracture surface are captured, and its slope is set as the second gradient boundary value. If the real-time growth gradient is less than the calibrated first gradient boundary value, it is determined that the multilayer heterogeneous material of the bag is in the elastic stress accumulation stage; when it is between the first and second gradient boundary values, it is determined that it has entered the microcrack initiation stage; when it is greater than the second gradient boundary value, it is determined that it has reached the macroscopic fiber fracture stage. When the microcrack initiation stage is detected, the control unit sends a trigger message to the bottom analog-to-digital conversion chip of the acoustic emission sensor array, and increases the data sampling frequency from the initial operating frequency multiplier to 5MHz to capture high-frequency fracture precursor characteristics, comparing the transient energy integral value with the physical judgment threshold. When the transient energy integral value exceeds the physical judgment threshold, the multi-layer heterogeneous material of the luggage is determined to be in a state of microstructural failure, and an early warning signal is output. The output quantitative result eliminates the physical interference of the extrusion force of the processing parts on the high-frequency sound wave transmission path and quantitatively characterizes the micro-damage intensity of the internal fiber layer fracture or interface peeling of the multi-layer heterogeneous material of the luggage. It overcomes the long-term acoustic drift caused by the difference in interface properties of materials from different batches. Based on the state feedback control principle of discrete system, a closed-loop optimization procedure for the material impedance compensation coefficient parameter is constructed. The system records the output physical judgment threshold for each connection cycle, receives the offline measured value of destructive tensile strength returned by the tensile testing machine after the finished product is off the line, calculates the relative deviation rate between the offline measured value and the process reference tensile strength, and the control unit uses the state update formula. Adjust the current impedance compensation coefficient of the system, where, The dimensionless gain coefficient representing the new impedance required for the next processing cycle; This represents the dimensionless initial impedance gain coefficient actually invoked within the current calculation cycle; The representative, based on the calibration of the basic material trial mold, ensures that the system learning step size is a real number greater than zero to guarantee the convergence of the feedback loop. The relative deviation rate obtained by the calculation is dimensionless. The system relies on the macroscopic physical damage index of the finished product to successively and adaptively correct the pre-dynamic acoustic compensation boundary. In this cross-scale closed-loop feedback mechanism, the deviation of the macroscopic destructive tensile strength of the finished product maps the global statistical evolution of the density of unclosed micropores and interface delamination defects inside the multilayer composite material at the mechanical level. The continuity of the micropores and interfaces is precisely the fundamental physical intrinsic parameter that determines the attenuation rate of high-frequency elastic wave scattering. Therefore, by extracting the degradation deviation gradient of macroscopic destructive mechanics, the control unit can indirectly deduce and linearly compensate for the overall drift of the microscopic acoustic damping characteristics inside the material in the form of physical positive correlation, thus establishing a self-consistent mechanism mapping closed loop.

[0035] Example 2: In this example, a wideband acoustic emission sensor with a sampling rate of 2MHz and an adjustable speed servo drive unit were selected. A physical test environment was constructed based on a heavy industrial sewing machine platform. Gaussian white noise with a signal-to-noise ratio of 20dB was injected into the acoustic emission signal acquisition link, and a 50Hz power frequency interference harmonic with an amplitude of 0.5V was superimposed. The sliding time window width was set according to the physical puncture frequency. When the physical puncture frequency of the processed part increased and the material showed a tendency to brittle fracture, the duration of the excited acoustic emission pulse shortened. 20μs was selected as the benchmark test value of the sliding time window. Three step speeds of 500rpm, 1500rpm, and 2500rpm were selected as the test gradient for sewing processing. An experimental group, control group 1, and control group 2 were established. The experimental group used a low-frequency pressure characteristic sequence to dynamically adjust the physical judgment threshold. Control group 1 used a fixed static threshold. Control group 2 set the sliding time window to 80μs and used a 50kHz low-pass filter.

[0036] The real-time amplitude of the low-frequency pressure feature sequence extracted by the broadband acoustic emission sensor under a puncture condition of 2500 rpm. Reaching 3.5 MPa, according to the formula Calculate the physical judgment threshold, where, This is the physical judgment threshold. The preset static baseline threshold, is the impedance gain coefficient, which is a dimensionless constant; This represents the real-time amplitude of the low-frequency pressure characteristic sequence. Using the preset material characteristic pressure constant, the calculated physical judgment threshold ranges from the static reference threshold of 15.0mV to 46.2mV. This calculation process generates a high-frequency judgment boundary that is dynamically compensated for by the extrusion force of the processed parts.

