Chain production data monitoring and optimizing system based on industrial internet of things

Abstract

The invention relates to the technical field of intelligent manufacturing, in particular to a chain production data monitoring and optimizing system based on industrial internet of things, which comprises a stress load monitoring module, a shock wave propagation analysis module, a vibration feature extraction module, a load optimization control module and a fracture risk early warning module. According to the method, through combination of stress data acquisition and coordinate index and gradient calculation, a transient stress overload area of chain plate stamping and pin shaft heat treatment is positioned, local extreme points are dynamically screened to improve anomaly detection precision, an energy accumulation coefficient is quantified by shock wave propagation speed and attenuation rate, and a vibration anomaly source of a chain pin shaft and a chain plate is associated. The vibration signals are subjected to short-time Fourier transform to analyze riveting and assembling characteristic frequencies, high-resolution stress mode characteristic values are generated to support parameter optimization, operation data are simulated to construct a stress accumulation curve, test parameters are dynamically adjusted in combination with a short-term stress limit to optimize load uniformity, and an impact energy accumulation value is matched with a safety threshold to calibrate load distribution; the fracture risk is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent manufacturing, and particularly to a chain production data monitoring and optimization system based on the industrial Internet of Things. Background Art

[0002] The technical field of intelligent manufacturing includes a manufacturing technology system based on digitization, networking, and intelligence, aiming to improve the intelligent level of manufacturing systems through the deep integration of information technology and manufacturing technology. The core contents of this technical field include links such as intelligent perception, intelligent control, intelligent decision-making, and intelligent collaboration, covering the entire process from design, production, management to service. Intelligent manufacturing uses technical means such as the industrial Internet of Things, big data, and artificial intelligence to achieve real-time monitoring of production equipment, fault prediction, quality optimization, and adaptive regulation of the production process. Its systematicness includes multiple aspects such as intelligent factories, intelligent production lines, intelligent equipment, and intelligent operation and maintenance. By establishing an efficient, flexible, and green manufacturing model, it improves production efficiency, reduces costs, and optimizes resource allocation.

[0003] Among them, the chain production data monitoring and optimization system refers to collecting, analyzing, and optimizing key data involved in the chain production process through industrial Internet of Things technology to improve production accuracy and stability. This system mainly focuses on technical matters such as data collection, anomaly detection, parameter optimization, and predictive analysis in chain production. It obtains data such as the operating status, processing accuracy, and equipment load on the chain production line through a sensor network, uses data modeling and analysis technology to optimize production parameters, and conducts trend analysis based on predictive algorithms to adjust processing parameters and optimize the production process. The system uses edge computing technology to preprocess data, reduce data transmission latency, and at the same time combines a cloud computing platform for data storage and in-depth analysis to form a data-driven decision-making support mechanism to ensure the efficiency and stability of the chain production process.

[0004] The existing technology relies on fixed measuring points and static threshold detection, and cannot capture the transient stress non-linear changes in the stamping forming and heat treatment processes. There are blind spots and lags in the identification of overloaded areas. Traditional vibration analysis uses static spectra, does not combine the shock wave propagation speed and attenuation law, has insufficient ability to quantify the energy accumulation trend, and has a high risk of misjudgment. Parameter optimization is based on offline modeling of historical data, lacks the ability to dynamically analyze characteristic frequencies, and is difficult to match real-time working conditions changes. The existing warning mechanism uses fixed safety thresholds and does not correlate with the real-time load distribution state, and there is a significant deviation between the fracture risk assessment and the actual stress. For example, when the equipment load suddenly changes, the traditional system may mis-trigger shutdown or miss detecting local stress mutations due to the inability to quantify the energy accumulation trend. Data transmission relies on centralized cloud processing, and the lack of edge preprocessing ability leads to the loss of key time-series data, and the optimization control timeliness is limited. Summary of the Invention

[0005] The object of the present invention is to solve the disadvantages existing in the prior art, and a chain production data monitoring and optimization system based on the industrial Internet of Things is proposed.

[0006] To achieve the above object, the present invention adopts the following technical solutions: The chain production data monitoring and optimization system based on the industrial Internet of Things includes: The stress load monitoring module arranges stress sensors in the chain plate stamping and forming area and the pin heat treatment area, collects transient data, determines the measuring point position by combining coordinate indexing, calculates the stress change rate, uses numerical differentiation to obtain the gradient, screens local extreme points, and obtains the stress overload area distribution value; The shock wave propagation analysis module obtains the vibration data of the key parts of the chain pin and the chain plate according to the stress overload area distribution value, calculates the propagation speed and attenuation rate, conducts a comparative analysis of the waveforms, and generates a shock wave energy accumulation coefficient; The vibration feature extraction module obtains the vibration signal during production based on the shock wave energy accumulation coefficient, distributes energy using the short-time Fourier transform, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning and assembly process, and generates the chain force mode characteristic value; The load optimization control module calls the chain simulation operation data to construct a stress accumulation curve based on the chain force mode characteristic value, calculates the short-term force limit of the chain, adjusts the chain test parameters, and generates the optimized chain load distribution value; The fracture risk warning module obtains the shock wave feedback data based on the optimized chain load distribution value, calculates the cumulative impact energy value, judges the load safety threshold, compares the stress states within the threshold range, adjusts the load distribution, and generates the chain fracture risk warning value.

[0007] As a further solution of the present invention, the stress overload area distribution value includes the measuring point position, the stress change rate, and the local extreme points. The shock wave energy accumulation coefficient is specifically the propagation speed, the attenuation rate, and the waveform comparison analysis result. The chain force mode characteristic value includes the characteristic frequency of the impact riveting process, the characteristic frequency of the tensioning and assembly process, and the energy distribution. The optimized chain load distribution value specifically refers to the stress accumulation curve, the short-term force limit, and the test parameter adjustment plan.

