A corrugated cardboard strength testing system

Abstract

The present application relates to the technical field, and disclose a kind of corrugated board strength test system, system includes: multi-dimensional environment simulation module, composite sensing detection module, abnormal condition identification module, collaborative influence quantification module, intelligent correction module and closed-loop detection module.This system not only solves the core technical problem of low precision of corrugated board strength detection in humid salt-containing environment, but also has the advantages of strong scene adaptability, full detection dimension and high intelligent degree, providing a new technical solution for corrugated board quality control in special environments such as coastal transportation and marine transportation.It can be widely used in packaging material production enterprises, logistics transportation enterprises and quality detection institutions, and has important industrial application value and promotion prospect.

Description

Technical Field

[0001] This invention relates to the field of corrugated board strength testing technology, specifically a corrugated board strength testing system. Background Technology

[0002] When corrugated cardboard is used in coastal transportation, ocean transportation, and high salt spray industrial environments, it faces not only the problem of moisture absorption caused by high humidity, but also the synergistic effect of physical adsorption and chemical erosion of salt ions. Salt ions not only reduce the bonding force between paper fibers, but also damage the internal structure of the fibers through penetration, while accelerating the aging of the adhesive layer, resulting in a cardboard strength decay rate far exceeding that of a single humidity environment.

[0003] Existing testing systems suffer from three major shortcomings: First, they focus solely on humidity, neglecting key influencing factors such as salt ion penetration depth and surface / internal structural damage, thus failing to reproduce realistic salt-humidity synergistic erosion scenarios. Second, they employ a limited range of testing dimensions, often only measuring edge crush strength while ignoring comprehensive mechanical properties like compressive strength and burst strength, and lacking quantitative methods for assessing internal structural damage. Third, they utilize linear models and simple error correction methods, making it difficult to handle the nonlinear synergistic effects of salt, humidity, and structural damage, resulting in low prediction accuracy and insufficient creativity. Therefore, there is an urgent need to construct a multi-dimensional, nonlinear, and scenario-adaptive corrugated cardboard strength testing system. Summary of the Invention

[0004] (a) Technical problems to be solved

[0005] To address the shortcomings of existing technologies, this invention provides a corrugated cardboard strength testing system to solve the aforementioned problems.

[0006] (II) Technical Solution

[0007] To achieve the above objectives, the present invention provides the following technical solution: a corrugated cardboard strength testing system, the system comprising: a multi-dimensional environment simulation module, a composite sensing detection module, an abnormal working condition identification module, a collaborative influence quantification module, an intelligent correction module, and a closed-loop detection module;

[0008] The multi-dimensional environment simulation module is used to simulate humid and saline environments with different humidity, salt concentration and salt ion penetration depth. It can reproduce the salt-humidity synergistic erosion scenario through gradient control and place the corrugated cardboard to be tested in the environment.

[0009] The composite sensing and detection module is used to simultaneously collect multi-dimensional physical parameters (moisture content, salt adsorption, salt ion erosion depth, ultrasonic propagation speed, ultrasonic attenuation coefficient), mechanical performance parameters (actual edge crush strength, actual compressive strength, actual burst strength), and predicted mechanical performance parameters of corrugated cardboard, while also collecting environmental parameters (humidity, salt concentration).

[0010] The abnormal working condition identification module is used to calculate the residual distribution between the predicted mechanical performance parameters and the actual mechanical performance parameters under each humid and saline environment by combining residual analysis with the 3σ criterion, and to screen out abnormal working conditions whose residuals exceed the 3σ range as the environment to be analyzed.

[0011] The synergistic impact quantification module is used to extract the core impact factors of multi-dimensional physical parameters based on principal component analysis (PCA), calculate the correlation weight between the core impact factors and the mechanical performance degradation through grey relational analysis, and combine the degree of internal structural damage reflected by ultrasonic testing to quantify the comprehensive impact of corrugated cardboard under each environment to be analyzed.

[0012] The intelligent correction module is used to predict mechanical performance parameters based on the output of the random forest regression model. It takes core influencing factors, environmental parameters and comprehensive impact as inputs, calculates the correlation strength between environmental parameters and prediction residuals through partial correlation coefficients, and adaptively corrects the predicted mechanical performance parameters based on the correlation strength and correlation weights to obtain the corrected mechanical performance parameters.

[0013] The closed-loop testing module is used to realize online testing of the strength of corrugated cardboard, pass / fail determination, and real-time feedback and updating of test data based on the corrected mechanical performance parameters.

[0014] As a preferred technical solution of the invention, the multi-dimensional environment simulation module includes a constant temperature and humidity salt spray infiltration chamber, a salt concentration gradient control unit, a humidity precision control unit, and an infiltration depth control unit.