[0037] Comparing the output warning signals under gradient operating conditions, the false alarm rate of control group 1 increased nonlinearly with the increase of rotational speed, from 2.1% at 500 rpm to 68.4% at 2500 rpm. Control group 2 caused energy aliasing of adjacent high-frequency transient characteristic sequences within an 80 μs window, reducing the detection rate of fiber layer fracture to 72.1%. The false alarm rate of the experimental group remained within the range of 1.1% to 1.4% under full gradient operating conditions, with a detection rate greater than 98.5%. When the test speed exceeded the physical load point of 2800 rpm, the pulse interval of the high-frequency transient characteristic sequence was less than 20 μs, and the transient energy integral value showed a saturation trend. The detection rate of the experimental group dropped to 85.2%. The above quantitative data defined the numerical boundaries of the dynamic threshold and characteristic frequency band in suppressing mechanical impedance interference. The low-frequency pressure characteristic sequence extracted by the low-pass filtering algorithm and the high-frequency transient characteristic sequence extracted by the band-pass filtering algorithm had a synergistic effect under the nonlinear compensation model.

[0038] Example 3: In a production line deployment scenario where multi-layered heterogeneous materials for bags are being replaced, the dynamic compensation model for the physical judgment threshold relies on the input of the material characteristic pressure constant and impedance gain coefficient. Using default parameters to replace the intrinsic physical parameters of a specific material disrupts the physical mapping relationship between acoustic impedance evolution and mechanical compaction, leading to the failure of energy shielding compensation for high-frequency transient characteristic sequences. This causes the monitoring and early warning system to experience judgment boundary drift and false alarms when faced with new composite fabrics. To address this, a preset material characteristic pressure constant and impedance gain coefficient are determined. A 10cm × 10cm sample of the multi-layered heterogeneous material to be tested is placed on a universal testing machine platform equipped with a force sensor. An acoustic emission sensor array of the same specifications as the production line is attached to the sample surface. The universal testing machine applies a vertical compressive load to the sample at a constant rate of 5mm / min, simultaneously acquiring reference acoustic emission signals and real-time mechanical stress values ​​under the applied load. The critical mechanical stress value at which the sample transitions from the elastic deformation stage to the plastic yielding stage is extracted and set as the material characteristic pressure constant. .

[0039] During the application of a vertical compressive load, a standard high-frequency probe pulse with a frequency of 200 kHz is continuously injected into the sample using an external ultrasonic generator. The peak voltage received after the standard high-frequency probe pulse passes through the sample is measured, and the initial peak voltage received when no load is applied and the pressure constant equal to the material's characteristic pressure are extracted. The target received peak voltage under load is calculated, and the difference between the initial received peak voltage and the target received peak voltage is calculated. This difference is divided by the initial received peak voltage to obtain the acoustic attenuation ratio. The acoustic attenuation ratio is set as the impedance gain coefficient. The envelope amplitude of the low-frequency pressure characteristic sequence is obtained. The absolute value of the low-frequency pressure characteristic sequence after low-pass filtering is extracted, and the full-wave rectified result is output to generate a unipolar pressure pulsation signal sequence. A digital moving average filter is used to smooth the time domain characteristics of the unipolar pressure pulsation signal sequence. The window length of the digital moving average filter is set to the time span of a single reciprocating puncture of the workpiece. The arithmetic mean of the unipolar pressure pulsation signal sequence within the window length is calculated and output, and this is used as the envelope amplitude of the low-frequency pressure characteristic sequence (i.e., the real-time amplitude). ); Extracted material characteristic pressure constant Impedance gain coefficient and real-time amplitude Substitute it into the formula as an input parameter. ,in, This is the physical judgment threshold. The preset static baseline threshold, is the impedance gain coefficient, which is a dimensionless constant; This represents the real-time amplitude of the low-frequency pressure characteristic sequence. The physical judgment threshold, calculated using the preset material characteristic pressure constant, is compared with the transient energy integral value to determine the microstructural failure state of the multilayer heterogeneous material of the bag.

[0040] Example 4: When the system faces the initial physical deployment of a newly set bag processing mold, the basic background noise of the working environment and the inherent time-domain vibration components generated by the no-load operation of the processing parts constitute the physical interference boundary for high-frequency transient feature sequence extraction. The heavy industrial sewing machine platform is controlled to run continuously at a preset working speed for 5 minutes in the no-load state without placing multi-layer heterogeneous materials of bags. The continuous acoustic emission reference signal in the no-load operation state is synchronously collected using an acoustic emission sensor array. The basic high-frequency background sequence in the continuous acoustic emission reference signal is extracted using a bandpass filtering algorithm from 150kHz to 400kHz. The basic high-frequency background sequence contains mechanical transmission characteristics and environmental interference characteristics. The transient amplitude square of the basic high-frequency background sequence is calculated within a sliding time window of 10μs to 50μs and the time accumulation is obtained. The reference energy integral sequence of the background disturbance intensity corresponding to the no-load state is output.