[0008] As a further solution of the present invention, the stress load monitoring module includes: The stress data acquisition sub-module arranges stress sensors in the chain plate stamping and forming area and the pin heat treatment area, obtains transient data, records the stress values of multiple measuring points, and determines the position of the measuring points by combining coordinate indexing, generating the measuring point stress distribution data; The stress gradient calculation sub-module calculates the stress difference between adjacent measuring points based on the measuring point stress distribution data, calculates the spatial displacement according to the measuring point coordinates, and simultaneously calculates the stress change rate using numerical differentiation, generating the measuring point stress change rate data; The stress overload area screening sub-module calls the stress change rate data of the measuring points, screens the local extreme points, extracts the stress gradient change in the overload area, and uses the formula: ; Performs operations to obtain the stress distribution value in the overload area, and obtains the stress overload area distribution value; Wherein, represents the stress difference between adjacent measuring points, represents the spatial distance between adjacent measuring points, represents the stress gradient weight parameter, represents the stress value of the measuring point, represents the average stress of all measuring points, represents the stress overload area distribution value, represents the number of stress gradient calculation points, represents the total number of measuring points, i represents a certain measuring point, j represents another measuring point related to i in a certain calculation step, and id represents the unique identifier of the measuring point.

[0009] As a further solution of the present invention, the shock wave propagation analysis module includes: The stress data analysis sub-module monitors the key parts of the chain pin and the chain plate based on the stress overload area distribution value, obtains the stress response data, records the data acquisition time and the corresponding stress value, and at the same time arranges the stress data in time series, calculates the time interval and stress change rate between different measuring points, screens the abnormal data and performs numerical interpolation processing to obtain the stress time series data set; The propagation speed calculation sub-module calls the stress time series data set, calculates the time difference of the shock wave arrival between different measuring points, calculates the local propagation speed based on the measuring point spacing, and at the same time performs a weighted summation process on the propagation speeds of all measuring points to obtain the average propagation speed, using the formula: ; Performs operations to obtain the shock wave propagation rate of multiple measuring points and calculates the global propagation speed to obtain the average shock wave propagation speed; Wherein, represents the average shock wave propagation speed, represents the th measuring point spacing, represents the th shock wave arrival time difference of the measuring point, represents the standard deviation of the shock wave arrival time differences of all measuring points, represents the th energy loss coefficient of the measuring point, represents the total number of measuring points, represents the time difference between measuring points; The impact energy analysis sub-module calls the average velocity of shock wave propagation, combines with the stress changes in the stress time series dataset, calculates the energy loss per unit propagation distance, and simultaneously calculates the total energy loss of the shock wave based on the stress attenuation rate of multiple measurement points, and accumulates the energy losses of all measurement points to obtain the shock wave energy accumulation coefficient.

[0010] As a further solution of the present invention, the vibration feature extraction module includes: The energy accumulation calculation sub-module calculates the shock wave energy distribution of the vibration signal under different time windows based on the shock wave energy accumulation coefficient, extracts the vibration amplitudes under multiple windows, calculates the cumulative energy per unit time, and obtains the energy peak distribution under different processes after normalization processing, and generates shock wave energy peak data; The characteristic frequency screening sub-module calls the shock wave energy peak data, uses the short-time Fourier transform to distribute energy, screens the key frequency bands corresponding to the energy peaks, calculates the contribution rates of multiple frequency bands, removes the frequency components with contribution rates lower than the set threshold, and uses the formula: ; Calculate the characteristic frequency screening value, obtain the characteristic frequency ranges of the chain impact riveting process and the chain tensioning assembly process, and generate process characteristic frequency data; Wherein, represents the characteristic frequency screening value, represents the th energy value of the frequency band, represents the th center frequency of the frequency band, represents the th vibration signal amplitude of the time window, represents the th time length of the time window, represents the total number of frequency bands, represents the number of frequency bands for contribution calculation, represents the total number of time windows; The force mode extraction sub-module calls the process characteristic frequency data, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning assembly process, calculates the spectral response characteristic values under multiple processes, performs normalization processing, matches the vibration mode categories, and establishes the chain force mode characteristic values.

[0011] As a further solution of the present invention, the load optimization control module includes: The force parameter analysis sub-module calls the chain simulation operation data based on the chain force mode characteristic values, obtains the force data under different operation states, extracts the force change trends of multiple nodes of the chain, calculates the force mean value, fluctuation range and force change rate under different operating conditions of the chain, and obtains the chain force characteristic parameters; The short-term force limit calculation sub-module of the chain calls the force characteristic parameters of the chain, analyzes the force states of the chain in different time periods, and uses the formula: ; Calculate the short-term force limit of the chain, adjust the short-term force threshold, and generate the short-term force limit value of the chain; Among them, represents the current short-term force limit of the chain, represents the actual force value in the th time period, represents the average force value, represents the standard deviation of the force value, represents the time point of each time period, represents the optimal time point, represents the standard deviation of time, represents the total number of time periods, represents the number of time periods for short-term force calculation; The chain load distribution optimization sub-module calls the short-term force limit value of the chain, adjusts the chain test parameters, compares the matching degree between the different chain force thresholds and the short-term limit force value, screens the optimal test parameters, and generates the optimized chain load distribution value.

[0012] As a further solution of the present invention, the system further includes: The fracture risk warning module, based on the optimized chain load distribution value, obtains shock wave feedback data, calculates the cumulative impact energy value, determines the load safety threshold, compares the force states within the threshold range, adjusts the load distribution, and generates a chain fracture risk warning value; The chain fracture risk warning value includes the cumulative impact energy value, the load safety threshold, and the comparison result of the force state.