[0015] The salt concentration gradient control unit precisely mixes NaCl solution and deionized water using dual-channel infusion pumps, achieving a continuously adjustable salt concentration of 0.3% to 6% by mass, with a control accuracy of ±0.05%.

[0016] The humidity control unit uses a combination of steam humidification and semiconductor dehumidification to achieve relative humidity regulation of 35%~98%, with a control accuracy of ±1%RH.

[0017] The penetration depth control unit, by regulating the ambient temperature and exposure time, combined with negative pressure assisted penetration technology, achieves gradient control of the salt ion erosion depth;

[0018] The constant temperature and humidity salt spray infiltration chamber is equipped with a salt spray uniform injection system, and the temperature, humidity and salt concentration uniformity error inside the chamber is ≤3%.

[0019] As a preferred technical solution of the invention, the composite sensing and detection module includes: a multi-parameter physical detection unit, a comprehensive mechanical detection unit, an environmental parameter detection unit, and a prediction model training unit;

[0020] The multi-parameter physical detection unit includes: a near-infrared moisture analyzer, an ion chromatograph, an X-ray fluorescence spectrometer, and an ultrasonic flaw detector;

[0021] The comprehensive mechanical testing unit uses a multi-station mechanical testing machine to simultaneously test edge crush strength, compressive strength, and bursting strength.

[0022] The environmental parameter detection unit includes a high-precision temperature and humidity sensor and an online salt concentration sensor;

[0023] The prediction model training unit collects information on the corrugation type, board thickness, ring crush strength, corrugation density, and sizing amount of corrugated cardboard. After standardization, the data is input into the random forest regression model for training, and the model outputs predicted edge crush strength, predicted compressive strength, and predicted burst strength.

[0024] As a preferred embodiment of the invention, the operation process of the abnormal operating condition identification module includes:

[0025] The residual between the predicted mechanical performance parameters and the actual mechanical performance parameters under each environment is calculated using the following formula: ;

[0026] Calculate the mean (μ) and standard deviation (σ) of the residual sequence, and construct a residual normal distribution model;

[0027] Environments with residuals exceeding the range of [μ−3σ,μ+3σ] are identified as environments to be analyzed, i.e., abnormal operating conditions with significant synergistic effects of salt and humidity.

[0028] As a preferred technical solution of the invention, the working process of the synergistic influence quantification module includes: standardizing multi-dimensional physical parameters (water content M, salt adsorption S, salt ion erosion depth D, ultrasonic propagation velocity V, and ultrasonic attenuation coefficient A), and extracting the first three principal components as core influencing factors through principal component analysis. Cumulative contribution rate ≥90%;

[0029] The formula for calculating the attenuation rate of actual mechanical performance parameters compared to those under standard conditions is as follows: ;

[0030] The correlation weights between each core influencing factor and the decay rate η were calculated using grey relational analysis. ,in The sum is 1;

[0031] Internal structure damage index constructed based on ultrasonic propagation velocity and ultrasonic attenuation coefficient: ,in Ultrasonic parameters under standard conditions. These are the weighting coefficients. ;

[0032] Overall impact , where α is the structural damage correction factor, with a value of 0.3 to 0.5.

[0033] As a preferred technical solution of the invention, the extraction process of the core influencing factors includes: constructing a raw data matrix of multi-dimensional physical parameters. Where m is the number of samples and n is the number of parameters;

[0034] The original data matrix is ​​standardized to obtain the standardized matrix. ;

[0035] Calculate the covariance matrix C of the normalized matrix and solve for the eigenvalues ​​of the covariance matrix. and the corresponding feature vectors ;

[0036] Sort the eigenvalues ​​from largest to smallest, select the eigenvectors corresponding to the top 3 eigenvalues ​​to construct the principal component loading matrix, and obtain the core influence factor through linear combination. .

[0037] As a preferred embodiment of the invention, the working process of the intelligent correction module includes:

[0038] The partial correlation coefficients between the residual sequence of predicted mechanical performance parameters and each core influencing factor are calculated, and the correlation strength R between environmental parameters and predicted residuals is obtained, with a value range of 0~1.

[0039] Based on the predicted value Y output by the random forest regression model, combined with the association strength R and association weights, the prediction is made. And calculate the correction amount based on the overall impact level Q. ;

[0040] Corrected mechanical property parameters .

[0041] As a preferred embodiment of the invention, the closed-loop detection module includes: a real-time transmission unit, a data fusion unit, a qualification determination unit, and a model update unit;

[0042] The real-time transmission unit adopts industrial Ethernet + LoRa dual-mode communication to achieve stable transmission of detection data with a latency of ≤50ms;

[0043] The data fusion unit performs spatiotemporal alignment and redundancy removal on the multi-source data from the composite sensing detection module to generate a unified detection dataset.

[0044] The pass / fail judgment unit compares the corrected mechanical performance parameters with the preset threshold and outputs the pass / fail judgment result and an analysis of the cause of the abnormality.

