A method and system for inspecting a packaging device

Abstract

The present application relates to the technical field of data processing, and discloses a method and system for inspecting a packing device. The method includes: collecting packaging process parameters through multiple sensors to form a working condition data set; extracting abnormal features by spectrum analysis to generate an operation deviation index; establishing an association rule by comparing quality and status data to obtain a fault precursor feature library; analyzing the station coordination by time series comparison to form a component status evaluation table; calculating the fault probability of key components to generate an inspection point sequence; formulating an inspection plan including parameters, standards, and cycles to form a preventive maintenance plan. By analyzing key parameters such as tension, temperature, and speed in the packaging process, the present application identifies the early features of abnormal operation states of the device, and determines the inspection priority order and maintenance cycle in combination with historical fault data and component status evaluation, realizing the transformation from passive maintenance to active prevention.

Description

Technical Field

[0001] This application relates to the technical field of data processing, and particularly to a method and system for inspecting and maintaining a packing device. Background Art

[0002] In the traditional maintenance and management of packing devices, equipment inspection and maintenance are usually carried out by combining regular maintenance and post-fault repair. Regular maintenance is to inspect and maintain the equipment at fixed time intervals, usually setting the maintenance cycle based on the maintenance manual provided by the equipment manufacturer and empirical values; post-fault repair is an emergency repair measure after the equipment has failed or its performance has significantly declined. This traditional inspection method can ensure the basic operation of the equipment to a certain extent, but lacks pertinence and foresight. With the development of automation technology, some packing devices have begun to adopt digital monitoring systems, which collect equipment operation data by installing various sensors and use simple threshold judgment methods for abnormal monitoring. This method has been improved compared with pure manual inspection and can detect equipment abnormalities more timely.

[0003] However, the existing inspection methods for packing devices have many deficiencies: First, traditional regular maintenance is often carried out at fixed intervals without considering the actual usage conditions and environmental factors of the equipment, resulting in waste of resources due to overly frequent maintenance in some cases and missing the best opportunity for fault prevention due to untimely maintenance in other cases; Second, although the existing digital monitoring systems can collect data, the analysis methods are relatively single, mainly relying on simple threshold judgment and unable to identify complex fault patterns and early signs; Third, during the inspection process, the components of the equipment are often treated in isolation, lacking systematic analysis and ignoring the mutual influence and fault propagation relationship between components; Finally, the allocation of inspection resources is not reasonable enough, without prioritizing according to the importance of components and the risk of failure, resulting in key components possibly not receiving enough attention while non-key components consume too much inspection resources. Summary of the Invention

[0004] This application provides a method and system for inspecting and maintaining a packing device, which is used to realize the early prediction and accurate positioning of equipment faults, thereby improving equipment reliability and maintenance efficiency and reducing the loss of unexpected shutdowns.

[0005] In a first aspect, the present application provides a method for inspecting a packaging device. The method for inspecting the packaging device includes: collecting packaging process parameters through a tension sensor, a sealing temperature probe, and a conveyor belt pulse encoder on the packaging device to form a comprehensive dataset of the operating conditions of the packaging device; according to the comprehensive dataset of the operating conditions, using spectral analysis to extract features such as abnormal consumption of packaging materials, fluctuations in the defective sealing rate, and uneven speed, and generating an operating deviation index for the packaging device; based on the operating deviation index, establishing an association rule by comparing historical data of product quality and device state parameters to obtain a fault precursor feature library for the packaging device; according to the fault precursor feature library, using a time series comparison method to analyze the action coordination between each station on the packaging line to form a status evaluation table for key components of the packaging device; according to the status evaluation table of the key components, combining historical maintenance records to calculate the failure probabilities of the cutting knife, heat seal strip, transmission gear, and motor bearing, and generating a priority sequence for inspection points of the packaging device; according to the priority sequence of the inspection points, formulating an inspection operation guidance plan including detection parameters, judgment criteria, and inspection periods to form a preventive maintenance plan for the packaging device.

[0006] In a second aspect, the present application provides a system for inspecting a packaging device. The system for inspecting the packaging device includes:

[0007] An acquisition module, configured to collect packaging process parameters through a tension sensor, a sealing temperature probe, and a conveyor belt pulse encoder on the packaging device to form a comprehensive dataset of the operating conditions of the packaging device;

[0008] A generation module, configured to, according to the comprehensive dataset of the operating conditions, use spectral analysis to extract features such as abnormal consumption of packaging materials, fluctuations in the defective sealing rate, and uneven speed, and generate an operating deviation index for the packaging device;

[0009] An establishment module, configured to, based on the operating deviation index, establish an association rule by comparing historical data of product quality and device state parameters to obtain a fault precursor feature library for the packaging device;

[0010] An analysis module, configured to, according to the fault precursor feature library, use a time series comparison method to analyze the action coordination between each station on the packaging line to form a status evaluation table for key components of the packaging device;

[0011] A calculation module, configured to, according to the status evaluation table of the key components, combine historical maintenance records to calculate the failure probabilities of the cutting knife, heat seal strip, transmission gear, and motor bearing, and generate a priority sequence for inspection points of the packaging device;

[0012] A maintenance module, configured to, according to the priority sequence of the inspection points, formulate an inspection operation guidance plan including detection parameters, judgment criteria, and inspection periods to form a preventive maintenance plan for the packaging device.

[0013] The third aspect of the present invention provides a computer device, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor invokes the instructions in the memory to cause the computer device to execute the above-mentioned inspection method for packaging equipment.

[0014] The fourth aspect of the present invention provides a computer-readable storage medium, wherein instructions are stored in the computer-readable storage medium, and when the instructions are run on a computer, the computer is caused to execute the above-mentioned inspection method for packaging equipment.

[0015] In the technical solution provided by this application, the packaging process parameters are collected through a tension sensor, a sealing temperature probe, and a conveyor belt pulse encoder to form a comprehensive dataset of working conditions. The spectrum analysis is used to extract the characteristics of abnormal consumption of packaging materials, fluctuations in the defective sealing rate, and uneven speed, so as to accurately identify the operation deviation of the equipment at an early stage, improving the timeliness and accuracy of abnormal detection compared with the traditional regular inspection method. At the same time, by comparing the historical data of product quality and equipment status parameters to establish association rules, a fault precursor feature library of the packaging equipment is obtained, realizing the correlation analysis between the equipment status and product quality, and providing a reliable basis for abnormal status judgment. In addition, the time series comparison method is used to analyze the action coordination between each station on the packaging line to form a status evaluation table for key components of the packaging equipment, overcoming the defect of the traditional method that only focuses on individual components and ignores system coordination, and being able to discover potential problems caused by improper cooperation between stations. Moreover, by combining historical maintenance records to calculate the failure probabilities of cutting knives, heat sealing strips, transmission gears, and motor bearings, a priority sequence of inspection points is generated, realizing the reasonable allocation of inspection resources and avoiding the problems of waste of inspection resources and insufficient maintenance of key components in the traditional method. Finally, by formulating an inspection operation guidance plan including detection parameters, judgment criteria, and inspection cycles to form a preventive maintenance plan, the inspection work is standardized and standardized, improving the maintenance efficiency and equipment reliability. The present invention applies artificial intelligence algorithms to the field of packaging equipment maintenance. By processing multi-source sensor data through advanced algorithms such as multi-scale wavelet decomposition, phase space reconstruction, Lyapunov exponent, and non-linear detrending analysis, complex data with weak correlation is deeply mined, enabling the system to extract valuable fault features from noisy background signals and establishing a fault prediction model through machine learning algorithms, realizing the transformation from traditional empirical judgment to data-driven decision-making. Especially when dealing with the correlation analysis between equipment status and product quality, lag analysis and time series comparison techniques are applied to solve the technical problem that it is difficult for traditional methods to accurately capture the impact of equipment status on product quality, providing a mathematical basis for accurately predicting the time of fault occurrence. At the same time, in the link of determining the inspection priority, by comprehensively considering multi-dimensional factors such as the remaining life of components, failure risk, and replacement cost, a decision-making model is established, breaking the limitation of traditional fixed-cycle inspection and realizing a risk-based differentiated inspection strategy, significantly improving the utilization efficiency of maintenance resources and the operation reliability of the equipment, and achieving the goals of reducing equipment failure rate, improving packaging quality, and optimizing maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 Schematic diagram of an embodiment of the inspection method for the packaging equipment in the embodiments of the present application;

[0018] Figure 2 Schematic diagram of an embodiment of the inspection system for the packaging equipment in the embodiments of the present application;

[0019] Figure 3 It is a schematic block diagram of the structure of the computer device in the embodiments of the present invention. Specific embodiments

[0020] The embodiments of the present application provide a method and a system for inspecting a packaging equipment. Terms such as "first", "second", "third", "fourth", etc. (if any) in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments described here can be implemented in an order different from that shown or described here. In addition, the terms "comprising" or "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0021] For ease of understanding, the following describes the specific process of the embodiments of the present application. Please refer to Figure 1 An embodiment of the inspection method for the packaging equipment in the embodiments of the present application includes:

[0022] Step S101: Collect packaging process parameters through the tension sensor, sealing temperature probe, and conveyor belt pulse encoder on the packaging equipment to form a comprehensive data set of the operating conditions of the packaging equipment;

[0023] Step S102: According to the comprehensive data set of the operating conditions, use spectrum analysis to extract the characteristics of abnormal consumption of packaging materials, fluctuations in the sealing defect rate, and uneven speed, and generate an operating deviation index for the packaging equipment;

[0024] Step S103: Based on the operating deviation index, establish an association rule by comparing the historical data of the product quality and the equipment status parameters to obtain a fault precursor feature library for the packaging equipment;

[0025] Step S104: According to the fault precursor feature library, adopt a time series comparison method to analyze the action coordination between each station on the packaging line to form a status evaluation table for the key components of the packaging equipment;

[0026] Step S105: According to the key component status evaluation form, calculate the failure probabilities of the cutting knife, heat seal strip, transmission gear, and motor bearing in combination with the historical maintenance records, and generate the priority sequence of the inspection points for the packaging equipment.

[0027] Step S106: Based on the priority sequence of the inspection points, formulate an inspection operation guidance plan including inspection parameters, judgment criteria, and inspection periods, and form a preventive maintenance plan for the packaging equipment.

[0028] It can be understood that the execution subject of this application can be the inspection system for the packaging equipment, or it can also be a terminal or a server, and specific limitations are not made here. In this embodiment of the application, the server is used as the execution subject for illustration.