[0013] As a further solution of the present invention, the fracture risk warning module includes: The load distribution calculation sub-module, based on the optimized chain load distribution value, calls the force conditions of the multi-node links, calculates the dynamic load distribution on the links, analyzes the load transfer relationship between different links, and adjusts the local load deviation to obtain the optimized link load distribution coefficient; The impact energy accumulation sub-module calls the optimized link load distribution coefficient, obtains shock wave feedback data, calculates the distribution of impact energy on the links, analyzes the transfer mode of impact energy on the links according to the change trend of the cumulative impact energy, and uses the formula: ; Calculate and obtain the cumulative impact energy value; Among them, represents the cumulative value of impact energy, represents the load at the link location, represents the link location speed, represents the link location mass, represents the total number of impact energy calculation nodes; The fracture risk assessment sub-module calls the cumulative value of impact energy, determines whether it exceeds the set load safety threshold, compares the stress states within the threshold range, adjusts the load distribution of the links, calculates the fracture risk level of the entire chain, and obtains the fracture risk warning value of the chain.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In the present invention, through stress data acquisition combined with coordinate indexing and gradient calculation, the transient stress overload regions of link plate stamping and pin heat treatment are located, and local extreme points are dynamically screened to improve the accuracy of anomaly detection. The shock wave propagation speed and attenuation rate quantify the energy accumulation coefficient, and accurately correlate the abnormal vibration sources of the chain pins and link plates. The vibration signals are analyzed by short-time Fourier transform to resolve the characteristic frequencies of riveting and assembly processes, generate high-resolution stress pattern characteristic values, and support dynamic optimization of parameters. Based on the simulated operation data, a stress accumulation curve is constructed, and the test parameters are dynamically adjusted in combination with the short-term stress limit to optimize the uniformity of load distribution. The cumulative value of impact energy is matched with the safety threshold in real time, the load distribution state is calibrated, and the risk of sudden fracture is reduced. Edge computing preprocessing synchronizes cloud analysis, shortens the closed-loop control response time, and improves decision-making efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is the system flow chart of the present invention; Figure 2 is the flow chart of the stress load monitoring module of the present invention; Figure 3 is the flow chart of the shock wave propagation analysis module of the present invention; Figure 4 is the flow chart of the vibration characteristic extraction module of the present invention; Figure 5 is the flow chart of the load optimization control module of the present invention; Figure 6 is the flow chart of the fracture risk warning module of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0016] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the drawings. These are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more, unless otherwise specifically defined.

[0018] Embodiment 1

[0019] Please refer to Figure 1 , the chain production data monitoring and optimization system based on the industrial Internet of Things includes: The stress load monitoring module arranges stress sensors in the chain plate stamping and forming area and the pin heat treatment area, collects transient data, determines the measuring point positions by combining coordinate indexing, calculates the stress change rate, uses numerical differentiation to obtain the gradient, screens local extreme points, and obtains the stress overload area distribution value; The shock wave propagation analysis module obtains the vibration data of the key parts of the chain pins and chain plates according to the stress overload area distribution value, calculates the propagation speed and attenuation rate, conducts a comparative analysis of the waveforms, and generates the shock wave energy accumulation coefficient; The vibration feature extraction module obtains the vibration signals during production based on the shock wave energy accumulation coefficient, distributes the energy using the short-time Fourier transform, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning and assembly process, and generates the chain force mode characteristic value; The load optimization control module constructs a stress accumulation curve by calling the chain simulation operation data based on the chain force mode characteristic value, calculates the short-term force limit of the chain, adjusts the chain test parameters, and generates the optimized chain load distribution value; The fracture risk warning module obtains the shock wave feedback data based on the optimized chain load distribution value, calculates the cumulative impact energy value, determines the load safety threshold, compares the stress states within the threshold range, adjusts the load distribution, and generates the chain fracture risk warning value.

[0020] The stress overload area distribution value includes the measuring point positions, the stress change rate, and the local extreme points. The shock wave energy accumulation coefficient specifically refers to the propagation speed, the attenuation rate, and the waveform comparative analysis result. The chain force mode characteristic value includes the characteristic frequencies of the impact riveting process, the tensioning and assembly process, and the energy distribution. The optimized chain load distribution value specifically refers to the stress accumulation curve, the short-term force limit, and the test parameter adjustment plan. The chain fracture risk warning value includes the cumulative impact energy value, the load safety threshold, and the stress state comparison result.

[0021] Please refer toFigure 2 , the stress load monitoring module includes: The stress data acquisition sub-module arranges stress sensors in the chain plate stamping forming area and the pin heat treatment area, obtains transient data, records the stress values of multiple measuring points, and determines the positions of the measuring points in combination with coordinate indexing to generate measuring point stress distribution data; Select stress sensors with appropriate ranges. In the stamping forming area, the sensors should be installed at positions where stress concentration is likely to occur, such as near the action points of stamping dies, the stressed corners of chain plates, and the direct transmission paths of stamping loads. In the pin heat treatment area, the sensors should be arranged in the key contact areas after the pins expand due to heat and in areas with large stress gradient changes. High-frequency sampling mode is used for data acquisition to ensure obtaining data on impact loads and transient stress changes during the heat treatment process. After the stress value of each measuring point is recorded, its specific position on the chain plate or pin is calibrated through the coordinate indexing system, and then the measuring point stress distribution data is generated. In specific implementation, if the stamping load is applied within 50 ms, the sampling frequency can be set to 10 kHz to ensure obtaining a complete stress change curve. For example, during the stamping process, the recorded stress value of the measuring point at the edge of the chain plate may be 120 MPa, while the local stress measuring point after the pin heat treatment may show 80 MPa. Through the spatial positioning of these data, the stress distribution data during the entire stamping forming and heat treatment processes is formed.

[0022] The stress gradient calculation sub-module calculates the stress difference between adjacent measuring points based on the measuring point stress distribution data, calculates the spatial displacement based on the measuring point coordinates, and simultaneously calculates the stress change rate using numerical differentiation to generate measuring point stress change rate data; If the stress values recorded by adjacent measuring points i and i + 1 are 110 MPa and 125 MPa respectively, the corresponding stress difference is 15 MPa. Then, the spatial displacement is calculated based on the measuring point coordinates. For example, if the distance between two measuring points Δx = 5 mm, the stress gradient at this position is calculated as MPa / mm. Further, the stress change rate is calculated using the numerical differentiation method. Assuming that the recorded stress value of a measuring point changes with time as σ(t), for a certain measuring point at t1 and t2 (interval Δt = 0.01 s), and their stresses are 90 MPa and 95 MPa respectively, the change rate is calculated as MPa / s. Finally, the measuring point stress change rate data is obtained. Through these calculations, a stress change rate distribution map of the entire area is constructed. The result shows that the stress change rate of the measuring points is relatively fast, and there may be a large stress concentration phenomenon in the area. Further analysis can be used to screen potential overloaded areas.