[0045] Every 100 sets of valid detection data accumulated, the model update unit automatically performs incremental training on the random forest regression model and updates the model parameters to improve long-term detection accuracy.

[0046] As a preferred technical solution of the invention, the composite sensing detection module further includes a surface morphology detection unit, which uses a laser confocal microscope to detect the surface roughness of the corrugated cardboard, in order to help determine the degree of surface structure damage caused by salt-humidity erosion.

[0047] As a preferred technical solution of the invention, it also includes a full lifecycle data management module, which is used to store environmental parameters, multi-dimensional detection data, correction data, judgment results and model update logs, and supports data traceability, trend analysis and abnormal working condition playback.

[0048] Compared with the prior art, the present invention provides a corrugated cardboard strength testing system, which has the following advantages:

[0049] This invention accurately reproduces the synergistic erosion scenario of salt, humidity, and penetration depth through a multi-dimensional environmental simulation module, overcoming the limitations of traditional single-humidity environmental simulation; the composite sensing and detection module achieves multi-source synchronous acquisition of physical parameters, mechanical performance parameters, and environmental parameters, filling the gap in the quantitative detection of internal structural damage; the abnormal working condition identification module uses residual analysis combined with the 3σ criterion to improve the accuracy of abnormal working condition identification; the synergistic influence quantification module achieves scientific quantification of multi-factor synergistic influence through principal component analysis and grey relational analysis; the intelligent correction module constructs an adaptive correction mechanism based on random forest regression and partial correlation analysis, significantly improving prediction accuracy; and the closed-loop detection module and the full life cycle data management module form a complete quality control and data traceability system.

[0050] Compared with existing technologies, this system not only solves the core technical problem of low accuracy in strength testing of corrugated cardboard in humid and saline environments, but also has the advantages of strong scene adaptability, comprehensive testing dimensions, and high degree of intelligence. It provides a brand-new technical solution for the quality control of corrugated cardboard in special environments such as coastal and marine transportation. It can be widely used in packaging material manufacturers, logistics and transportation companies, and quality testing institutions, and has important industrial application value and promotion prospects. Attached Figure Description

[0051] Figure 1 This is the overall system architecture diagram (core logic diagram) of the present invention;

[0052] Figure 2 This is a diagram showing the internal structure of the multi-dimensional environment simulation module of the present invention;

[0053] Figure 3 This is a diagram showing the internal structure of the composite sensing and detection module of the present invention.

[0054] Figure 4 This is a diagram showing the internal structure of the closed-loop detection module of the present invention;

[0055] Figure 5 This is a schematic diagram of the system workflow of the present invention;

[0056] Figure 6 This is a schematic diagram of the three-dimensional influence matrix of salt-humidity-penetration depth of the present invention;

[0057] Figure 7 This is a flowchart illustrating the steps for calculating the overall impact of the present invention. Detailed Implementation

[0058] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only a part of the embodiments of the invention, and not all of them. Unless otherwise specified, the embodiments and features described in this application can be combined with each other. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0059] It should be noted that if the embodiments of the invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0060] Furthermore, "multiple" refers to two or more. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of a person skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the invention.

[0061] This invention provides a corrugated cardboard strength testing system, characterized in that the system includes: a multi-dimensional environment simulation module, a composite sensing detection module, an abnormal working condition identification module, a collaborative influence quantification module, an intelligent correction module, a closed-loop detection module, and a full life cycle data management module.

[0062] Please refer to Figure 2. The multi-dimensional environmental simulation module is used to reproduce the synergistic erosion effect of salt, humidity, and penetration depth in a humid and saline environment. The specific implementation methods and parameter settings of each component are as follows:

[0063] The constant temperature and humidity salt spray permeation chamber uses a customized 150L chamber (made of 316L stainless steel to prevent salt spray corrosion). The internal dimensions are 500mm × 500mm × 600mm. It is equipped with a transparent observation window (quartz glass) and an adjustable sample rack (with a load capacity ≥ 5kg). The salt spray uniform spraying system uses four rotating nozzles (made of PTFE), evenly distributed around the top of the chamber. The spray angle is adjustable from 30° to 60°, and the spray pressure is precisely controlled at 0.05~0.2MPa via a pressure regulating valve, ensuring uniform distribution of salt spray within the chamber. Actual measurements show that the salt concentration difference between any two points within the chamber is ≤3%, and the humidity difference is ≤2%.