[0029] Specifically, the packaging process parameters are collected by the tension sensor, sealing temperature probe, and conveyor belt pulse encoder on the packaging equipment to form a comprehensive dataset of working conditions. Among them, the tension sensor continuously monitors the tension change of the packaging material during transportation, such as the tensile force change of the packaging film between the winding wheel and the feeding wheel; the sealing temperature probe records the temperature fluctuation of the heat seal strip during the sealing process in real time; and the conveyor belt pulse encoder records the change signal of the packaging line speed. During the collection process, multi-point sampling is performed on these signals. For example, the tension data is collected at a frequency of 200 Hz per second, the temperature data is collected at a frequency of 50 Hz per second, and the speed data is collected at a frequency of 100 Hz per second. Then, the working units are divided according to the operating cycle of the packaging equipment. For example, if each packaging cycle is 2 seconds, each working unit contains 400 tension data points, 100 temperature data points, and 200 speed data points. Then, the key time nodes in the complete packaging cycle are identified, such as the start time point of material loading, the start time point of heat sealing, the action time point of the cutting knife, etc., and the multi-sensor data is phase-aligned according to these time nodes to eliminate the difference in signal transmission delay and ensure that the data of different sensors can be accurately corresponding in time. After that, background interference is removed from the phase-aligned signal by the wavelet denoising method to extract the effective working condition signal. Wavelet denoising is to decompose the signal into different frequency components by wavelet transform, and then perform threshold processing on the noise components to reconstruct the denoised signal. A three-dimensional working condition feature space is constructed, with tension-temperature-speed as the coordinate axes to map the operating state of the equipment, and data point density analysis is performed in this space to form a comprehensive dataset of the working conditions of the packaging equipment. According to the comprehensive dataset of working conditions, spectrum analysis is used to extract the abnormal consumption of packaging materials, the fluctuation of the defective sealing rate, and the uneven speed characteristics, and an operation deviation index is generated. The tension signal in the comprehensive dataset of working conditions is subjected to multi-scale wavelet decomposition to construct a tension fluctuation characteristic tree spectrum. Multi-scale wavelet decomposition is to decompose the signal at different scales to obtain the signal characteristics in different frequency ranges. The abnormal energy distribution region is extracted from the tension fluctuation characteristic tree spectrum, and the material tension peak deviation ratio is calculated to form an identification fingerprint for abnormal consumption of packaging materials. The tension peak deviation ratio refers to the difference ratio between the actual tension peak and the standard tension peak, which reflects whether the consumption of packaging materials is abnormal. Then, the phase space reconstruction is performed on the sealing temperature data sequence, and the temperature trajectory attractor diagram is drawn. Phase space reconstruction is to convert a one-dimensional time series into a trajectory in a multi-dimensional phase space for analyzing the characteristics of a non-linear dynamic system. By calculating the Lyapunov exponent of the temperature trajectory attractor diagram, the stability of the sealing system is quantified, and an early warning index for the fluctuation of the defective sealing rate is established. The Lyapunov exponent is an index to measure the sensitivity of the system to the initial conditions, and the larger the value, the more unstable the system. Then, non-linear detrending analysis is used for the conveyor belt pulse sequence to extract the hidden pattern of tiny speed fluctuations. Non-linear detrending analysis is a method to remove the influence of the long-term trend in the signal and highlight the short-term fluctuation characteristics.Perform cross-correlation analysis on the micro velocity fluctuation hiding mode and the characteristic mode during standard operation to form a velocity non-uniformity characteristic spectrum. Through the multi-dimensional characteristic space mapping algorithm, integrate the abnormal identification fingerprint of packaging material consumption, the warning index of the fluctuation of the defective sealing rate, and the velocity non-uniformity characteristic spectrum into the operation deviation index of the packaging equipment. Based on the operation deviation index, establish association rules by comparing the historical data of product quality and equipment state parameters to obtain the fault precursor feature library of the packaging equipment. Extract the packaging integrity score, the sealing strength test value, and the appearance qualification rate data from the product quality inspection database to construct the product quality historical record matrix. The packaging integrity score is a score for the packaging sealing performance; the sealing strength test value is the numerical value of the sealing strength obtained through tensile testing; the appearance qualification rate refers to the proportion of products with qualified appearance inspections. Pair the product quality historical record matrix with the operation deviation index corresponding to time to form a quality-equipment state two-factor data chain, connecting the product quality index and the equipment state index. Conduct lag analysis on the quality-equipment state two-factor data chain to determine the time window from the occurrence of operation deviation to the decline of product quality, forming a state-quality influence cycle diagram. Lag analysis is to study the time delay relationship in which one variable changes following the change of another variable. Group through the state-quality influence cycle diagram to distinguish the development characteristics of different fault types and draw the fault evolution trajectory curve. Extract the key inflection points and threshold features from the fault evolution trajectory curve and mark the early warning signal points. The key inflection point is the turning point where the change accelerates during the fault development process, and the threshold feature is the critical value for triggering the warning. Associate and label the early warning signal points with the specific fault types in the historical maintenance records to obtain the fault precursor feature library of the packaging equipment.

[0030] According to the fault precursor feature library, the time series comparison method is used to analyze the action coordination among various workstations on the packaging line, and a status evaluation form for key components of the packing equipment is formed. The action time series signals of the feeding station, folding station, sealing station, cutting station, and output station are collected from the control system of the packing equipment to construct the action time series of the packing stations. The feeding station is responsible for feeding the packaging materials; the folding station is responsible for folding the packaging materials into shape; the sealing station is responsible for heat-sealing the edges of the packaging; the cutting station is responsible for separating individual packages; the output station is responsible for outputting the finished products. The action time series of the packing stations is phase-synchronized with the standard time series template in the fault precursor feature library to generate a map of action deviation between stations. Phase synchronization is the process of aligning two time series according to the key event points. Based on the map of action deviation between stations, the action conversion delay value between adjacent stations is extracted to form a metric for abnormal station connection. The action conversion delay value refers to the time interval from the end of the action of one station to the start of the action of the next station. The metric for abnormal station connection is decomposed to identify the mechanism sources causing the delay, and the key nodes of the drive chain between stations are marked. The motion characteristic changes of each key component are analyzed through the key nodes of the drive chain between stations, and the deviation degree of the component operation state is calculated. The deviation degree of the component operation state is the degree of difference between the actual operation state and the ideal state. The deviation degree of the component operation state is compared with the component health threshold to form a status evaluation form for key components of the packing equipment.

[0031] According to the key component status evaluation form, calculate the failure probabilities of the cutting knife, heat sealing strip, transmission gear, and motor bearing by combining historical maintenance records, and generate the priority sequence of inspection points for the packaging equipment. Extract the historical failure records of the cutting knife, heat sealing strip, transmission gear, and motor bearing from the equipment management system, and construct a distribution map of component failure time. The cutting knife is a device responsible for cutting the packaging material; the heat sealing strip is a heating element that melts the edges of the packaging material to form a seal; the transmission gear is a mechanical component that transmits power; the motor bearing is a component that supports the rotating part of the motor. Compare the component failure time distribution map with the current state parameters in the key component status evaluation form, and calculate the predicted remaining life value of the component. The predicted remaining life value of the component is the time that the component can still work normally estimated based on historical data and the current state. Based on the predicted remaining life value of the component and the component state deterioration rate, form the component failure risk coefficient. The component state deterioration rate is the speed at which the component performance declines. Perform a weighted treatment on the component failure risk coefficient, and the weight factors include component replacement cost, downtime impact degree, and spare part availability, to generate the comprehensive impact score of component failure. Divide the inspection urgency level according to the comprehensive impact score of component failure, and it is divided into three inspection levels: daily inspection items, key attention items, and emergency treatment items. Arrange the components of different inspection levels in the order of the production line layout to generate the priority sequence of inspection points for the packaging equipment. Based on the priority sequence of inspection points, formulate an inspection operation guidance plan including detection parameters, judgment criteria, and inspection cycles, and form a preventive maintenance plan for the packaging equipment. For each component in the priority sequence of inspection points, extract a dedicated set of detection parameters from the standard detection specification library, including the cutting sharpness detection value of the cutting knife, the heat sealing strip temperature uniformity index, the transmission gear meshing clearance range, and the motor bearing vibration spectrum threshold. The cutting sharpness detection value of the cutting knife is an index to measure the cutting ability of the cutting knife; the heat sealing strip temperature uniformity index is a measure of the consistency of the heat sealing strip temperature distribution; the transmission gear meshing clearance range is the interval range between meshing gears; the motor bearing vibration spectrum threshold is the normal limit value of the bearing vibration frequency distribution. Associate the set of detection parameters with the abnormal threshold in the failure precursor feature library to construct a component state determination standard table. Set different inspection cycles based on the component failure risk coefficient to generate a component inspection time interval matrix. Integrate the set of detection parameters, the determination standard table, and the inspection time interval matrix into a component inspection operation card. According to the production plan arrangement and the maintenance resource allocation situation, schedule the time of the component inspection operation card to generate an inspection execution calendar. Combine the component inspection operation card with the inspection execution calendar to generate a preventive maintenance plan for the packaging equipment.

[0032] Taking the candy packaging line of a certain food production enterprise as an example, the tension of the packaging film is collected by a tension sensor installed on the film roll shaft and fluctuates within the normal value range of 1.5N ± 0.1N. The temperature probe of the heat seal bar monitors that the temperature of the heat seal bar changes within the set value range of 185°C ± 2°C, and the linear velocity measured by the conveyor belt pulse encoder maintains at 60 products / minute ± 1 product / minute. After data processing, it is found that the Lyapunov exponent of the heat seal bar temperature fluctuation increases from the normal value of 0.05 to 0.12, indicating a decrease in system stability. By matching with the established fault precursor feature library, a similar situation usually indicates poor contact of the heat seal bar. Further time series comparison analysis shows that the action conversion delay from the sealing station to the cutting station increases by 15ms, exceeding the normal range, and it is concluded that the deviation degree of the heat seal bar working state reaches 68%, exceeding the health threshold of 60%. Combining historical data, the fault risk coefficient of the heat seal bar is calculated to be 0.78, and the comprehensive fault impact score is 85 points, belonging to the key attention items. Therefore, in the preventive maintenance plan, the inspection cycle of the heat seal bar is adjusted from once a week to once a day, and it is specified that the detected temperature uniformity shall not exceed ±3°C, and the contact pressure shall be between 0.4MPa and 0.6MPa to ensure the packaging sealing quality.