[0023] The stress overloaded area screening sub-module calls the measuring point stress change rate data, screens local extreme points, extracts the stress gradient changes in the overloaded areas, and uses the formula: ; The stress distribution value of the overload area is obtained by calculation to obtain the stress overload area distribution value; in, Represents the stress difference between adjacent measuring points, Represents the spatial distance between adjacent measuring points, represents the stress gradient weight parameter, Represents the stress value of the measuring point, represents the mean stress value of all measuring points, Represents the stress overload area distribution value, represents the number of stress gradient calculation points, Represents the total number of measurement points, i represents a measurement point, j represents another measurement point related to i in a certain calculation step, and id represents the unique identification of the measurement point.

[0024] For example, if the stress change rate data measured in the stamping area of a chain plate is {500, 450, 300, 700, 250} MPa / s, the local extreme point may be 700 MPa / s. When calculating the stress gradient change in the overload area, the formula is used: ; Assuming that the stress difference between adjacent measuring points is {15, 20, 18, 25} MPa, and the spacing between adjacent measuring points is Δx = {5, 4, 6, 5} mm, the calculated stress gradients are {3, 5, 3, 5} MPa / mm, respectively. Assuming the corresponding weights w = {1, 1.2, 1, 1.1}, the weighted sum is calculated as follows: ; Then calculate the stress mean of all measuring points. Assuming that the stress values of the measuring points are {100, 120, 110, 130} MPa, the mean is: ; Calculate the mean square error: ; Finally, the stress overload area distribution value is obtained: ; The results indicate that the stress overload distribution value in this area is high, exceeding the set overload identification threshold of 35. Therefore, this area can be judged as a stress overload area, which means that there may be a risk of stress concentration or material fatigue in this area. Subsequent structural analysis or experimental verification can be combined to further confirm whether this area requires structural optimization or material strengthening.

[0025] Table 1 Stress measurement point data

[0026] As shown in Table 1, the stress values at different positions are recorded at this measurement point, and the stress gradient between adjacent measurement points is calculated for subsequent stress gradient calculation and overload area screening.

[0027] Please refer to Figure 3 , the shock wave propagation analysis module includes: Based on the stress overload area distribution value, the stress data analysis sub-module monitors the key parts of the chain pin and chain plate, obtains stress response data, records the data acquisition time and the corresponding stress value, and at the same time arranges the stress data in time series, calculates the time interval and stress change rate between different measurement points, screens out abnormal data and performs numerical interpolation processing to obtain a stress time series data set; In the structural monitoring system, measurement points are set, and the distance between measurement points is fixed at 5 mm. The stress response data of each measurement point is recorded by a stress sensor under the action of a shock wave. The stress data storage format includes the measurement point number, stress value (unit: MPa), and data acquisition time (unit: ms). An example of data recording is shown in Table 2. During the data processing, first, the stress values are arranged according to the time series, and the time difference between adjacent measurement points is calculated. For example, the first measurement point records a stress of 5 MPa at 0.5 ms, and the second measurement point records a stress of 4.8 MPa at 0.7 ms, then the time interval is 0.2 ms. Further calculate the stress change rate between measurement points, that is, the stress difference divided by the time interval. For example MPa / ms. When the stress change rate of a certain measurement point exceeds the set threshold of ±3 MPa / ms, it is considered that the data of this measurement point is abnormal, and the numerical interpolation method is used for correction. For example, the abnormal measurement point is filled with the average value of the previous and subsequent measurement points. If the stress of measurement point A is 4.6 MPa and the stress of measurement point C is 4.4 MPa, then the stress of interpolation point B is set to 4.5 MPa. Finally, a stress time series data set is obtained, including the stress value of each measurement point and the corresponding time information, as shown in Table 2 for details.

[0028] Table 2 Stress time series data table of monitoring points

[0029] Table 2 lists the stress data distribution between measurement points. The data has been arranged in time series and processed by removing abnormal data and interpolation to ensure the integrity and accuracy of the data.

[0030] The propagation speed calculation sub-module calls the stress time series data set, calculates the time difference of the shock wave arrival time between different measurement points, calculates the local propagation speed based on the measurement point distance, and at the same time performs weighted summation processing on the propagation speeds of all measurement points to obtain the average propagation speed. The formula is: ; Calculate the shock wave propagation rate of multiple measurement points through operations, and calculate the global propagation speed to obtain the average shock wave propagation speed; Among them, represents the average propagation velocity of the shock wave, represents the spacing between the th time difference of the shock wave arrival at the th measurement point, represents the standard deviation of the shock wave arrival time differences at all measurement points, represents the energy loss coefficient of the th measurement point, represents the total number of measurement points, represents the time difference between measurement points; First, calculate the time difference of the shock wave arrival at every two adjacent measurement points. For example, if the recording time at measurement point 1 is 0.5 ms and the recording time at measurement point 2 is 0.7 ms, then the time difference ms. Calculate the propagation velocity for all measurement points by dividing the spacing between adjacent measurement points (5 mm) by the time difference. For example, mm / ms. Calculate the weighted sum of the velocities of all measurement points to obtain the average propagation velocity. The calculation formula is as follows: where, is the spacing between measurement points of 5 mm, is the time difference between measurement points, and the standard deviation Calculate the standard deviation of all time differences. The energy loss coefficient is set to 0.02, and the total number of measurement points . Calculate the propagation velocity of each measurement point: ; The standard deviation is calculated as follows: ; Substitute the data , the mean value , to obtain: ; Calculate the weighted sum: ; ; This result indicates that the average propagation velocity of the shock wave under the current measurement point distribution is 7.07 mm / ms, indicating that the shock wave maintains a stable propagation velocity during propagation and does not exhibit obvious energy attenuation or abnormal propagation phenomena.