[0064] Salt concentration gradient control unit: Utilizing two Reif BT100L high-precision peristaltic pumps (flow range 0.001~100mL / min, accuracy ±0.5%), one pump delivers NaCl standard solution (10% concentration, analytical grade) and the other delivers deionized water (conductivity ≤10μS / cm). The flow ratio of the two pumps is adjusted via a PLC controller, achieving continuously adjustable mass fraction from 0.3% to 6%. For example, when the target salt concentration is 3%, the NaCl solution pump flow rate is set to 30mL / min, and the deionized water pump flow rate is set to 70mL / min. After mixing, the solution is atomized into salt mist and introduced into the chamber. The salt concentration control accuracy, after calibration, can reach ±0.05%.

[0065] Humidity Precision Control Unit: The humidification section uses a stainless steel steam generator (evaporation rate adjustable from 0 to 5 kg / h), which heats deionized water to produce saturated steam, which is then introduced into the chamber after being depressurized by a pressure reducing valve. The dehumidification section uses a semiconductor refrigeration dehumidifier (cooling power adjustable from 0 to 300 W), which removes moisture from the chamber through condensation. Temperature and humidity sensors provide real-time feedback data to the PLC, and a PID control algorithm is used to achieve precise control of relative humidity from 35% to 98%, with a control accuracy of ±1%RH and a dynamic response time ≤10 minutes (when humidity changes by 10%).

[0066] Penetration depth control unit: The ambient temperature is regulated by an internal heating element and cooling coil, ranging from 20 to 40℃ with a control accuracy of ±0.5℃. The exposure time is set via a PLC timer and can be adjusted arbitrarily within the range of 1 to 72 hours. Negative pressure assisted penetration technology uses a miniature vacuum pump (ultimate vacuum -0.08MPa) to apply a negative pressure of -0.02 to -0.05MPa through the extraction port at the bottom of the chamber, accelerating the penetration of salt ions into the corrugated cardboard. Combined with temperature and time control, a salt ion erosion depth gradient of 0.1 to 5 mm is achieved. For example, at a temperature of 35℃ and a negative pressure of -0.04MPa, an erosion depth of approximately 2 mm can be achieved after 24 hours of exposure, and approximately 3.5 mm after 48 hours of exposure, with a penetration depth error ≤ ±0.1 mm.

[0067] Please refer to Figure 3. The composite sensing detection module adopts multi-sensor fusion technology to simultaneously collect the physical parameters, mechanical performance parameters, and environmental parameters of the corrugated cardboard. The specific configuration and detection process of each unit are as follows:

[0068] S1, Multi-parameter physical detection unit, performs multi-parameter physical detection:

[0069] Near-infrared moisture analyzer (Sartorius MA37): Detection wavelength range 1.2~2.5μm, using diffuse reflectance measurement method, the distance between the detection probe and the sample surface is controlled at 5mm, detection range 0~30%, accuracy ±0.15%. During detection, the sample is laid flat on the detection platform, the probe scans three different areas on the sample surface, and the average value is taken as the moisture content data of the sample, avoiding errors caused by uneven local moisture absorption.

[0070] Ion chromatograph (Thermo Fisher ICS-5000+): Equipped with an AS-DV autosampler and an IonPac AS19 analytical column (4×250mm). Salt adsorption detection procedure: ① Cut the etched sample into 1cm×1cm fragments, accurately weigh 0.5g (accuracy ±0.0001g) and place in a centrifuge tube; ② Add 20mL of deionized water and ultrasonically extract for 30min (power 100W, frequency 40kHz); ③ Centrifuge (8000r / min, time 10min), and filter the supernatant through a 0.22μm filter membrane; ④ Detect the Cl⁻ ion concentration using the ion chromatograph, and calculate according to the formula... Calculate the salt adsorption capacity (C is the Cl⁻ concentration in mg / L, V is the extract volume in mL, and m is the sample mass in g), with a detection accuracy of ±0.005 mg / g.

[0071] X-ray fluorescence spectrometer (Brook S8 Tiger model): Rh target X-ray tube, tube voltage 50kV, tube current 60mA, detection range 0.1~5mm. Salt ion erosion depth detection principle: Based on the correlation between the characteristic fluorescence spectral intensity of salt ions (Cl⁻) and depth, the erosion depth is calculated by scanning spectral signals at different depths and combining them with a calibration curve (pre-calibrated using standard samples of known depths), with a detection accuracy of ±0.02mm. During detection, the sample is fixed on a dedicated stage, and a line scan method (scanning speed 1mm / s) is used to acquire the depth data of three parallel scan lines. The average value is taken as the final result.

[0072] Ultrasonic flaw detector (Olympus EPOCH 650): 5MHz straight probe (10mm diameter), glycerol as the coupling agent (to ensure good ultrasonic coupling). Ultrasonic propagation velocity detection: Measure the propagation time t of the ultrasound in the sample, and calculate the propagation velocity using the formula V=d / t, combined with the sample thickness d. Detection range: 1000~5000m / s, accuracy ±10m / s; Ultrasonic attenuation coefficient detection: Measure the incident wave amplitude. With the transmitted wave amplitude A, through the formula Calculate the attenuation coefficient α (d is the sample thickness), with an accuracy of ±0.5dB / cm.