[0033] In the embodiments of the present application, process parameters of the packaging process are collected through a tension sensor, a sealing temperature probe, and a conveyor belt pulse encoder to form a comprehensive dataset of working conditions. Spectral analysis is used to extract features such as abnormal consumption of packaging materials, fluctuations in the defective sealing rate, and uneven speed, so as to accurately identify equipment operation deviations at an early stage, improving the timeliness and accuracy of anomaly detection compared with traditional regular inspection methods. At the same time, by comparing historical data of product quality and equipment status parameters to establish association rules, a fault precursor feature library of the packaging equipment is obtained, realizing the correlation analysis between equipment status and product quality and providing a reliable basis for anomaly status judgment. In addition, a time series comparison method is used to analyze the action coordination between each station on the packaging line to form a status evaluation table of key components of the packaging equipment, overcoming the defect of traditional methods that only focus on individual components and ignore system coordination, and being able to discover potential problems caused by improper cooperation between stations. Furthermore, by combining historical maintenance records to calculate the failure probabilities of cutting knives, heat sealing strips, transmission gears, and motor bearings, a priority sequence of inspection points is generated, realizing the reasonable allocation of inspection resources and avoiding the problems of waste of inspection resources and insufficient maintenance of key components in traditional methods. Finally, by formulating an inspection operation guidance plan including detection parameters, judgment criteria, and inspection cycles, a preventive maintenance plan is formed, standardizing and standardizing the inspection work and improving the maintenance efficiency and equipment reliability. The present invention applies artificial intelligence algorithms to the field of packaging equipment maintenance. By processing multi-source sensor data through advanced algorithms such as multi-scale wavelet decomposition, phase space reconstruction, Lyapunov exponent, and non-linear detrending analysis, complex data with low correlation is deeply mined, enabling the system to extract valuable fault features from noisy background signals and establishing a fault prediction model through machine learning algorithms, realizing the transformation from traditional experience-based judgment to data-driven decision-making. Especially when dealing with the correlation analysis between equipment status and product quality, lag analysis and time series comparison techniques are applied to solve the technical problem that traditional methods are difficult to accurately capture the impact of equipment status on product quality, providing a mathematical basis for accurately predicting the time of fault occurrence. At the same time, in the link of determining the inspection priority, by comprehensively considering multi-dimensional factors such as the remaining life of components, fault risk, and replacement cost, a decision-making model is established, breaking the limitation of traditional fixed-cycle inspections, realizing a risk-based differential inspection strategy, significantly improving the utilization efficiency of maintenance resources and the operation reliability of equipment, and achieving the goals of reducing equipment failure rate, improving packaging quality, and optimizing maintenance costs.

[0034] In a specific embodiment, the process of executing step S101 may specifically include the following steps:

[0035] (1) Perform multi-point sampling on the instantaneous tension waveform of the tension sensor during the conveying process of the packaging material, the temperature curve of the sealing temperature probe during the heat sealing process, and the speed change signal of the conveyor belt pulse encoder.

[0036] (2)Divide the sampling signal into working units according to the operating cycle of the packing device, and identify the key time nodes in the complete packing cycle;

[0037] (3)Align the phases of the multi-sensor data according to the key time nodes to eliminate the signal transmission delay difference;

[0038] (4)Remove the background interference from the phase-aligned signal by wavelet denoising method, and extract the effective working condition signal;

[0039] (5)Construct a three-dimensional working condition feature space, and map the operating state of the device with tension-temperature-speed as the coordinate axes;

[0040] (6)Conduct data point density analysis in the three-dimensional working condition feature space to form a comprehensive data set of the working conditions of the packing device.

[0041] Specifically, multi-point sampling is performed on the instantaneous tension waveform of the tension sensor during the conveying process of the packaging material, the temperature curve of the sealing temperature probe during the heat sealing process, and the speed change signal of the conveyor belt pulse encoder. Multi-point sampling refers to the process of discretely collecting continuous signals at certain frequency intervals on the time axis. Specifically, the sampling frequency of the tension sensor is set to 500 Hz, that is, 500 data points are collected per second, and the minute changes in the tension of the packaging material are recorded in real time; the sampling frequency of the sealing temperature probe is 100 Hz to capture the dynamic changes in the temperature of the heat sealing strip; the conveyor belt pulse encoder collects the speed signal at a frequency of 1000 Hz to accurately record the fluctuations in the operating speed of the device. Through this high-frequency sampling method, the subtle change characteristics during the operation of the device can be obtained.

[0042] Divide the sampling signal into working units according to the operating cycle of the packing device, and identify the key time nodes in the complete packaging cycle. A working unit refers to the time period during which the packing device completes a complete packaging action, usually bounded by the starting point of product packaging. The key time nodes include the start time of material feeding, the start time of heat sealing, the cutting action time, etc. The method of identifying these time nodes is to analyze the mutation points in the signal characteristic curve. For example, a sudden rise point in the tension signal indicates the start of feeding, a rapid rise point in the temperature signal indicates the start of heat sealing, and a short pause point in the speed signal indicates the cutting action. By setting a threshold detection algorithm, when the signal change rate exceeds the preset threshold, it is determined as a key time node. For example, when the tension change rate exceeds 0.5 N / ms, it is marked as the material feeding point; when the temperature rise rate exceeds 10 °C / s, it is marked as the heat sealing start point. In this way, the continuous signal data is divided into independent working units, which is convenient for subsequent analysis. Then, phase alignment of the multi-sensor data is performed according to the key time nodes to eliminate the signal transmission delay difference. Phase alignment is the process of corresponding the data points collected by different sensors at the same logical moment, which solves the time delay problem caused by different physical positions of sensors and different signal transmission paths. The specific implementation method is to use a certain key time node as a reference point and perform time translation alignment on the data of other sensors. For example, taking the heat sealing start point as a reference, the tension signal and the speed signal are respectively translated by Δt1 and Δt2 to align the three groups of data on the time axis. Δt1 and Δt2 are fixed delay values determined through experiments, which reflect the physical delay of signal transmission. After phase alignment, the data of the three sensors at the same logical moment can be directly corresponding for analysis.

[0043] Subsequently, background interference is removed from the phase-aligned signal by wavelet denoising method to extract the effective working condition signal. Wavelet denoising is a signal processing method based on wavelet transform, which can effectively remove noise while retaining the main features of the signal. The specific steps include: selecting an appropriate wavelet basis function, such as Daubechies wavelet or Symlet wavelet; performing multi-scale wavelet decomposition on the signal to obtain wavelet coefficients of different frequencies; performing threshold processing on the wavelet coefficients, setting the coefficients smaller than the threshold to zero; and performing wavelet reconstruction to obtain the denoised signal. In the signal processing of the packing device, 3-level Daubechies4 wavelet decomposition is selected, and the wavelet coefficients are processed by the soft threshold method, and the threshold value is taken as 3 times the standard deviation of the wavelet coefficients. This processing method can effectively remove random noise and electromagnetic interference and retain the important features in the signal. Compared with the original signal, the signal-to-noise ratio of the denoised tension signal, temperature signal and speed signal is increased by about 10 dB.

[0044] Construct a three-dimensional operating condition feature space, and map the operating state of the equipment with tension-temperature-speed as the coordinate axes. The three-dimensional operating condition feature space is an Euclidean space with three key parameters as the coordinate axes, and each point represents the operating state of the equipment at a specific moment. The specific construction method is as follows: normalize the denoised signal to unify signals with three different dimensions to the interval [0,1]; then establish a three-dimensional coordinate system with the normalized tension value as the x-axis, the temperature value as the y-axis, and the speed value as the z-axis; map the data points at the same moment after phase alignment into this coordinate system to form the trajectory of the equipment operating state. Through mapping, the operating state of the equipment is represented as a trajectory in three-dimensional space, and the trajectory under normal working conditions usually has stable morphological characteristics. Conduct data point density analysis in the three-dimensional operating condition feature space to form a comprehensive dataset of the operating conditions of the packaging equipment. Data point density analysis is a method for statistically analyzing the distribution density of data points in different regions of three-dimensional space, used to identify the main distribution regions and abnormal regions of the equipment operating state. Specifically, divide the three-dimensional space into grid cells of equal size; then count the number of data points contained in each grid cell; determine the main distribution regions and deviation degrees of the equipment operating state according to the number of data points. The size of the grid cell is determined according to the data accuracy requirements, and usually selected so that the main distribution region under normal operating conditions contains about 10-20 grid cells. By calculating the statistical characteristics of the data points in each grid cell, such as mean, variance, skewness, kurtosis, etc., further quantify the stability and abnormality degree of the equipment operating state, and form a comprehensive dataset of operating conditions containing multi-dimensional operating characteristics of the equipment.

[0045] Taking a food packaging production line as an example, when the packaging equipment is running, the tension sensor records that the tension of the packaging film fluctuates between 0.8N and 1.2N, and the sampling frequency is 500Hz; the temperature probe monitors that the temperature of the heat seal bar changes between 175°C and 180°C, and the sampling frequency is 100Hz; the pulse encoder shows that the running speed of the equipment is between 35 and 37 packages per minute, and the sampling frequency is 1000Hz. Through signal analysis, it is found that a packaging cycle is completed every 2.5 seconds. Taking the sharp rise point of the temperature signal as the heat seal start time node, at the same time, the tension signal shows an instantaneous drop and the speed signal is briefly stable. Taking the heat seal start point as the reference point, the tension signal is shifted forward by 15ms and the speed signal is shifted backward by 8ms to achieve phase alignment. The aligned signal is denoised by 3-level Daubechies4 wavelet decomposition to filter out environmental vibration and electrical interference. Normalize the denoised tension, temperature, and speed signals to the interval [0,1], construct a three-dimensional feature space, and find that the data points in the normal operating state are mainly concentrated in the central region of the feature space to form a stable ellipsoidal distribution. When foreign objects adhere to the heat seal bar, the data points will shift in the positive direction of the temperature axis to form a local high-density area. This feature is marked as an early indicator of heat seal abnormality, providing an effective basis for equipment inspection.