[0041] The arithmetic mean of the reference energy integral sequence under no-load operation is extracted as the baseline of the basic interference energy. The baseline of the basic interference energy is multiplied by a preset tolerance margin coefficient to obtain a scalar amplification value, which is then set as the static reference threshold. , set static benchmark threshold Material characteristic pressure constant And the impedance gain coefficient is substituted into the formula. ,in, This is the physical judgment threshold. is the static reference threshold, and is the impedance gain coefficient, which is a dimensionless constant; This represents the real-time amplitude of the low-frequency pressure characteristic sequence. The material's characteristic pressure constant is compared with the physical judgment threshold. Compared with the real-time acquired transient energy integral value, when the transient energy integral value is greater than the physical judgment threshold... It outputs a warning signal in real time, and the output warning signal quantitatively characterizes the microstructural failure state of the multilayer heterogeneous material of the bag.

[0042] Example 5: When facing the trial production of a new type of composite fiber woven fabric, the filter boundary and early warning tolerance parameters set by the solidified numerical values ​​deviate from the physical sound emission law of the new material, causing the high-frequency transient characteristic sequence extraction to be mixed with environmental noise or missing fracture pulses; a 10cm×10cm sample of the bag material to be tested is selected and clamped in the tensile fixture of the universal testing machine; the universal testing machine is controlled to apply a linearly increasing tensile load to the sample until the sample physically fractures; a broadband acoustic emission sensor attached to the sample surface is used to collect the acoustic wave signal of the entire tensile failure process, and the 5ms before and after the corresponding tensile load transient drop moment is extracted. The abrupt acoustic wave sequence within the range is analyzed; the fast Fourier transform algorithm is used to convert the abrupt acoustic wave sequence to the frequency domain and output the energy spectral density distribution curve; the absolute peak point of the energy amplitude is found in the energy spectral density distribution curve, and the lower cutoff frequency and upper cutoff frequency containing 90% of the total energy distribution of the absolute peak point of the energy amplitude are calculated by bidirectional integration to low and high frequencies; the lower cutoff frequency and upper cutoff frequency are set as the boundary parameters of the bandpass filtering algorithm to define the preset physical characteristic frequency band from 150kHz to 300kHz; this physical measurement procedure locks the frequency range of the filtering algorithm within the objective acoustic release boundary of the material molecular chain breakage.

[0043] A continuous 10-second sampling segment of the acoustic signal during the linear elastic tensile deformation stage of the sample was extracted as a background noise control sample; the standard deviation of the energy integral sequence of the background noise control sample within a preset physical characteristic frequency band was calculated, and the highest transient energy integral peak value of the abrupt acoustic signal sequence within the preset physical characteristic frequency band was obtained; according to the formula... Calculate the preset tolerance margin coefficient, where... This is a preset tolerance margin coefficient, which is a dimensionless constant. The highest transient energy integral peak value, The arithmetic mean of the energy integral sequences of the background noise control samples. The standard deviation of the energy integral sequence of the background noise control sample is used to substitute the calculated preset tolerance margin coefficient into the pre-calibration process to establish the static benchmark threshold of the specific material. The above joint calibration parameters establish extraction rules according to the tensile fracture phenomenon and energy statistical distribution law. The established calculation path eliminates the parameter blind spots of the filter boundary and tolerance coefficient setting, and maintains the consistency of quantitative judgment of the monitoring and early warning system in different batches of material processing production lines.

[0044] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit of this application and the scope of protection of this invention, and all of these forms are within the protection scope of this application.