[0073] Laser confocal microscope (Keyence VK-X200): 50x objective magnification, scanning range 0.1~10μm, accuracy ±0.01μm. By scanning the three-dimensional morphology of the sample surface, the surface roughness parameter Ra (arithmetic mean deviation) is calculated to help determine the degree of surface fiber detachment and adhesive layer damage caused by salt-humidity erosion.

[0074] S2, Comprehensive Mechanical Testing Unit, performs comprehensive mechanical testing:

[0075] The Instron 5982 multi-station mechanical testing machine was selected, equipped with 3 independent testing stations, which can simultaneously complete edge crush, compressive strength, and burst strength tests.

[0076] Edge crush strength test: A standard sample of 100mm×25mm is used. The test speed is 1mm / min. The maximum pressure value is collected by a pressure sensor (range 0~6000N / m), which is the edge crush strength with an accuracy of ±0.5N / m.

[0077] Compressive strength test: A standard sample of 150mm×150mm is used, the test speed is 2mm / min, the pressure sensor range is 0~10000N, and the accuracy is ±1N;

[0078] Bursting strength test: A standard sample of 100mm×100mm is used, equipped with a rubber diaphragm (1mm thick), the test speed is 3mm / min, the pressure sensor range is 0~500kPa, and the accuracy is ±0.5kPa.

[0079] S3. Environmental Parameter Detection Unit: The temperature and humidity sensor is the Vaisala HMP110, based on the capacitive sensing principle, with a humidity measurement range of 0~100%RH and an accuracy of ±0.8%RH; the salt concentration sensor is the Honeywell HSC030N, based on the conductivity measurement principle, with an accuracy of ±0.03%, and the data sampling frequency is set to 2Hz, transmitting data to the data acquisition card in real time via an RS485 interface.

[0080] S4. Predictive Model Training Unit: This unit trains the predictive model by collecting key parameters of the corrugated cardboard, including corrugation type (A / B / C / E), cardboard thickness (accuracy ±0.001mm, measured with a micrometer), ring crush strength (RCT, accuracy ±3N / m, tested with a ring crush strength tester), corrugation density (number of corrugations / 100mm, counted using machine vision), and sizing amount (g / m², measured by weighing). These parameters are used as input features and processed using Min-Max normalization (formula...). After that, the random forest regression model (100 decision trees, maximum depth 15, minimum number of sample splits 5, minimum number of sample leaf nodes 2) was input for training. The training dataset contained 500 sets of sample data under different working conditions. Five-fold cross-validation was used to optimize the model parameters. The model output predicted edge pressure strength, predicted compressive strength, and predicted burst strength. The goodness of fit of the training set was R²≥0.92.

[0081] As a concrete example, the abnormal operating condition identification module accurately identifies abnormal intensity degradation conditions caused by salt-humidity synergy through residual analysis combined with the 3σ criterion. The specific workflow is as follows:

[0082] Residual calculation: For each test sample in a humid and saline environment, calculate the residuals of the three mechanical property parameters separately, using the following formula: ;

[0083] Residual distribution modeling: Collect residual data from 50 test samples in the same batch, construct a residual sequence, and calculate the mean μ and standard deviation σ of the residual sequence. For example, if the residual sequence of the edge pressure strength of a certain batch of samples has μ=5.2N / m and σ=8.3N / m, construct a normal distribution model N(5.2,8.32);

[0084] Abnormal operating condition determination: According to the 3σ criterion, when the residual exceeds the range of [μ−3σ,μ+3σ], it is determined to be an abnormal operating condition. Continuing with the above example, the 3σ range is [5.2−3×8.3,5.2+3×8.3]=[−19.7,29.9] N / m. If the edge pressure strength residual in a certain environment is 32.5 N / m, exceeding this range, then that environment is determined to be the environment to be analyzed. This method can effectively eliminate random errors (probability only 0.27%) and accurately locate operating conditions where the salt-humidity synergistic effect is significant.

[0085] Please refer to Figures 6 and 7. The synergistic effect quantification module quantifies the synergistic effects of salt, humidity, and structural damage through principal component analysis, grey relational analysis, and structural damage fusion. The specific implementation steps are as follows:

[0086] S1. Multi-dimensional physical parameter standardization: Collect multi-dimensional physical parameters (water content M, salt adsorption S, salt ion erosion depth D, ultrasonic propagation velocity V, ultrasonic attenuation coefficient A) of the environment to be analyzed and its surrounding environment to construct an original data matrix Xm×n (m=30 samples, n=5 parameters). For example, the original data of a sample are: M=12.5, S=0.85mg / g, D=1.8mm, V=2800m / s, A=3.2dB / cm. Z-score standardization is used, and the formula is... ( Let j be the mean of the j-th parameter. (where the standard deviation is the j-th parameter), thus obtaining the standardized matrix. ;

[0087] S2, Core Impact Factor Extraction (PCA):

[0088] Calculate the covariance matrix C (size 5×5) of the standardized matrix Z. The covariance formula is: ;

[0089] Solving for the eigenvalues ​​of the covariance matrix and the corresponding feature vectors The eigenvalues ​​are sorted as follows: The cumulative contribution rate of the first three eigenvalues ​​is , and the principal component loading matrix is ​​constructed by selecting the eigenvectors corresponding to the first three eigenvalues.