[0046] In a specific embodiment, the process of executing step S102 may specifically include the following steps:

[0047] (1) Perform multi-scale wavelet decomposition on the tension signal in the comprehensive working condition dataset to construct a tension fluctuation characteristic tree spectrum;

[0048] (2) Extract the energy distribution abnormal region from the tension fluctuation characteristic tree spectrum, calculate the peak deviation ratio of the material tension, and form an abnormal consumption identification fingerprint of the packaging material;

[0049] (3) Perform phase space reconstruction on the sealing temperature data sequence and draw a temperature trajectory attractor graph;

[0050] (4) Quantify the stability of the sealing system by calculating the Lyapunov exponent of the temperature trajectory attractor graph, and establish a warning index for the fluctuation of the defective sealing rate;

[0051] (5) Adopt non-linear detrending analysis for the conveyor belt pulse sequence to extract the hidden mode of minute speed fluctuations;

[0052] (6) Perform cross-correlation analysis on the hidden mode of minute speed fluctuations and the characteristic mode during standard operation to form a speed non-uniformity characteristic spectrum;

[0053] (7) Integrate the abnormal consumption identification fingerprint of the packaging material, the warning index for the fluctuation of the defective sealing rate, and the speed non-uniformity characteristic spectrum into an operation deviation index of the packing equipment through a multi-dimensional feature space mapping algorithm.

[0054] Specifically, perform multi-scale wavelet decomposition on the tension signals in the comprehensive working condition dataset to construct a tension fluctuation characteristic tree spectrum. Multi-scale wavelet decomposition is a time-frequency analysis technique that can decompose signals at different scales and extract features in different frequency ranges. The specific implementation steps are as follows: Select a wavelet basis function suitable for the characteristics of the tension signal, such as the Daubechies wavelet basis or the Morlet wavelet basis; Pass the tension signal through high-pass and low-pass filters in sequence to obtain detail coefficients and approximation coefficients; Continue to decompose the approximation coefficients to obtain the detail coefficients and approximation coefficients of the next level; Repeat the above steps until the predetermined decomposition level. Usually, a decomposition depth of 5 layers is selected, which can cover the characteristics of each frequency band from high-frequency noise to low-frequency trends. After decomposition, the formed tension fluctuation characteristic tree spectrum contains detail coefficient matrices of different frequency levels and an approximation coefficient matrix of the lowest frequency. These coefficient matrices jointly describe the complete characteristics of the tension signal in the time-frequency domain. Extract the energy distribution abnormal area from the tension fluctuation characteristic tree spectrum, calculate the ratio of the peak deviation of the material tension, and form an abnormal consumption identification fingerprint of the packaging material. The energy distribution abnormal area refers to the area in certain time-frequency intervals of the characteristic tree spectrum where the signal energy significantly deviates from the normal operating state. The extraction method is as follows: Calculate the energy distribution of the detail coefficients at each level of the characteristic tree spectrum, that is, the sum of the squares of each coefficient; Compare it with the standard energy distribution during normal operation established in advance; When the energy deviation in a certain time-frequency interval exceeds the set threshold, mark it as an abnormal area. The ratio of the peak deviation of the material tension is defined as the difference between the ratio of the actual tension peak to the standard tension peak and 1, which reflects the abnormal degree of the packaging material tension. Combine the detected abnormal features at each level according to their importance to form an abnormal consumption identification fingerprint of the packaging material. This fingerprint can accurately reflect different types of material consumption abnormalities, such as excessive stretching and insufficient relaxation.

[0055] Perform phase space reconstruction on the sealing temperature data sequence and draw a temperature trajectory attractor graph. Phase space reconstruction is a method to restore the dynamic characteristics of a system from a one-dimensional time series. Based on the embedding theorem, the one-dimensional time series is mapped into a high-dimensional phase space. The specific implementation steps are as follows: Determine the embedding dimension m and the time delay τ; Construct an m-dimensional vector sequence, where each vector consists of m points with an interval of τ in the original time series; Plot these vector points in the m-dimensional phase space to form the system state trajectory. The embedding dimension is usually determined by the false nearest neighbor method, and the time delay is determined by the mutual information method or the autocorrelation function. For the sealing temperature data, typical embedding dimensions are 3 or 4, and the time delay is 10 - 20 sampling points. In the reconstructed phase space, the temperature change trajectory forms a specific geometric shape, called the temperature trajectory attractor graph, which reflects the inherent dynamic characteristics of the sealing temperature system.

[0056] By calculating the Lyapunov exponent of the temperature trajectory attractor diagram, the stability of the sealing system is quantified, and an early warning index for the fluctuation of the defective sealing rate is established. The Lyapunov exponent is a quantitative index that describes the separation rate of adjacent trajectories in a dynamic system. A positive value indicates that the system has chaotic characteristics, and a negative value indicates that the system is stable. The calculation formula of the Lyapunov exponent is as follows:

[0057]

[0058] Among them, represents the Lyapunov exponent, is the number of iterations, is the sampling time interval, represents the initial distance of the i-th adjacent trajectory point pair in the phase space, represents the distance after one iteration. According to the correlation analysis between the Lyapunov exponent and the historical defective sealing rate, an early warning index for the fluctuation of the defective sealing rate (SDWI) is established:

[0059]

[0060] Among them, is the early warning index for the fluctuation of the defective sealing rate, , , , are the fitting coefficients, is the Lyapunov exponent, is the temperature coefficient of variation. When exceeds the preset threshold, an early warning for defective sealing is triggered. Nonlinear detrending analysis is used for the conveyor belt pulse sequence to extract the hidden pattern of micro-velocity fluctuations. Nonlinear detrending analysis is a method to eliminate the long-term trend of the signal and highlight the characteristics of short-term fluctuations. The specific implementation steps are as follows: divide the pulse sequence into several non-overlapping segments; fit a multi-order polynomial trend curve for each segment of data; subtract the trend curve from the original data to obtain the detrended fluctuation sequence; perform statistical analysis on the detrended fluctuation sequence to extract the fluctuation characteristics. The polynomial order is usually selected as 3-5 orders, which can effectively capture the nonlinear trend. After removing the long-term trend, the hidden pattern of micro-velocity fluctuations will be more obvious, and these patterns often contain important information about the operating state of the equipment.

[0061] Perform cross-correlation analysis between the micro velocity fluctuation hiding pattern and the characteristic pattern during standard operation to form a velocity non-uniformity characteristic spectrum. Cross-correlation analysis is a technique for evaluating the similarity between two signals. By calculating the correlation coefficients at different time delays, the correlation between the signals is obtained. The specific implementation steps are as follows: calculate the correlation coefficients at different time delays between the extracted micro velocity fluctuation hiding pattern and the standard pattern; plot the curve of the correlation coefficient changing with the time delay; analyze the peak position, peak magnitude, and curve shape characteristics of the correlation curve. Cross-correlation analysis can effectively detect abnormal patterns in the device speed change, such as periodic fluctuations, sudden fluctuations, etc. These characteristics constitute the velocity non-uniformity characteristic spectrum, which directly reflects the health status of the device transmission system. Through the multi-dimensional feature space mapping algorithm, the abnormal identification fingerprint of packaging material consumption, the warning index of the fluctuation of the defective sealing rate, and the velocity non-uniformity characteristic spectrum are integrated into the operation deviation index of the packaging equipment. Multi-dimensional feature space mapping is a technique for unifying the representation and comprehensive evaluation of multiple heterogeneous features. The calculation formula of the operation deviation index (PODI) of the packaging equipment is as follows:

[0062]

[0063] Wherein, is the operation deviation index of the packaging equipment, , , are the weight coefficients of each feature, satisfying + + = 1, MPDF is the score of the abnormal identification fingerprint of material consumption, SDWI is the warning index of the fluctuation of the defective sealing rate, and SVSI is the index of the velocity non-uniformity characteristic spectrum. The weight coefficients are determined through historical data analysis and expert experience, reflecting the influence degree of each factor on the device state. The higher the PODI value, the greater the operation deviation of the device, and more attention is needed.

[0064] Taking a beverage packaging line as an example, when abnormal fluctuations in the tension signal are detected during the operation of the equipment, a tension fluctuation characteristic tree spectrum is constructed through 5-layer Daubechies4 wavelet decomposition. It is found that there are regions with abnormally increased energy in the detail coefficients of the 3rd and 4th layers, and the corresponding frequency range is 5 - 15 Hz, which is usually related to faults in the packaging film tension control system. The peak deviation ratio of the material tension is calculated to be 0.28, exceeding the normal range of 0.1, forming an abnormal consumption identification fingerprint of the packaging material. At the same time, phase space reconstruction is performed on the sealing temperature data. The embedding dimension is selected as 3, and the time delay is 15 sampling points. The temperature trajectory attractor graph is drawn, and it is found that the trajectory changes from the usual compact ring to a loose spiral. The Lyapunov exponent is calculated to be 0.15, significantly higher than the normal value of 0.05. Substituting this value into the warning index formula, the warning index for the fluctuation of the defective sealing rate is obtained as 0.68, approaching the warning threshold of 0.7. Nonlinear detrending analysis is performed on the conveyor belt pulse sequence, and the periodic fluctuation pattern is extracted. Cross-correlation analysis is performed with the standard pattern, and it is found that the peak of the correlation curve drops and the position shifts, indicating a decrease in the synchronization of the drive system. Combining these three indicators, the operation deviation index of the packaging equipment is calculated to be 0.58, exceeding the normal operation threshold of 0.4, indicating that the tension control system and the drive system need to be inspected and maintained to prevent possible packaging quality problems.

[0065] In a specific embodiment, the process of executing step S103 may specifically include the following steps:

[0066] (1) Extract the packaging integrity score, sealing strength test value, and appearance qualification rate data from the product quality inspection database to construct a product quality historical record matrix;

[0067] (2) Pair the product quality historical record matrix with the corresponding operation deviation index in time series to form a quality - equipment status two - factor data chain;

[0068] (3) Perform lag analysis on the quality - equipment status two - factor data chain to determine the time window from the occurrence of the operation deviation to the decline of the product quality, and form a status - quality influence period graph;

[0069] (4) Group through the status - quality influence period graph to distinguish the development characteristics of different fault types, and draw a fault evolution trajectory curve;

[0070] (5) Extract the key inflection points and threshold characteristics from the fault evolution trajectory curve, and mark the early warning signal points;

[0071] (6) Associate and label the early warning signal points with the specific fault types in the historical maintenance records to obtain a fault precursor feature library of the packaging equipment.