[0031] The shock energy analysis sub-module calls the average propagation velocity of the shock wave, combines it with the stress changes in the stress time series dataset, calculates the energy loss per unit propagation distance, and simultaneously calculates the total energy loss of the shock wave based on the stress attenuation rate of multiple measurement points, and accumulates the energy losses of all measurement points to obtain the shock wave energy accumulation coefficient.

[0032] Based on the average propagation speed mm / ms, and combined with the stress time series data to calculate the energy loss. First, the energy loss per unit propagation distance is calculated as follows: ; For example, the stress change from measuring point 1 to measuring point 2 MPa, then the energy loss per unit propagation distance: ; Calculate all measuring points: ; Total energy loss of the shock wave: ; The cumulative energy loss coefficient is calculated as follows: ; The result shows that the shock wave will have energy loss due to the influence of the medium during propagation, and the final cumulative loss is 0.8 MPa, which means that the shock wave gradually attenuates during the contact process between the chain pin and the link plate, and the influence on the subsequent shock effect gradually decreases. At the same time, it can be used as an important basis for subsequent energy compensation and shock wave optimization.

[0033] Please refer to Figure 4 , the vibration feature extraction module includes: The energy accumulation calculation sub-module calculates the shock wave energy distribution of the vibration signal under different time windows based on the shock wave energy accumulation coefficient, extracts the vibration amplitudes under multiple windows, and calculates the cumulative energy per unit time. After normalization, the energy peak distribution under different processes is obtained, and the shock wave energy peak data is generated; First, the collected vibration signal is segmented, and the length of each time window is set to 0.05 seconds. The instantaneous energy is calculated based on the vibration amplitude in the time domain. The specific calculation method is to sum the squares of the amplitudes of the vibration signal within the time window. For example, within the window time If the vibration signal amplitude data is , then the instantaneous energy of this window is: ; And so on, after calculating all windows, sum the instantaneous energies of each time window to form the cumulative energy per unit time, and then perform normalization processing to make the energy values under different processes within the range of [0,1]. The normalization uses the maximum-minimum normalization method, that is: ; Assume that the minimum energy value of the vibration signal is 0.05 and the maximum energy value is 1.2. Then, the normalized energy value of 0.1724 for a certain window is as follows: ; After processing all windows, the energy peak distribution is obtained, and then the energy peak characteristics of each process are identified. Finally, the shock wave energy peak data is formed. The result shows that there are differences in the energy distribution of the vibration signal under different time windows, and the impact energy changes under different processes can be classified through the normalized energy peak data, providing basic data for subsequent characteristic frequency screening.

[0034] The characteristic frequency screening sub-module calls the shock wave energy peak data, uses the short-time Fourier transform to distribute energy, screens the key frequency bands corresponding to the energy peaks, calculates the contribution rates of multiple frequency bands, removes the frequency components with contribution rates lower than the set threshold, and uses the formula: ; [[ID=1"2]] Calculate the characteristic frequency screening value, obtain the characteristic frequency ranges of the chain impact riveting process and the chain tensioning assembly process, and generate the process characteristic frequency data; Among them, represents the characteristic frequency screening value, represents the th energy value of the frequency band, represents the th center frequency of the frequency band, represents the th amplitude of the vibration signal of the time window, represents the th time length of the time window, represents the total number of frequency bands, represents the number of frequency bands for contribution calculation, represents the total number of time windows; Set the window length to 256 points and the step size to 128 points. Calculate the frequency energy distribution of each time period, and screen out the key frequency bands corresponding to the energy peaks. For example, within a certain time window, the calculated energy peaks appear in the 300 Hz, 450 Hz, and 700 Hz frequency bands, and the corresponding energy values are respectively. Then, the contribution rate is calculated as follows: ; For the contribution rate of 300 Hz: ; The contribution rate of 450 Hz: ; The contribution rate of 700 Hz: ; Set the contribution rate threshold to 0.2, and frequencies below this value are removed. Therefore, 700 Hz is removed, and only 300 Hz and 450 Hz are retained. Then, calculate the final characteristic frequency screening value: ; If the amplitude of the vibration signal is , and the window time is 0.05 seconds, then: ; Substitute the screened frequency values: ; Finally, obtain the characteristic frequency range of the chain impact riveting process and the chain tensioning assembly process, and form the process characteristic frequency data. This result shows that the characteristic frequency screening value can reflect the main energy distribution of the vibration signal. The calculated 11286 indicates that in the riveting and tensioning processes, the main energy is concentrated at 300 Hz and 450 Hz, and the contribution rate of the vibration signal meets the screening threshold, further proving that this process has stable vibration characteristics and providing reliable characteristic frequency data for the subsequent extraction of the force mode.

[0035] The force mode extraction sub-module calls the process characteristic frequency data, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning assembly process, calculates the spectral response characteristic values under multiple processes, performs normalization processing, matches the vibration mode categories, and establishes the chain force mode characteristic values.

[0036] For example, in the impact riveting process, the characteristic frequencies are mainly distributed at 300 Hz and 450 Hz, while in the tensioning assembly process, they are mainly distributed at 250 Hz and 500 Hz. To perform mode matching, calculate the spectral response characteristic values and use the normalization calculation method: ; Assume that the characteristic frequency value of a certain test is 350 Hz, then after normalization: ; Calculate all the characteristic frequency data in this way, and perform vibration mode classification. Match the test data with the existing modes, and finally establish the chain force mode characteristic value data. This result shows that the characteristic frequency ranges under different processes can be normalized through the spectral response characteristic values. The value 0.4 indicates that the characteristic frequency of this test signal is closer to the riveting process and has a higher matching degree with the characteristics in the process database. Therefore, it can be judged that this test signal is more likely to come from the impact riveting process, thus providing a basis for the identification of the process state.

[0037] Table 3: Shock wave energy peak data

[0038] As shown in Table 3, the vibration signal amplitudes, the calculated instantaneous energy values, and the energy values after normalization under multiple time windows are recorded. These data are used for subsequent characteristic frequency screening and force mode extraction analysis. The results show that the vibration signal energy presents a regular distribution under different time windows. Combining with spectrum analysis can further determine the characteristic frequency distribution, laying a foundation for the matching of the chain force mode.