[0090] The core impact factor is obtained through a linear combination: ,in Characterizing the salt-humidity synergistic erosion effect, Characterizing the internal structural damage effect, Characterizes the superposition effect of erosion depth;

[0091] S3. Calculation of Strength Attenuation Rate: Mechanical performance parameters under standard environmental conditions (50% humidity, 0% salt concentration, 0mm erosion depth) are obtained through prior calibration, such as standard edge pressure strength. Actual edge pressure strength under a certain environment to be analyzed Then the edge pressure strength attenuation rate ;

[0092] S4. Calculation of Association Weights (Grey Relational Analysis):

[0093] Using the intensity attenuation rate η as the reference sequence Core Influence Factor For comparing sequences ;

[0094] Calculate the absolute difference between the reference sequence and the comparison sequence. We obtain the maximum absolute difference Δmax and the minimum absolute difference Δmin;

[0095] Calculate the correlation coefficient (ζ is the resolution coefficient, taken as 0.5);

[0096] Association weight Calculated The sum of the weights is 1;

[0097] S5. Construction of Internal Structure Damage Index: Ultrasonic parameters under standard conditions were obtained through calibration. ,Pick The structural damage index For example, in a certain environment to be analyzed ;

[0098] S6. Calculation of the overall impact: Taking the structural damage correction factor α = 0.4, then... For example, a certain environment to be analyzed A larger Q value indicates that the corrugated cardboard is more significantly affected by the synergistic effects of salt, humidity, and structural damage.

[0099] Furthermore, the intelligent correction module, based on the random forest regression model and partial correlation analysis, achieves adaptive correction of the predicted mechanical performance parameters. The specific process is as follows:

[0100] S1. Correlation Strength Calculation: Select the residual sequence e of the environment to be analyzed and similar historical environments, along with the core influencing factors. The partial correlation coefficient is calculated to eliminate mutual interference between factors. The formula for calculating the partial correlation coefficient is as follows: (i,j,k are different core influencing factors). The calculated correlation strength R is 0.87 (the value ranges from 0 to 1, and the closer it is to 1, the stronger the correlation between the environmental parameters and the residuals).

[0101] S2. Correction Calculation: Based on the predicted value Y_predicted from the random forest regression model, combined with the association strength R and association weights. And considering the overall impact level Q, the correction formula is as follows: For example, in a certain environment to be analyzed (The coefficient here needs to be adjusted according to the actual working conditions to avoid excessive correction. In practical applications, the range can be limited by introducing a correction coefficient.)

[0102] S3. Corrected mechanical property parameters: When ΔY>0 This indicates that the predicted value is too high, and the intensity attenuation caused by the combined effect of salt and humidity needs to be deducted; when ΔY≤0, Y_{correction}=Y_{prediction}, indicating that the predicted value is close to the actual value and no correction is needed. Continuing with the above example, if... The error between the actual edge compression strength of 1850 N / m and the actual edge compression strength is only 1.9%, and the correction effect is significant.

[0103] Furthermore, please refer to Figure 4. The closed-loop detection module realizes real-time transmission, fusion processing, pass / fail determination, and model updating of detection data, forming a closed-loop optimization mechanism.

[0104] Real-time transmission unit: Employs dual-mode communication using Industrial Ethernet (TCP / IP protocol) and LoRa (spread spectrum communication technology). Industrial Ethernet is used for high-speed transmission over short distances (≤100m), with a data rate ≥100Mbps and latency ≤50ms. LoRa is used for long-distance (≤1km) wireless transmission, with a data rate ≥50kbps and strong anti-interference capabilities, ensuring stable transmission of test data in complex industrial environments. Transmitted data includes environmental parameters, multi-dimensional physical parameters, mechanical performance parameters, correction data, and judgment results.

[0105] Data fusion unit: Employs a spatiotemporal alignment algorithm to achieve synchronous matching of multi-source data. Time synchronization is based on the NTP protocol (time synchronization accuracy ≤10ms), and spatial synchronization is achieved by binding sensor numbers with sample numbers. Redundant data removal uses a variance-based screening method. The variance of multiple measurements of the same parameter is calculated, and outlier data with variances greater than a threshold (preset to 0.05) are removed, while stable data is retained to generate a unified detection dataset.