[0072] Specifically, extract the packaging integrity score, sealing strength test value, and appearance pass rate data from the product quality inspection database to construct a product quality history record matrix. The packaging integrity score is a quantitative evaluation of the packaging sealing performance through methods such as airtightness testing and water immersion testing, usually represented by 0-100 points; the sealing strength test value is the tensile strength of the sealing joint measured by a tensile tester when subjected to tensile force, with the unit of Newton (N); the appearance pass rate data refers to the proportion of the number of qualified products in the total number of inspected products during the product appearance inspection. These three indicators together reflect different aspects of the product packaging quality. The method of constructing the product quality history record matrix is to arrange these three indicators in a time series to form an n×4 matrix, where n is the number of historical data points, and the 4 columns are time, packaging integrity score, sealing strength test value, and appearance pass rate respectively. Each row in the matrix represents the product quality status at a certain time point, and this matrix structure facilitates subsequent temporal matching with the equipment status data. Pair the product quality history record matrix with the corresponding running deviation indicators in time series to form a quality-equipment status two-factor data chain. The running deviation indicators are generated in the previous steps and contain multi-dimensional information such as the abnormal identification fingerprint of packaging material consumption, the early warning index of the fluctuation of the sealing defect rate, and the uneven speed characteristic spectrum. The method of time series pairing is to correspond the product quality data and the equipment operation data one by one according to the time stamp. When the acquisition times of the product quality data and the equipment operation data are not exactly the same, the nearest time point matching or linear interpolation method is used for pairing. The quality-equipment status two-factor data chain formed after pairing is a comprehensive data structure containing product quality indicators and equipment operation status indicators. Each data point contains a time stamp, quality indicators (packaging integrity score, sealing strength test value, appearance pass rate), and equipment status indicators (running deviation indicators). This data chain structure can clearly reflect the correlation between product quality and equipment status.

[0073] Then, perform lag analysis on the quality-equipment status two-factor data chain to determine the time window from the occurrence of operation deviation to the decline of product quality, and form a status-quality influence cycle diagram. Lag analysis is a method for studying the time-delay relationship in which one variable changes after another in sequence data. The specific implementation steps are as follows: Shift the time of the equipment operation deviation index sequence to construct sequences with different lag times; Calculate the correlation coefficient between the operation deviation index and the product quality index at each lag time; Plot the curve of the correlation coefficient changing with the lag time, that is, the lag correlation diagram; Determine the lag time when the correlation coefficient reaches the maximum value or exceeds the significance threshold. Through this analysis, the time window in which the change of equipment status affects product quality can be determined, that is, the time required from equipment status anomaly to product quality decline. This time window is of great significance for predictive maintenance, as it determines the maximum allowable time from equipment anomaly detection to the necessary maintenance measures. The results of the lag analysis form a status-quality influence cycle diagram, which intuitively shows the time relationship in which different equipment status indicators affect different product quality indicators. Grouping is carried out through the status-quality influence cycle diagram to distinguish the development characteristics of different fault types, and the fault evolution trajectory curve is plotted. The grouping method is based on the time relationship characteristics in the status-quality influence cycle diagram, and historical fault cases are clustered according to similarity. Commonly used clustering algorithms include K-means clustering, hierarchical clustering, etc. The characteristics considered during clustering include: the change rate of equipment status indicators, the decline amplitude of product quality indicators, the length of the influence time window, etc. The clustering results represent different fault types or fault development modes. For each type of fault, select typical cases and plot the curves of equipment status indicators and product quality indicators changing with time from the beginning of equipment status anomaly to the final occurrence of the fault, that is, the fault evolution trajectory curve. The curve reflects the complete development process of different fault types from early signs to final failure.

[0074] Subsequently, extract the key inflection points and threshold characteristics from the fault evolution trajectory curve and mark the early warning signal points. The key inflection point refers to the point where the change trend in the fault evolution trajectory curve changes significantly, usually indicating that the fault development enters a new stage. The methods for extracting key inflection points include curve slope analysis method, curvature calculation method, etc. Specifically, smooth the trajectory curve to remove the influence of short-term fluctuations; Then calculate the first derivative (slope) and second derivative (curvature) of the curve; Identify the points with significant changes in slope or curvature by setting thresholds. The threshold characteristic refers to the characteristic point when the equipment status indicator or product quality indicator reaches a specific value, indicating that the equipment is in a specific fault development stage. These characteristic points are usually determined based on historical data statistics and expert experience. The key inflection points and threshold characteristics together constitute the early warning signal points, which are the indicators and values that need to be focused on in equipment status monitoring.

[0075] Associate the early warning signal points with the specific fault types in the historical maintenance records to obtain a precursor feature library of packaging equipment faults. The historical maintenance records contain information such as the specific fault types, fault locations, and maintenance measures when the equipment fails. The method of association annotation is to establish a corresponding relationship between each early warning signal point and the fault type in the historical maintenance records, forming an early warning rule of "if the equipment status index reaches the first threshold and the product quality index is lower than the second threshold, the equipment may be about to have a type P fault". These rules are statistically verified to ensure their early warning accuracy and timeliness. All early warning rules form a precursor feature library of packaging equipment faults, which is an important knowledge basis for equipment status monitoring and preventive maintenance.

[0076] Taking a certain biscuit packaging line as an example, by analyzing the production data of the past year, product quality data such as the packaging integrity score (out of 100), the sealing strength test value (unit: Newton), and the appearance qualification rate (percentage) were extracted from the quality inspection database, and a product quality historical record matrix containing 1095 rows of data for 365 days and 3 shifts per day was constructed. At the same time, the equipment operation deviation index data for the same period was collected, and these two sets of data were integrated into a quality-equipment status two-factor data chain through timestamp matching. By performing a lag analysis on the data chain, it was found that there is an 8-hour lag relationship between the heat seal temperature fluctuation index and the packaging integrity score, that is, after the heat seal temperature fluctuates abnormally, the packaging integrity score will significantly decrease after about 8 hours; there is a 4-hour lag relationship between the tension control deviation and the sealing strength test value; there is a 2-hour lag relationship between the uneven speed of the transmission system and the appearance qualification rate. According to these lag time characteristics, the historical fault cases were divided into three categories: heat seal system fault category, tension control system fault category, and transmission system fault category. For the heat seal system fault category, the trajectory curves of the heat seal temperature uniformity index, the sealing temperature fluctuation range, and the packaging integrity score changing with time were plotted. Key inflection points were identified from the trajectory curves: when the heat seal temperature uniformity index fluctuates from within the normal 3°C to above 5°C and lasts for more than 2 hours, it is the first early warning point; when the sealing temperature fluctuation range expands from the normal ±2°C to ±4°C and lasts for more than 1 hour, it is the second early warning point. Matching these early warning points with the historical maintenance records of 14 heat seal failures, it was found that a local wear fault of the heat seal will occur on average 12 hours after the first early warning point appears, and a failure of the heat seal temperature controller will occur on average 6 hours after the second early warning point appears.

[0077] In a specific embodiment, the process of executing step S104 may specifically include the following steps:

[0078] (1) Collect the action timing signals of the feeding station, folding station, sealing station, cutting station, and output station from the packaging equipment control system to construct a packaging station action time series;

[0079] (2) Phase-synchronize the time series of the packing station operation with the standard time series template in the fault precursor feature library to generate an operation deviation map of the station;

[0080] (3) Extract the operation conversion delay value between adjacent stations based on the operation deviation map of the station to form a measure of abnormal station connection;

[0081] (4) Decompose the measure of abnormal station connection to identify the mechanism source causing the delay and mark the key nodes of the drive chain between stations;

[0082] (5) Analyze the change in the motion characteristics of each key component through the key nodes of the drive chain between stations and calculate the deviation degree of the component operation state;

[0083] (6) Compare the deviation degree of the component operation state with the component health threshold to form a key component state evaluation table for the packing equipment.

[0084] Specifically, collect the action timing signals of the feeding station, folding station, sealing station, cutting station, and output station from the packaging equipment control system to construct the action time series of the packaging stations. The feeding station is the part responsible for transporting packaging materials into the equipment; the folding station is responsible for folding the packaging materials into a preset shape; the sealing station heat-seals or bonds the edges of the packaging; the cutting station divides the continuous packaging into independent units; the output station transports the completed packaging products out of the equipment. The action timing signal refers to the data record of the start time, duration, and end time of the actions performed by each station. The collection method is to install position sensors, photoelectric switches, or encoders on the key drive components of each station to record the precise time points of the actions of each station. For example, the rotation signal of the feeding roller at the feeding station, the pressing plate action signal at the folding station, the pressure application signal of the heat-sealing head at the sealing station, the cutting knife action signal at the cutting station, and the conveyor belt start signal at the output station. These signals are recorded in the form of timestamps and status values (0 or 1) to form the action time series of each station. Sort and merge the time series of each station according to time to form a complete action time series of the packaging stations, which completely records the timing relationship of the actions of each station during the entire packaging process. Synchronize the phase of the action time series of the packaging stations with the standard timing template in the fault precursor feature library to generate the action deviation map of the stations. The standard timing template is a typical action time series of the equipment under normal operating conditions selected from historical data. Phase synchronization refers to the process of aligning two time series according to a specific reference point to eliminate the influence of time offset and facilitate direct comparison of the series characteristics. The specific implementation steps include: determining the reference synchronization point, usually selecting the starting point of the packaging cycle, such as the feeding start moment of the feeding station; then aligning the currently collected time series and the standard template at the reference point; calculating the time deviation of the actions of each station at the corresponding time points. This deviation calculation uses a point-to-point comparison method, that is, for each corresponding point in the series, calculate the difference between its time value and the time value of the corresponding point in the standard template. Arrange these deviation values in chronological order and display them graphically to form the action deviation map of the stations. This map intuitively shows the timing difference between the current equipment operating state and the standard state, and different deviation patterns often correspond to different types of equipment anomalies.