Claims

1. A monitoring and early warning method for bag production, characterized in that, Includes the following steps: Step S101: Using an array of acoustic emission sensors arranged around the bag processing mold, broadband acoustic emission signals induced by transient mechanical stress are collected during the physical connection process of sewing or welding of multi-layer heterogeneous materials of bags. Step S102: Extract the low-frequency pressure feature sequence, which characterizes the real-time density evolution and acoustic impedance change of the multilayer heterogeneous material of the bag under pressure, and the high-frequency transient feature sequence, which characterizes the release of elastic potential energy by microscopic damage inside the multilayer heterogeneous material of the bag, from the broadband acoustic emission signal using a frequency division decoupling algorithm. Step S103: Based on the real-time amplitude fluctuation law of the low-frequency pressure characteristic sequence, establish a nonlinear compensation model for the attenuation of sound wave transmission by the acoustic impedance of the medium in the physical connection process, and calculate the physical judgment threshold that changes in real time with the degree of compaction of the multi-layer heterogeneous material of the bag based on the nonlinear compensation model, so that the physical judgment threshold can automatically offset the acoustic characteristic distortion interference caused by the extrusion of the processing parts. Step S104: Calculate the transient energy integral value of the high-frequency transient characteristic sequence in the preset physical characteristic frequency band from 150kHz to 300kHz. The transient energy integral value is used to quantitatively characterize the microscopic damage intensity of internal fiber layer fracture or interface peeling in the multilayer heterogeneous material of the bag. Step S105: Compare the transient energy integral value with the physical judgment threshold. When the transient energy integral value exceeds the physical judgment threshold, determine that the multi-layer heterogeneous material of the bag is in a state of microstructural failure and output a warning signal.

2. The monitoring and early warning method for bag production according to claim 1, characterized in that, Step S103 includes the following sub-steps: Step S1031, obtaining the envelope amplitude intensity of the low-frequency pressure characteristic sequence; Step S1032, mapping the envelope amplitude intensity to an attenuation correction amount using the material impedance compensation coefficient; Step S1033, superimposing the attenuation correction amount onto the static reference threshold to obtain a physical judgment threshold, so as to counteract the energy shielding interference caused by the pressure density evolution of the multilayer heterogeneous material of the bag to the high-frequency transient characteristic sequence.

3. The monitoring and early warning method for bag production according to claim 1, characterized in that, Step S102 specifically includes: using a low-pass filtering algorithm with a cutoff frequency of 20kHz to smooth the broadband acoustic emission signal and extract the low-frequency pressure feature sequence reflecting the processing pressure fluctuation; at the same time, using a band-pass filtering algorithm from 150kHz to 400kHz to remove the broadband mechanical noise generated by mechanical operation in order to obtain the high-frequency transient feature sequence.

4. The monitoring and early warning method for bag production according to claim 1, characterized in that, Step S104 specifically includes: within a sliding time window of 10μs to 50μs, squaring the transient amplitude of the high-frequency transient characteristic sequence and calculating the time accumulation to obtain the transient energy integral value characterizing the strain energy release intensity inside the multilayer heterogeneous material of the bag, which is used to determine the initiation of hidden microcracks.

5. A monitoring and early warning method for bag production according to claim 1, characterized in that, Step S105 further includes: calculating the growth gradient of the transient energy integral value within a preset connection period; identifying the evolution stage of the multilayer heterogeneous material of the bag from stress accumulation to fiber breakage based on the growth gradient; and dynamically adjusting the audible and visual warning frequency or data sampling rate of the warning signal according to the degree of danger of the evolution stage.

6. A monitoring and early warning method for bag production according to claim 1, characterized in that, Step S101 includes: capturing elastic wave signals by multiple acoustic emission sensors arranged around the processing position; transmitting the signals captured by each acoustic emission sensor to an impedance matching device for impedance equalization through shielded cables; and using a signal conditioning unit to pre-amplify and anti-aliasing filter the equalized signals to convert them into digital broadband acoustic emission signals.

7. A monitoring and early warning method for bag production according to claim 1, characterized in that, After the warning signal is output in step S105, the following operations are also included: controlling the connected equipment to reduce the mold closing rate or to stop the machine in an emergency; at the same time, using a marking inkjet device to spray physical marks on the damage coordinate positions on the surface of the multi-layer heterogeneous materials of the bag, and writing the damage data corresponding to the warning signal into the electronic quality traceability file of the batch of products.

8. A monitoring and early warning method for bag production according to claim 1, characterized in that, It also includes an environmental noise reduction step: collecting environmental background noise signals under no-load processing conditions; calculating the spectral distribution characteristics of the environmental background noise signals; and using the spectral distribution characteristics to perform real-time spectral subtraction processing on the broadband acoustic emission signals obtained in step S101 to suppress power frequency interference and high-frequency pulse noise in the production line environment.

9. A monitoring and early warning method for bag production according to claim 2, characterized in that, It also includes a parameter closed-loop optimization step: performing a correlation analysis between the physical judgment threshold of each processing cycle and the destructive tensile strength data of the finished product; and automatically updating the material impedance compensation coefficient in step S1032 based on the deviation trend obtained from the correlation analysis, so as to achieve adaptive parameter optimization for different batches of bag materials.

Images