[0106] Acceptance Judgment Unit: Preset strength thresholds (can be customized according to packaging requirements), such as edge crush strength ≥1200N / m, compressive strength ≥2500N, and burst strength ≥350kPa. The corrected mechanical property parameters are compared with the thresholds one by one. If all parameters meet the threshold requirements, a "qualified" judgment result is output; if any parameter is below the threshold, a "unqualified" result is output. The unit also analyzes the cause of the anomaly by considering the overall impact and the weight of each core influencing factor, for example, "excessive salt ion erosion depth (D=3.2mm>2.5mm), resulting in insufficient compressive strength."

[0107] Model Update Unit: The model update trigger condition is set to accumulate 100 sets of valid detection data (valid data refers to the complete dataset after removing outliers). An incremental training algorithm is used, merging the new dataset with the original training set, freezing the first 80% of the decision tree parameters, and updating only the last 20% of the decision trees to reduce training time and retain the model's original generalization ability. After training, the model performance is verified using a test set (20% of the new dataset). If the goodness-of-fit R² on the test set improves by ≥5%, the model parameters are updated; otherwise, the original model parameters are retained. Actual testing shows that after incremental training, the edge pressure intensity prediction error decreased from 8.5% to 3.2%, significantly improving long-term detection accuracy.

[0108] Furthermore, the full lifecycle data management module adopts a dual storage mode of "local storage + cloud backup" to ensure data security and traceability:

[0109] Please refer to Figure 1. The local storage uses a 2TB industrial-grade solid-state drive (SSD, read / write speed ≥500MB / s), which supports stable operation in harsh environments (operating temperature -40~85℃).

[0110] Cloud backup utilizes Alibaba Cloud OSS object storage service, and data transmission employs AES-256 encryption algorithm to ensure data security. Stored data includes environmental parameter data, multi-dimensional detection data, correction data, judgment results, and model update logs. Data is stored in a structured format of "date-environmental parameter-sample number," and supports multi-dimensional retrieval by time, environmental parameter, detection result, and other dimensions.

[0111] It has data trend analysis capabilities and can use Matlab tools to draw intensity decay trend curves, environmental parameter influence curves, etc., providing data support for packaging material selection and process optimization.

[0112] It supports abnormal operating condition playback function, which can retrieve all test data and process curves under abnormal operating conditions to assist in the analysis of abnormal causes. The data retention period is ≥10 years, which meets the requirements of product quality traceability and industry supervision.

[0113] Please refer to Figure 5 for the system workflow.

[0114] S1. Sample Pretreatment: The corrugated cardboard to be tested is cut into standard samples of 100mm×25mm (edge ​​crush test), 150mm×150mm (compression test), and 100mm×100mm (bursting test) according to GB / T2679.3-2015 (edge ​​crush strength), GB / T6543-2008 (compression strength), and GB / T454-2002 (bursting strength). Impurities and wrinkles on the sample surface are removed, and the samples are placed in a standard environment (50% humidity, 0% salt concentration, 23℃) for 24 hours to balance the moisture content and ambient humidity.

[0115] S2. Environmental Parameter Setting: Set the target environmental parameters through the touch screen of the multi-dimensional environmental simulation module, such as humidity 75%, salt concentration 3%, temperature 30℃, exposure time 24h, negative pressure -0.03MPa. After clicking start, the PLC controller automatically adjusts each unit to make the environmental parameters in the chamber reach the set value and stabilize for 60 minutes (to ensure environmental uniformity).

[0116] S3. Sample Erosion and Data Acquisition: Pre-treated samples are evenly placed on the sample rack inside the chamber (sample spacing ≥20mm to avoid mutual obstruction of salt spray). The chamber door is closed to begin salt-humidity synergistic erosion. After the exposure time, the chamber door is opened, the samples are removed, and quickly transferred to the detection platform of the composite sensing module. Simultaneously, physical parameters such as moisture content, salt adsorption, salt ion erosion depth, ultrasonic parameters, and surface roughness are collected, along with actual mechanical performance parameters such as edge crush resistance, compressive strength, and burst strength. At the same time, the environmental parameter detection unit records the environmental parameter change curves throughout the erosion process. The prediction model training unit inputs the collected key structural parameters of the corrugated cardboard into the trained random forest regression model and outputs predicted mechanical performance parameters.

[0117] S4. Abnormal Operating Condition Identification: The abnormal operating condition identification module reads the predicted and actual values, calculates the residuals of the three mechanical performance parameters, and filters the environment to be analyzed based on the 3σ criterion. If an environment is identified as an environment to be analyzed, it proceeds to the next step of synergistic influence quantification; if it is not identified as an environment to be analyzed, it directly proceeds to the qualification judgment stage.