[0085] Then, based on the station action deviation map, extract the action conversion delay values between adjacent stations to form the station connection anomaly measurement value. The action conversion delay value refers to the time interval between the end of one station action and the start of the next station action. Under normal operating conditions, these delay values should fluctuate within a certain range; when an abnormality occurs in a certain part of the equipment, the delay values between the corresponding stations will change significantly. The extraction method is as follows: identify the start and end time points of each station action from the station action deviation map; then calculate the time intervals between adjacent stations, such as the time from the end of the feeding station to the start of the folding station, the time from the end of the folding station to the start of the sealing station, etc.; compare these time intervals with the corresponding intervals in the standard template and calculate the deviation values. The station connection anomaly measurement value is defined as the deviation rate of the actual connection delay from the standard connection delay, and this indicator can directly reflect the degree of abnormality in the coordination between stations. Usually, a threshold is set, and when the anomaly measurement value exceeds the threshold, it is determined as a connection anomaly. Decompose the station connection anomaly measurement value to identify the institutional sources of the delay and mark the key nodes of the drive chain between stations. The decomposition method is based on the equipment structure analysis, decompose the drive chain between stations into several key nodes, such as bearings, gears, drive belts, connecting rods, etc., and then allocate the abnormal delay values to these nodes. The specific steps include: establish the equipment drive chain structure model to clarify the power transmission path between each station; then, through additional sensor information, such as auxiliary data like drive motor current, bearing vibration, gear meshing sound, etc., locate the abnormal delay; combine the equipment structure knowledge to determine the most likely institutional parts causing the delay. This decomposition analysis can accurately locate the macroscopic station connection anomaly to specific mechanical components, greatly narrowing the scope of fault troubleshooting. The identified key nodes of the drive chain are usually vulnerable or key institutional components in the equipment, such as spindle bearings, transmission gears, synchronous belt pulleys, etc.

[0086] Subsequently, analyze the motion characteristic changes of each key component through the key nodes of the drive chain between stations, and calculate the deviation degree of the component operating state. The calculation formula for the component operating state deviation degree (CDSD) is as follows:

[0087]

[0088] Where, is the component operating state deviation degree; is the number of key performance parameters; , , are the weight coefficients, satisfying + + = 1; and are the dynamic stiffness parameters in the current and standard states respectively; and They are the motion frequency parameters in the current and standard states respectively; and They are the amplitude parameters under the current and standard states, respectively. The dynamic stiffness parameter reflects the characteristics of the component's response to external forces, the motion frequency parameter reflects the periodic characteristics of the component's motion, and the amplitude parameter reflects the amplitude characteristics of the component's motion. These three types of parameters together describe the motion state of the component. This formula weights and synthesizes parameter deviations of different properties to obtain a comprehensive state deviation index. Different weight configurations are used for different types of components. For example, for bearings, the vibration frequency deviation weight is higher; for gears, the dynamic stiffness deviation weight related to meshing clearance is higher.

[0089] The component operation state deviation is compared with the component health threshold to form a key component state evaluation table for packaging equipment. The component health threshold is a critical value of state deviation determined based on historical data and expert experience. Exceeding this value indicates that the component state has entered the abnormal range. The comparison process is: compare the operation state deviation of each key component with the corresponding health threshold, and determine the component health state level based on the relationship between the deviation and the threshold. The health status is usually divided into four levels: normal, slightly abnormal, moderately abnormal, and severely abnormal. The health status evaluation results of all key components are summarized in a table format, including information such as component name, current state deviation, health threshold, health status level, and recommended treatment measures, to form a key component state evaluation table for packaging equipment. The evaluation table is used to display the health status of each key component of the equipment.

[0090] In a specific embodiment, the process of executing step S105 may specifically include the following steps:

[0091] (1) Extract the historical fault records of cutters, heat seals, transmission gears, and motor bearings from the equipment management system and construct a component failure time distribution map;

[0092] (2) Compare the component failure time distribution diagram with the current state parameters in the key component state assessment table to calculate the predicted value of the component's remaining life;

[0093] (3) Based on the predicted value of the remaining service life of the component and the degradation rate of the component state, a component failure risk coefficient is formed;

[0094] (4) Weighting the component failure risk factor, with weight factors including component replacement cost, downtime impact, and spare parts availability, to generate a comprehensive impact score for component failure;

[0095] (5) The inspection urgency is divided into three levels: daily inspection items, key focus items, and emergency handling items, according to the comprehensive impact score of component failures;

[0096] (6) Arrange the components of different inspection levels in the order of the production line layout to generate the priority sequence of the inspection points for the packaging equipment.

[0097] Specifically, extract the historical failure records of cutting knives, heat-sealing strips, transmission gears, and motor bearings from the equipment management system to construct a component failure time distribution map. A cutting knife is a sharp component in the packaging equipment used to cut the packaging material; a heat-sealing strip is a heating element that melts and seals the edge of the packaging material by heating; a transmission gear is a toothed mechanical component that transmits power; a motor bearing is an antifriction component that supports the rotating part of the motor. The historical failure records include information such as the installation time of the component, the failure occurrence time, the failure type, and the repair records. The method of extracting these data is to query the repair record database of the equipment management system and filter out all repair events related to these four types of key components. Classify the filtered data according to the component type, calculate the service time of each component from installation to failure, and form a component life data set. Then, conduct statistical analysis on the life data of each type of component, and use methods such as chi-square fitting to determine the most suitable life distribution model, such as Weibull distribution, lognormal distribution, etc. According to the fitted distribution model, draw a component failure time distribution map to show the variation law of the failure probability of various components with the service time. Compare the component failure time distribution map with the current state parameters in the key component status evaluation form. The key component status evaluation form is formed in the previous steps and contains information such as the current state deviation degree of each key component. The method of calculating the remaining life of the component is based on the proportional hazards model, using the current state parameters of the component as covariates to correct the basic life distribution. The specific steps include: obtaining the basic life distribution parameters of the component from the failure time distribution map; then calculating the state adjustment factor according to the state deviation degree in the status evaluation form; correcting the basic life distribution through the state adjustment factor to obtain the remaining life prediction value considering the current state. For example, if the basic average life of a certain heat-sealing strip is 2000 hours and the current state deviation degree is 0.3, and the calculated state adjustment factor is 0.7, then the remaining life prediction value of this heat-sealing strip is 2000 hours × 0.7 = 1400 hours. This method combines historical statistical data with the current actual state, improving the accuracy of life prediction. Then, based on the remaining life prediction value of the component and the component state deterioration rate, form a component failure risk coefficient. The component state deterioration rate refers to the degree of change of the component state parameters over time, usually determined by the change rate of consecutive multiple state monitoring data. The calculation method is to conduct linear or nonlinear regression analysis on the recent state monitoring data to obtain the change trend of the state parameters over time. The calculation of the component failure risk coefficient comprehensively considers two factors: the remaining life prediction value and the state deterioration rate. For two components with the same remaining life, the component with a faster state deterioration rate has a higher failure risk; similarly, for two components with the same state deterioration rate, the component with a shorter remaining life has a higher failure risk. The specific calculation method is to take the reciprocal of the remaining life prediction value and perform a weighted average with the state deterioration rate to obtain a comprehensive failure risk coefficient.The weight coefficient is determined according to the component type and historical experience. Generally, the weight of the state deterioration rate is between 0.3 and 0.5.

[0098] The component failure risk coefficient is weighted. The weight factors include the component replacement cost, the impact degree of downtime, and the availability of spare parts, to generate a comprehensive impact score of component failure. The component replacement cost refers to the material cost and labor cost required to replace the component; the impact degree of downtime refers to the size of the downtime loss caused by the failure of this component, including direct economic losses and indirect impacts; the availability of spare parts refers to the difficulty of obtaining spare parts for this component, including factors such as the procurement cycle and the number of suppliers. The specific method of weighting is: quantitatively score these three weight factors, usually using a 1-10 point system; then determine the weight coefficient of each factor, and the sum of the weight coefficients is 1; multiply the component failure risk coefficient by the weighted sum of the three weighted factors to obtain the comprehensive impact score of component failure. This scoring method takes into account both the failure risk of the component itself and the comprehensive impact of the failure on production and maintenance, providing a scientific basis for the subsequent prioritization of inspection rounds. According to the comprehensive impact score of component failure, the urgency of inspection is divided into three inspection levels: routine inspection items, key attention items, and emergency treatment items. The classification standard is based on the threshold setting of the score, usually determined by the percentile method or expert experience method. For example, the scores can be arranged in ascending order, and the first 60% of the components are classified as routine inspection items, 60%-90% of the components are classified as key attention items, and the last 10% of the components are classified as emergency treatment items. Or directly set the score threshold, such as scores below 50 are routine inspection items, 50-80 are key attention items, and scores above 80 are emergency treatment items. Routine inspection items are the items that need to be routinely inspected during regular inspections, with a lower inspection frequency; key attention items are the items that require an increased inspection frequency and detail; emergency treatment items are the items that need to be inspected or replaced preventively with priority. This hierarchical management method can reasonably allocate limited inspection resources and improve inspection efficiency.

[0099] Arrange the components of different inspection levels in the order of the production line layout to generate the priority sequence of the inspection points for the packaging equipment. The production line layout order refers to the sequence in the direction of material flow or the spatial position of the equipment. The arrangement method is as follows: Group the components according to the inspection level, with the emergency treatment items ranked at the front, followed by the key attention items, and then the daily inspection items; then, within each level, arrange them in the order of the production line layout, from upstream to downstream, or from the equipment feed end to the discharge end; form a complete list of the priority sequence of inspection points, including information such as component name, location, inspection level, and inspection key points. This sorting method takes into account both the urgency of the inspection and the rationality of the inspection path, facilitating the efficient execution of on-site inspectors. Taking the biscuit packaging line of a certain food packaging enterprise as an example, by analyzing the equipment maintenance records of the past three years, the historical failure data of the cutting knife, heat seal strip, transmission gear, and motor bearing were extracted. It was found that the average service life of the cutting knife is 800 hours, with a standard deviation of 120 hours, and the main failure types are wear and chipping; the average life of the heat seal strip is 600 hours, with a standard deviation of 90 hours, and the main failures are uneven temperature and adhesion; the average life of the transmission gear is 1500 hours, with a standard deviation of 200 hours, and the main failures are tooth surface wear and broken teeth; the average life of the motor bearing is 2000 hours, with a standard deviation of 300 hours, and the main failures are increased noise and overheating. Fit these data to the Weibull distribution respectively to form the failure time distribution diagrams of each component. Then, combined with the previously generated key component status evaluation table, it is known that the current deviation degree of the cutting knife is 0.45, the heat seal strip is 0.28, the transmission gear is 0.33, and the motor bearing is 0.15. Through the proportional hazards model, the remaining life of the cutting knife is calculated to be 350 hours, the heat seal strip is 400 hours, the transmission gear is 950 hours, and the motor bearing is 1700 hours. According to the analysis of the recent three state monitoring data, the state deterioration rate of the cutting knife is 0.03 / hour, the heat seal strip is 0.02 / hour, the transmission gear is 0.01 / hour, and the motor bearing is 0.005 / hour. Combining the remaining life and the deterioration rate, calculate the failure risk coefficients of each component: the cutting knife is 0.85, the heat seal strip is 0.65, the transmission gear is 0.40, and the motor bearing is 0.25. Further considering that the cutting knife replacement cost score is 3 (low cost), the shutdown impact degree score is 8 (great impact), and the spare part availability score is 2 (easy to obtain), the scores of the heat seal strip are 4, 7, 3 respectively, the scores of the transmission gear are 6, 6, 5 respectively, and the scores of the motor bearing are 5, 9, 7 respectively, generate the comprehensive impact score of component failure: the cutting knife is 82, the heat seal strip is 73, the transmission gear is 56, and the motor bearing is 62. According to the scores, classify the cutting knife as an emergency treatment item, the heat seal strip and the motor bearing as key attention items, and the transmission gear as a daily inspection item. Arrange the components in the order of the production line layout (feeding, transmission, cutting, sealing, output) to form the inspection priority sequence: the cutting knife in the emergency treatment item; the motor bearing and the heat seal strip in the key attention items; the transmission gear in the daily inspection item.The inspection personnel can timely detect potential fault risks and ensure the stable operation of the equipment by following this sequence for inspections.