[0039] Please refer to Figure 5 , the load optimization control module includes: Based on the characteristic values of the chain force mode, the force parameter analysis sub-module calls the simulated operation data of the chain, obtains the force data under different operating states, extracts the force change trends of multiple nodes of the chain, calculates the force mean value, the fluctuation range, and the force change rate under different operating conditions of the chain, and obtains the force characteristic parameters of the chain; First, call the simulation data of the chain under different operating states, and analyze the force conditions within each time period. For the force data of the chain under different operating conditions, first divide it into multiple time periods according to the time series. For example, take 0.5 seconds as a time interval, and the force value of the chain in each time period is recorded and stored in the data table. Subsequently, calculate the mean value and the standard deviation of the force data for all time periods, that is, calculate the arithmetic mean of each force value and the degree of deviation of the force value in each time period relative to the average value. Further, based on the force changes of multiple nodes within each time period, extract the force fluctuation trend. The specific method is to calculate the force change rate between two consecutive time periods for each node, that is , where represents the time interval. By statistically analyzing the force change rates of each node within all time periods, obtain the force change trend of the entire chain under different operating conditions. After the data is sorted out, by setting the force mean value and its fluctuation range, that is times (such as ), screen out the abnormal force values, and at the same time calculate the maximum force change rate to analyze the dynamic force characteristics of the chain. For example, under a certain operating condition, the force mean value is 800 N and the standard deviation is 50 N, then the force fluctuation range can be set to 700 N - 900 N. If the force value in a certain time period exceeds this range, such as 950 N, it is considered that the chain may be in a sudden force state during this time period. At this time, extract the force change rates of the adjacent time periods before and after it. If the change rate exceeds the set threshold (such as 500 N / s), it is identified as a force mutation point, and then summarize the force characteristic parameters. The results show that the force fluctuation characteristics of the chain in different time periods can be quantified by calculating the mean value and the standard deviation, and the analysis based on the force change rate can effectively identify abnormal force conditions, which provides key data support for the subsequent calculation of the short-term force limit of the chain.

[0040] The short-term force limit calculation sub-module of the chain calls the force characteristic parameters of the chain to analyze the force states of different time periods of the chain, and uses the formula: ; Calculate the short-term force limit of the chain, adjust the short-term force threshold, and generate the short-term force limit value of the chain; Among them, represents the current short-term force limit of the chain, represents the actual force value in the th time period, represents the average force value, represents the standard deviation of the force value, represents the time point of each time period, represents the optimal time point, represents the standard deviation of time, represents the total number of time periods, represents the number of time periods for short-term force calculation; First, determine the calculation range of the short-term force limit based on the aforementioned obtained force characteristic parameters of the chain, and then call the formula: ; Among them, represents the force value in the th time period. For example, select a set of force data from the aforementioned calculation as , the average force , the standard deviation . For the time point , assume that the optimal force point in a certain time period is set to 2.5 s, the time standard deviation , calculate the force limit value at a certain moment: 1. Calculate the normalization of the force difference in each time period: ; The calculation result is .

[0041] 2. Calculate the time weight term: ; For example, the corresponding to the time period are [2.1, 2.3, 2.5, 2.7, 2.9] seconds respectively, and calculate its exponential weight: ; 3. Calculate the final short-term force limit: ; ; The result shows that the calculated value is used to measure the ultimate force-bearing range that the chain can withstand in a short period of time. The magnitude of the value determines whether the force threshold needs to be adjusted under specific operating conditions of the chain to ensure a reasonable force state of the chain and reduce the risk of sudden force changes.

[0042] The chain load distribution optimization sub-module calls the short-term ultimate force value of the chain, adjusts the chain test parameters, compares the matching degree between the differential chain force thresholds and the short-term ultimate force values, screens the optimal test parameters, and generates the optimized chain load distribution value.

[0043] First, call the previously calculated short-term ultimate force value of the chain , set the initial force threshold for each operating condition. For example, if the set force threshold of the chain is 850 N and the short-term ultimate force is , then calculate its threshold correction amount . During the subsequent chain load optimization process, for different test parameters, such as chain tension, drive wheel torque, lubrication state, etc., adjust the test parameters item by item. For example, reduce the drive wheel torque by 10 Nm, or adjust the chain tension to the set optimization range (such as 600 N - 700 N), and then compare the force conditions of the chain after adjusting each parameter. Based on the short-term ultimate force value calculate whether the force distribution is reasonable, and select the optimal test parameters through multiple rounds of optimization. Finally, generate the optimized chain load distribution value. The result shows that by optimizing the load distribution, the operating state of the chain can be ensured within the short-term ultimate force range, reducing the impact of overloading or sudden force changes on the chain performance, thereby improving the service life and operating stability of the chain.

[0044] Please refer to Figure 6 , the fracture risk warning module includes: The load distribution calculation sub-module, based on the optimized chain load distribution value, calls the force conditions of multiple-node chain links, calculates the dynamic load distribution on the chain links, analyzes the load transfer relationship between differential chain links, and adjusts the local load deviation to obtain the optimized chain link load distribution coefficient; In practical applications, when the chain is under load, the loads on each link are different. Especially under different working conditions, these links may experience different load fluctuations. For example, during the speed change process of the chain, the force states of the links may vary significantly due to torque changes. Therefore, during the execution of this module, it is first necessary to conduct a detailed mechanical analysis of the chain under the current load state. By using finite element analysis or directly measuring the forces on each node of the link (including vertical forces, tangential forces, etc.), and inputting this data into the calculation model, further calculate the dynamic load distribution of each link. This process requires multi-level parameter calibration and load calculation based on the physical properties of the link, the movement speed, and the external load on the chain, so as to ensure that the load deviation of each link is minimized as much as possible, making the chain more stable during actual use. During the specific execution process, assume that the initial loads of the links are respectively and and etc., and calculate the change trends of these loads during the movement of the chain. Through comprehensive optimization of the load characteristics of the chain, the final load distribution coefficient of the link approaches an equilibrium state. During this process, methods such as dynamically adjusting the connection angle of the link and using different materials to enhance the bearing capacity can be adopted to optimize the load distribution. Finally, calculate the optimized load distribution coefficient of the link through calculation, such as coefficients and and etc. This optimization result can be used for subsequent impact energy analysis and directly affects the fatigue life and fracture risk assessment of the chain.