[0100] In a specific embodiment, the process of performing step S106 may specifically include the following steps:

[0101] (1) For each component in the inspection point priority sequence, extract a dedicated set of inspection parameters from the standard inspection specification library, including the cutting tool sharpness detection value, the heat-sealing strip temperature uniformity index, the transmission gear meshing clearance range, and the motor bearing vibration spectrum threshold;

[0102] (2) Associate the set of inspection parameters with the abnormal thresholds in the fault precursor feature library to construct a component status determination standard table;

[0103] (3) Set different inspection cycles based on the component fault risk coefficient to generate a component inspection time interval matrix;

[0104] (4) Integrate the set of inspection parameters, the determination standard table, and the inspection time interval matrix into a component inspection operation card;

[0105] (5) Schedule the time for the component inspection operation card according to the production plan arrangement and maintenance resource allocation situation to generate an inspection execution calendar;

[0106] (6) Combine the component inspection operation card with the inspection execution calendar to generate a preventive maintenance plan for the packaging equipment.

[0107] Specifically, by querying the pre-established device detection standard database, the professional detection items and methods for each component are obtained. The standard detection specification library is a collection of technical specifications and industry standards provided by equipment manufacturers, including the detection methods, judgment criteria, and operating procedures for various components. For the cutting knife, the sharpness detection value is extracted, that is, the quantitative index for measuring the cutting ability of the cutting knife with a standard medium under a standard pressure; for the heat-sealing strip, the temperature uniformity index is extracted, that is, the maximum deviation value of the surface temperature distribution of the heat-sealing strip; for the transmission gear, the meshing clearance range is extracted, that is, the measured value of the clearance between the meshing gears; for the motor bearing, the vibration spectrum threshold is extracted, that is, the upper limit of the vibration amplitude in each frequency band during the normal operation of the bearing. Associating the detection parameter set with the abnormal thresholds in the fault precursor feature library to construct a component status judgment standard table means pairing the detection parameters extracted in the previous step with the corresponding abnormal feature thresholds in the fault precursor feature library. The fault precursor feature library stores the typical change patterns of equipment state parameters before different fault types occur, including the abnormal thresholds of each parameter. The association method is to query the fault precursor feature library by comparison, extract the abnormal thresholds corresponding to each detection parameter, and set the judgment criteria. For example, when the sharpness of the cutting knife is lower than 80% of the standard value, it is determined that replacement is needed; when the deviation of the temperature uniformity of the heat-sealing strip exceeds 5°C, it is determined that adjustment is needed; when the meshing clearance of the transmission gear exceeds 15% of the standard range, it is determined that replacement is needed; when the amplitude of the characteristic frequency in the vibration spectrum of the motor bearing exceeds the standard threshold, it is determined that maintenance is needed. These judgment criteria form the component status judgment standard table, which contains information such as detection parameters, normal range, warning range, and danger range.

[0108] The process of setting different inspection intervals based on the component failure risk coefficient and generating the component inspection time interval matrix is to set different inspection frequencies for different components according to the previously calculated component failure risk coefficients. Components with a high failure risk coefficient require more frequent inspections, while components with a low failure risk coefficient can have reduced inspection frequencies. The specific setting method is to establish a mapping relationship between the risk coefficient and the inspection interval. For example, components with a risk coefficient above 0.8 are inspected once per shift, those with a risk coefficient of 0.5 - 0.8 are inspected once a day, those with a risk coefficient of 0.3 - 0.5 are inspected once a week, and those with a risk coefficient below 0.3 are inspected once a month. At the same time, considering the importance of the components and their historical failure modes, the mapping relationship is adjusted. The finally formed inspection time interval matrix is a two-dimensional table, with the row being the component name and the column being the time unit (hours, days, weeks, etc.), and the cell value being the number of inspections. Integrating the detection parameter set, the judgment standard table, and the inspection time interval matrix into the component inspection operation card is to structurally integrate the data obtained in the previous three steps to form a standardized document convenient for on-site execution. The inspection operation card contains the following content: basic component information (name, location, number, etc.); the detection parameter set and the description of the detection method; the status judgment standard table; the inspection cycle and time arrangement; the guidance on handling measures when abnormalities are found; the historical inspection record area, etc. For different components, dedicated operation card templates are designed to ensure that the inspection personnel can conduct inspections and make judgments according to the standard procedures.

[0109] The process of scheduling the component inspection operation card according to the production plan arrangement and maintenance resource allocation situation and generating the inspection execution calendar is to coordinate the inspection operations with the actual production plan to avoid inspection activities affecting normal production. The specific method is as follows: Obtain the production plan arrangement for a future period (usually one month), including production time, downtime, product changeover time, etc.; then match the work schedules and skills of the maintenance personnel; according to these constraints, assign specific execution times and responsible persons to each inspection operation card. Priority is given to arranging inspection items that affect production during the equipment downtime period, while inspection items that do not affect production can be arranged during the equipment operation period. The generated inspection execution calendar is usually displayed in the form of a Gantt chart or a calendar, visually showing the inspection task arrangements for each day. Combining the component inspection operation card with the inspection execution calendar to generate the preventive maintenance plan for the packaging equipment is to integrate the operation card and the execution calendar formed in the previous two steps into a complete maintenance management document. The preventive maintenance plan contains the following content: an overview of the plan, including maintenance objectives, scope of application, responsibility division, etc.; the inspection execution calendar, showing the time arrangements for all inspection tasks; the inspection operation cards for each component, serving as specific execution guides; the list of maintenance resource requirements, including personnel, tools, spare parts, etc.; the evaluation criteria and methods for maintenance effects; the plan adjustment and emergency handling mechanism, etc. This integration process ensures the systematicness and continuity of preventive maintenance activities, providing clear action guidelines for equipment management personnel and inspection personnel.

[0110] Taking a beverage packaging production line as an example, for the cutting knife classified as an emergency handling item, the cutting knife sharpness detection standard was extracted from the standard detection specification library, stipulating that the cutting performance should be tested under a fixed pressure using a reference material of standard thickness; for the heat seal strip in the key attention items, the temperature uniformity detection standard was extracted, requiring the use of an infrared thermal imager to scan the surface temperature distribution of the heat seal strip; for the motor bearing, the vibration spectrum detection standard was extracted, stipulating that a vibration analyzer should be used to collect the bearing vibration data and analyze the characteristic frequency. These detection standards were associated with the abnormal thresholds in the fault precursor feature library, determining that the cutting knife needs to be replaced when its sharpness is lower than 85% of the standard value, the heat seal strip needs to be adjusted when the temperature difference between each point exceeds 8°C, and the motor bearing needs to be repaired when the vibration amplitude of the characteristic frequency exceeds 0.7g. Based on the fault risk coefficients (cutting knife 0.85, heat seal strip 0.65, motor bearing 0.62), it was set to inspect the cutting knife once every 4 hours, the heat seal strip once per shift (8 hours), and the motor bearing once a day. These information were integrated into the inspection operation cards for each component, and combined with the production plan (two shifts of production from Monday to Friday, single shift of production on weekends, and major equipment maintenance shutdown on Wednesday of one week per month) and the scheduling of maintenance personnel, a monthly inspection execution calendar was generated. Finally, all the cards and calendars were integrated into a complete preventive maintenance plan, providing detailed detection indicators, judgment criteria, and execution schedules for the inspection personnel, ensuring the healthy operation of key components.

[0111] The above described the inspection method for the packaging equipment in the embodiment of the present application. Next, the inspection system for the packaging equipment in the embodiment of the present application will be described. Please refer to Figure 2 One embodiment of the inspection system for the packaging equipment in the embodiment of the present application includes:

[0112] Collect the packaging process parameters through the tension sensor, sealing temperature probe, and conveyor belt pulse encoder on the packaging equipment to form a comprehensive data set of the operating conditions of the packaging equipment;

[0113] According to the comprehensive data set of the operating conditions, use spectral analysis to extract the characteristics of abnormal consumption of packaging materials, fluctuations in the sealing defect rate, and uneven speed, and generate the operation deviation indicators of the packaging equipment;

[0114] Based on the operation deviation indicators, establish association rules by comparing the historical data of product quality and equipment status parameters to obtain the fault precursor feature library of the packaging equipment;

[0115] According to the fault precursor feature library, adopt the time series comparison method to analyze the action coordination between each station on the packaging line to form the status evaluation table of the key components of the packaging equipment;

[0116] According to the status evaluation table of the key components, combine the historical maintenance records to calculate the fault probabilities of the cutting knife, heat seal strip, transmission gear, and motor bearing, and generate the priority sequence of the inspection points of the packaging equipment.

[0117] According to the above inspection point priority sequence, formulate an inspection operation guidance plan including detection parameters, judgment criteria, and inspection cycles, and form a preventive maintenance plan for the packaged equipment.

[0118] Refer to Figure 3 , in an embodiment of the present invention, a computer device is further provided. The computer device may be a server, and its internal structure may be as Figure 3 shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0119] Those skilled in the art can understand that Figure 3 the structure shown in

[0120] is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0121] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0122] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0123] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs and other various media that can store program codes.