[0045] The impact energy accumulation sub-module calls the optimized load distribution coefficient of the link, obtains the shock wave feedback data, calculates the distribution of the impact energy on the link, and analyzes the transmission mode of the impact energy on the link according to the change trend of the accumulated impact energy. Use the formula: ; Calculate and obtain the cumulative value of the impact energy; where represents the cumulative value of the impact energy, represents the load at the link location, represents the speed at the link location, represents the mass at the link location, represents the total number of impact energy calculation nodes; In practical applications, the impact energy of the chain often comes from sudden external load changes. For example, when the equipment starts or stops, the chain will encounter instantaneous impacts. In this module, first, it is necessary to obtain the shock wave feedback data that the chain may encounter during operation through multiple measurements. These data can be obtained through strain gauges or pressure sensors. In actual calculations, it is assumed that each node of the chain is subjected to different impact loads. Assume that at a certain moment, the impact load of a link is , the speed is , and the mass is . At this time, through the formula: ; the impact energy can be calculated. For example, assume that there are a total of 3 nodes to be calculated, and the given loads, speeds, and masses are respectively: , , ; , , ; , , , then calculate the impact energy of each node one by one: ; ; ; The total impact energy is: ; According to this cumulative impact energy, the next step is to analyze the change trend of the impact energy and provide a basis for subsequent load adjustment and fracture risk assessment based on the change trend.

[0046] The fracture risk assessment sub-module calls the cumulative value of the impact energy, determines whether it exceeds the set load safety threshold, compares the stress states within the threshold range, adjusts the load distribution of the links, calculates the fracture risk level of the entire chain, and obtains the fracture risk warning value of the chain.

[0047] In actual operation, setting a reasonable load safety threshold is crucial for ensuring the service life of the chain. When setting the safety threshold, it can be based on empirical data, historical tests, or obtained through simulation experiments. For example, assuming the safety threshold is 3000 J, when the impact energy exceeds this threshold, there may be a relatively high risk of fracture. Therefore, when the impact energy reaches 3642.26 J, a risk assessment mechanism should be initiated to analyze the stress state of each node of the chain. By adjusting the load distribution of the chain links (such as reducing the load on the nodes with heavier loads and increasing the load on the nodes with lighter loads), the overall energy transfer of the chain can be made more uniform, reducing the risk of overload. At the same time, the fracture risk level of the chain can be calculated through a series of test data and mathematical models. For example, it is set that when the impact energy is greater than 3000 J, the fracture risk level is "high", when it is greater than 2500 J, it is "medium", and when it is less than 2500 J, it is "low". At this time, since the cumulative impact energy is 3642.26 J, the final assessment result is "high", and the warning system should be triggered to remind the operator to check or maintain the chain.

[0048] The above are only the preferred embodiments of the present invention and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A chain production data monitoring and optimization system based on the industrial Internet of Things, characterized in that, The system includes: The stress load monitoring module arranges stress sensors in the chain plate stamping forming area and the pin heat treatment area, collects transient data, determines the measuring point positions by combining coordinate indexing, calculates the stress change rate, obtains the gradient by using numerical differentiation, screens local extreme points, and obtains the stress overload area distribution value; The shock wave propagation analysis module obtains the vibration data of the key parts of the chain pin and the chain plate according to the stress overload area distribution value, calculates the propagation speed and the attenuation rate, conducts a comparative analysis of the waveforms, and generates the shock wave energy accumulation coefficient; The vibration feature extraction module obtains the vibration signals during production based on the shock wave energy accumulation coefficient, distributes the energy by using the short-time Fourier transform, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning assembly process, and generates the chain force mode characteristic value; The load optimization control module constructs a stress accumulation curve by calling the chain simulation operation data based on the chain force mode characteristic value, calculates the short-term force limit of the chain, adjusts the chain test parameters, and generates the optimized chain load distribution value.

2. The chain production data monitoring and optimization system based on the industrial Internet of Things according to claim 1, wherein The stress overload area distribution value includes the measuring point position, the stress change rate, and the local extreme points. The shock wave energy accumulation coefficient specifically refers to the propagation speed, the attenuation rate, and the waveform comparative analysis result. The chain force mode characteristic value includes the characteristic frequencies of the impact riveting process and the tensioning assembly process, and the energy distribution. The optimized chain load distribution value specifically refers to the stress accumulation curve, the short-term force limit, and the test parameter adjustment scheme.

3. The chain production data monitoring and optimization system based on industrial Internet of Things according to claim 2, wherein The stress load monitoring module includes: The stress data acquisition sub-module arranges stress sensors in the chain plate stamping forming area and the pin heat treatment area, obtains transient data, records the stress values of multiple measuring points, and determines the positions of the measuring points by combining coordinate indexing, generating the measuring point stress distribution data; The stress gradient calculation sub-module calculates the stress difference between adjacent measuring points based on the measuring point stress distribution data, calculates the spatial displacement according to the measuring point coordinates, and simultaneously calculates the stress change rate by using numerical differentiation, generating the measuring point stress change rate data; The stress overload area screening sub-module calls the measuring point stress change rate data, screens local extreme points, extracts the stress gradient change in the overload area, and uses the formula: ； Performs operations to obtain the stress distribution value in the overload area, and obtains the stress overload area distribution value; Among them, represents the stress difference between adjacent measurement points, represents the spatial distance between adjacent measurement points, represents the stress gradient weight parameter, represents the stress value of the measurement point, represents the average stress of all measurement points, represents the distribution value of the stress overload area, represents the number of stress gradient calculation points, represents the total number of measurement points, i represents a certain measurement point, j represents another measurement point related to i in a certain calculation step, and id represents the unique identifier of the measurement point.

4. The chain production data monitoring and optimization system based on the industrial Internet of Things according to claim 3, wherein The shock wave propagation analysis module includes: The stress data analysis sub-module monitors the key parts of the chain pin and the chain plate based on the stress overload area distribution value, obtains the stress response data, records the data acquisition time and the corresponding stress values, and simultaneously arranges the stress data in time series, calculates the time interval and the stress change rate between different measuring points, screens abnormal data and conducts numerical interpolation processing, obtaining the stress time series data set; The propagation speed calculation sub-module calls the stress time series data set, calculates the shock wave arrival time difference between different measuring points, calculates the local propagation speed according to the measuring point spacing, and simultaneously conducts weighted summation processing on the propagation speeds of all measuring points, obtaining the average propagation speed, and using the formula: ； Performs operations to obtain the shock wave propagation rates of multiple measuring points, and calculates the global propagation speed, obtaining the average shock wave propagation speed; Among them, represents the average propagation velocity of shock waves, represents the spacing of the th measurement point, represents the time difference of the shock wave arrival at the th measurement point, represents the standard deviation of the time differences of the shock wave arrivals at all measurement points, represents the energy loss coefficient of the th measurement point, represents the total number of measurement points, represents the time difference between measurement points; The impact energy analysis sub-module calls the average propagation velocity of the shock wave, combines the stress changes in the stress time series data set, calculates the energy loss per unit propagation distance, and simultaneously calculates the total energy loss of the shock wave based on the stress attenuation rate of multiple measurement points, and accumulates the energy losses of all measurement points to obtain the shock wave energy accumulation coefficient.

5. The chain production data monitoring and optimization system based on the industrial Internet of Things according to claim 4, wherein The vibration feature extraction module includes: The energy accumulation calculation sub-module calculates the shock wave energy distribution of the vibration signal under different time windows based on the shock wave energy accumulation coefficient, extracts the vibration amplitudes under multiple windows, calculates the accumulated energy per unit time, and obtains the energy peak distribution under different processes after normalization processing, generating shock wave energy peak data; The characteristic frequency screening sub-module calls the shock wave energy peak data, uses the short-time Fourier transform to distribute energy, screens the key frequency bands corresponding to the energy peaks, calculates the contribution rates of multiple frequency bands, removes the frequency components with contribution rates lower than the set threshold, and uses the formula: ； Calculate the characteristic frequency screening value, obtain the characteristic frequency ranges of the chain impact riveting process and the chain tensioning assembly process, and generate process characteristic frequency data; Among them, represents the characteristic frequency screening value, represents the energy value of the th frequency band, represents the central frequency of the th frequency band, represents the amplitude of the vibration signal of the th time window, represents the number of frequency bands for contribution calculation, represents the total number of time windows. i represents the measurement point number, frequency band number, or time window number, and E represents the energy value of a certain frequency band; The force mode extraction sub-module calls the process characteristic frequency data, screens the characteristic frequencies of the chain impact riveting process and the chain tensioning assembly process, calculates the spectral response characteristic values under multiple processes, performs normalization processing, matches the vibration mode categories, and establishes the chain force mode characteristic values.

6. The chain production data monitoring and optimization system based on the industrial Internet of Things according to claim 5, characterized in that, The load optimization control module includes: The force parameter analysis sub-module calls the chain simulation operation data based on the chain force mode characteristic values, obtains the force data under different operating states, extracts the force change trends of multiple nodes of the chain, calculates the force mean value, fluctuation range, and force change rate under different operating conditions of the chain, and obtains the chain force characteristic parameters; The short-term force limit calculation sub-module of the chain calls the chain force characteristic parameters, analyzes the force states of the chain in different time periods, and uses the formula: ； Calculate the short-term force limit of the chain, adjust the short-term force threshold, and generate the short-term force limit value of the chain; in, Represents the current short-term stress limit of the chain, Representatives in the The actual force value within a time period, Represents the average force value, Represents the standard deviation of the force values, Represents the time point of each time period, represents the optimal time point, represents the standard deviation of time, Represents the total number of time periods, Represents the number of short-term force calculation time periods; The chain load distribution optimization sub-module calls the short-term force limit value of the chain, adjusts the chain test parameters, compares the matching degree between the different chain force thresholds and the short-term limit force values, screens the optimal test parameters, and generates the optimized chain load distribution value.

7. The chain production data monitoring and optimization system based on industrial Internet of Things according to claim 6, characterized in that The system further includes: The fracture risk warning module obtains the shock wave feedback data based on the optimized chain load distribution value, calculates the impact energy cumulative value, judges the load safety threshold, compares the force states within the threshold range, adjusts the load distribution, and generates the chain fracture risk warning value; The chain fracture risk warning value includes the impact energy cumulative value, the load safety threshold, and the force state comparison result.

8. The chain production data monitoring and optimization system based on the industrial Internet of Things according to claim 7, characterized in that, The fracture risk warning module includes: The load distribution calculation sub-module calls the force conditions of multiple-node links based on the optimized chain load distribution value, calculates the dynamic load distribution on the links, analyzes the load transfer relationship between different links, and adjusts the local load deviation to obtain the optimized link load distribution coefficient; The impact energy accumulation sub-module calls the optimized link load distribution coefficient, obtains shock wave feedback data, calculates the distribution of impact energy on the link, analyzes the transfer mode of impact energy on the link according to the change trend of the accumulated impact energy, and uses the formula: ； Calculate and obtain the cumulative value of impact energy; Among them, represents the cumulative value of impact energy, represents the load at the link location, represents the speed at the link location, represents the mass at the link location, represents the total number of impact energy calculation nodes; The fracture risk assessment sub-module calls the cumulative value of the impact energy, determines whether it exceeds the set load safety threshold, compares the stress states within the threshold range, adjusts the link load distribution, calculates the fracture risk level of the whole chain, and obtains the fracture risk warning value of the chain.

Images