[0124] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A method for inspecting a packing device, characterized in that The method for inspecting the packing equipment includes: Collecting the packaging process parameters through the tension sensor, sealing temperature probe, and conveyor belt pulse encoder on the packing equipment to form a comprehensive data set of the working conditions of the packing equipment; According to the comprehensive data set of the working conditions, using spectrum analysis to extract the characteristics of abnormal consumption of packaging materials, fluctuations in the defective sealing rate, and uneven speed, and generating the operation deviation index of the packing equipment; Based on the operation deviation index, establishing an association rule by comparing the historical data of product quality and equipment status parameters to obtain a fault precursor feature library of the packing equipment, including: extracting the packaging integrity score, sealing strength test value, and appearance qualification rate data from the product quality inspection database to construct a historical record matrix of product quality; performing time series pairing on the historical record matrix of product quality and the operation deviation index corresponding to time to form a quality-equipment status two-factor data chain; performing lag analysis on the quality-equipment status two-factor data chain to determine the time window from the occurrence of the operation deviation to the decline of product quality, and forming a state-quality influence period diagram; grouping through the state-quality influence period diagram to distinguish the development characteristics of different fault types, and drawing a fault evolution trajectory curve; extracting the key inflection points and threshold characteristics from the fault evolution trajectory curve, and marking the early warning signal points; associating and annotating the early warning signal points with the specific fault types in the historical maintenance records to obtain a fault precursor feature library of the packing equipment; According to the fault precursor feature library, using the time series comparison method to analyze the action coordination between each station on the packaging line to form a status evaluation table of the key components of the packing equipment, including: collecting the action time series signals of the feeding station, folding station, sealing station, cutting station, and output station from the control system of the packing equipment to construct a time series of the action of the packing station; synchronizing the phase of the time series of the action of the packing station with the standard time series template in the fault precursor feature library to generate a deviation map of the action of the station; extracting the action conversion delay value between adjacent stations based on the deviation map of the action of the station to form a measure value of abnormal connection between stations; decomposing the measure value of abnormal connection between stations to identify the mechanism source causing the delay, and marking the key nodes of the drive chain between stations; analyzing the change of the motion characteristics of each key component through the key nodes of the drive chain between stations, and calculating the deviation degree of the component operation state; comparing the deviation degree of the component operation state with the component health threshold to form a status evaluation table of the key components of the packing equipment; According to the status evaluation table of the key components, combining the historical maintenance records to calculate the fault probabilities of the cutting knife, heat seal strip, transmission gear, and motor bearing, and generating a priority sequence of the inspection points of the packing equipment; According to the priority sequence of the inspection points, formulating an inspection operation guidance plan including inspection parameters, judgment criteria, and inspection periods to form a preventive maintenance plan for the packing equipment.

2. The inspection method for the packaging equipment according to claim 1, wherein, The collecting the packaging process parameters through the tension sensor, sealing temperature probe, and conveyor belt pulse encoder on the packing equipment to form a comprehensive data set of the working conditions of the packing equipment includes: Performing multi-point sampling on the instantaneous tension waveform of the tension sensor during the conveying process of the packaging material, the temperature curve of the sealing temperature probe during the heat sealing process, and the speed change signal of the conveyor belt pulse encoder. Divide the sampling signal into working units according to the operating cycle of the packing equipment, and identify the key time nodes in the complete packaging cycle; Align the multi-sensor data according to the key time nodes to eliminate the signal transmission delay difference; Remove the background interference from the phase-aligned signal by wavelet denoising method, and extract the effective working condition signal; Construct a three-dimensional working condition feature space, and map the equipment operating state with tension-temperature-speed as the coordinate axes; Conduct data point density analysis in the three-dimensional working condition feature space to form a comprehensive data set of the packing equipment working conditions.

3. The inspection method for the packaging equipment according to claim 1, wherein According to the comprehensive data set of the working conditions, use the spectrum analysis method to extract the abnormal consumption of packaging materials, the fluctuation of the defective sealing rate, and the uneven speed characteristics, and generate the operation deviation index of the packing equipment, including: Perform multi-scale wavelet decomposition on the tension signal in the comprehensive data set of the working conditions, and construct a tension fluctuation characteristic tree spectrum; Extract the abnormal energy distribution area from the tension fluctuation characteristic tree spectrum, calculate the peak deviation ratio of the material tension, and form the identification fingerprint of the abnormal consumption of packaging materials; Perform phase space reconstruction on the sealing temperature data sequence, and draw the temperature trajectory attractor diagram; Quantify the stability of the sealing system by calculating the Lyapunov exponent of the temperature trajectory attractor diagram, and establish an early warning index for the fluctuation of the defective sealing rate; Adopt non-linear detrending analysis for the conveyor belt pulse sequence to extract the hidden mode of small speed fluctuations; Perform cross-correlation analysis on the hidden mode of small speed fluctuations and the characteristic mode during standard operation to form an uneven speed characteristic spectrum; Through the multi-dimensional feature space mapping algorithm, integrate the identification fingerprint of the abnormal consumption of packaging materials, the early warning index of the fluctuation of the defective sealing rate, and the uneven speed characteristic spectrum into the operation deviation index of the packing equipment.

4. The inspection method for the packaging equipment according to claim 1, characterized in that According to the key component status evaluation form, calculate the failure probabilities of the cutting knife, heat sealing strip, transmission gear, and motor bearing in combination with the historical maintenance records, and generate the priority sequence of the inspection points of the packing equipment, including: Extract the historical failure records of the cutting knife, heat sealing strip, transmission gear, and motor bearing from the equipment management system, and construct a component failure time distribution diagram; Compare the component failure time distribution diagram with the current state parameters in the key component status evaluation form, and calculate the predicted remaining life value of the component; Based on the predicted remaining life value of the component and the component state deterioration rate, form a component failure risk coefficient; Perform weighted processing on the component failure risk coefficient, and the weight factors include component replacement cost, shutdown impact degree, and spare part availability, and generate a comprehensive impact score of component failure; Divide the inspection urgency according to the comprehensive impact score of component failure, and divide it into three inspection levels: daily inspection items, key attention items, and emergency handling items; Arrange the components of different inspection levels in the order of the production line layout to generate the priority sequence of the inspection points of the packing equipment.

5. The inspection method for the packaging equipment according to claim 1, characterized in that, According to the priority sequence of the inspection points, formulate an inspection operation guidance plan including detection parameters, judgment criteria, and inspection cycles, and form a preventive maintenance plan for the packing equipment, including: For each component in the priority sequence of the inspection points, extract a dedicated detection parameter set from the standard detection specification library, including the detection value of the cutting knife sharpness, the temperature uniformity index of the heat sealing strip, the meshing clearance range of the transmission gear, and the vibration spectrum threshold of the motor bearing; Associate the detection parameter set with the abnormal thresholds in the fault precursor feature library to construct a component status determination standard table; Set differential inspection cycles based on the component fault risk coefficient to generate a matrix of component inspection time intervals; Integrate the detection parameter set, the determination standard table, and the inspection time interval matrix into a component inspection operation card; Schedule the time of the component inspection operation card according to the production plan arrangement and maintenance resource allocation situation to generate an inspection execution calendar; Combine the component inspection operation card with the inspection execution calendar to generate a preventive maintenance plan for the packaging equipment.

6. A patrol inspection system for a packaging device, which is used to implement the patrol inspection method of the packaging device described in any one of claims 1-5, characterized in that, The packaging equipment inspection system includes: An acquisition module for collecting packaging process parameters through a tension sensor, a sealing temperature probe, and a conveyor belt pulse encoder on the packaging equipment to form a comprehensive dataset of the operating conditions of the packaging equipment; A generation module for extracting features of abnormal packaging material consumption, fluctuating sealing defect rate, and uneven speed from the comprehensive dataset of operating conditions using spectrum analysis to generate operating deviation indicators for the packaging equipment; A construction module for establishing an association rule by comparing historical data of product quality and equipment status parameters based on the operating deviation indicators to obtain a fault precursor feature library for the packaging equipment, including: extracting packaging integrity scores, sealing strength test values, and appearance qualification rate data from the product quality inspection database to construct a historical record matrix of product quality; performing time series pairing of the historical record matrix of product quality with the corresponding operating deviation indicators in time to form a quality-equipment status two-factor data chain; performing lag analysis on the quality-equipment status two-factor data chain to determine the time window from the occurrence of the operating deviation to the decline of product quality to form a state-quality influence cycle diagram; grouping through the state-quality influence cycle diagram to distinguish the development characteristics of different fault types and drawing a fault evolution trajectory curve; extracting key inflection points and threshold features from the fault evolution trajectory curve and marking the early warning signal points; associating and annotating the early warning signal points with specific fault types in the historical maintenance records to obtain a fault precursor feature library for the packaging equipment; An analysis module for analyzing the action coordination between each station on the packaging line using the time series comparison method based on the fault precursor feature library to form an evaluation table of the key component status of the packaging equipment, including: collecting action time series signals of the feeding station, folding station, sealing station, cutting station, and output station from the control system of the packaging equipment to construct an action time series of the packaging stations; synchronizing the phase of the action time series of the packaging stations with the standard time series template in the fault precursor feature library to generate a station action deviation map; extracting the action conversion delay value between adjacent stations based on the station action deviation map to form a measure of abnormal station connection; decomposing the measure of abnormal station connection to identify the mechanism source causing the delay and marking the key nodes of the drive chain between stations; analyzing the change of the motion characteristics of each key component through the key nodes of the drive chain between stations and calculating the deviation degree of the component operating state; comparing the deviation degree of the component operating state with the component health threshold to form an evaluation table of the key component status of the packaging equipment; A calculation module, configured to calculate the failure probabilities of cutting knives, heat-sealing strips, transmission gears, and motor bearings according to the critical component status evaluation table and in combination with historical maintenance records, and generate a priority sequence of inspection points for the packing equipment; A maintenance module, configured to formulate an inspection operation guidance plan including inspection parameters, judgment criteria, and inspection periods according to the priority sequence of inspection points, and form a preventive maintenance plan for the packing equipment.

7. A computer device, characterized in that, It includes a memory and a processor, and the memory stores a computer program that can run on the processor. It is characterized in that when the processor executes the computer program, the inspection method of the packing equipment described in any one of claims 1 to 5 is implemented.

8. A computer-readable storage medium, on which a computer program is stored, and when the computer program is run by a processor, the processor is caused to execute the inspection method of the packing equipment described in any one of claims 1 to 5.

Images