[0118] S5. Synergistic Impact Quantification and Intelligent Correction: The synergistic impact quantification module extracts the core impact factors, calculates the correlation weights and the overall impact degree according to the above steps; the intelligent correction module calculates the correlation strength and correction amount to obtain the corrected mechanical performance parameters.

[0119] S6. Loop Detection and Model Update: The pass / fail judgment unit of the loop detection module outputs pass / fail results and anomaly cause analysis, and the full lifecycle data management module stores all detection data; after accumulating 100 sets of valid data, the model update unit automatically starts incremental training, updates the random forest regression model parameters, and completes loop optimization.

[0120] Finally, to verify the detection accuracy and innovativeness of this system, three different types of corrugated cardboard (A-flute, B-flute, and C-flute) were selected and compared in three typical humid and saline environments (this system vs. a traditional single humidity detection system). The experimental results are as follows:

[0121]

[0122]

[0123]

[0124]

[0125] Experimental results show that the detection error of this system is ≤2.2% under different humid and saline environments, which is far lower than the 6.4%~24.4% of traditional systems. The accuracy of abnormal working condition identification reaches 98.5%, and the detection accuracy is further improved by ≥10% after incremental model training. At the same time, this system adds detection dimensions such as salt ion penetration depth and internal structural damage, and innovatively adopts a quantization and correction mechanism that integrates multiple nonlinear algorithms, solving the technical pain point that traditional systems cannot adapt to salt-humidity co-erosion scenarios.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0127] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0128] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A corrugated cardboard strength testing system, characterized in that, The system includes: a multi-dimensional environment simulation module, a composite sensing and detection module, an abnormal working condition identification module, a collaborative impact quantification module, an intelligent correction module, and a closed-loop detection module; The multi-dimensional environment simulation module is used to simulate humid and saline environments with different humidity, salt concentration and salt ion penetration depth. It can reproduce the salt-humidity synergistic erosion scenario through gradient control and place the corrugated cardboard to be tested in the environment. The composite sensing and detection module is used to simultaneously collect multi-dimensional physical parameters, mechanical performance parameters, and predicted mechanical performance parameters of corrugated cardboard, while also collecting environmental parameters. The abnormal working condition identification module is used to calculate the residual distribution between the predicted mechanical performance parameters and the actual mechanical performance parameters under each humid and saline environment by combining residual analysis with the 3σ criterion, and to screen out abnormal working conditions whose residuals exceed the 3σ range as the environment to be analyzed. The collaborative impact quantification module is used to extract the core impact factors of multi-dimensional physical parameters based on principal component analysis, calculate the correlation weight between the core impact factors and the mechanical performance degradation through grey relational analysis, and combine the degree of internal structural damage reflected by ultrasonic testing to quantify the comprehensive impact of corrugated cardboard under each environment to be analyzed. The intelligent correction module is used to predict mechanical performance parameters based on the output of the random forest regression model. It takes core influencing factors, environmental parameters and comprehensive impact as inputs, calculates the correlation strength between environmental parameters and prediction residuals through partial correlation coefficients, and adaptively corrects the predicted mechanical performance parameters based on the correlation strength and correlation weights to obtain the corrected mechanical performance parameters. The closed-loop testing module is used to realize online testing of the strength of corrugated cardboard, pass / fail determination, and real-time feedback and updating of test data based on the corrected mechanical performance parameters.

2. The corrugated cardboard strength testing system according to claim 1, characterized in that: The multi-dimensional environmental simulation module includes a constant temperature and humidity salt spray infiltration chamber, a salt concentration gradient control unit, a humidity precision control unit, and an infiltration depth control unit. The salt concentration gradient control unit precisely mixes NaCl solution and deionized water using dual-channel infusion pumps, achieving a continuously adjustable salt concentration of 0.3% to 6% by mass, with a control accuracy of ±0.05%. The humidity control unit uses a combination of steam humidification and semiconductor dehumidification to achieve relative humidity regulation of 35%~98%, with a control accuracy of ±1%RH. The penetration depth control unit, by regulating the ambient temperature and exposure time, combined with negative pressure assisted penetration technology, achieves gradient control of the salt ion erosion depth; The constant temperature and humidity salt spray infiltration chamber is equipped with a salt spray uniform injection system, and the temperature, humidity and salt concentration uniformity error inside the chamber is ≤3%.

3. The corrugated cardboard strength testing system according to claim 1, characterized in that: The composite sensing and detection module includes: a multi-parameter physical detection unit, a comprehensive mechanical detection unit, an environmental parameter detection unit, and a prediction model training unit; The multi-parameter physical detection unit includes: a near-infrared moisture analyzer, an ion chromatograph, an X-ray fluorescence spectrometer, and an ultrasonic flaw detector; The comprehensive mechanical testing unit uses a multi-station mechanical testing machine to simultaneously test edge crush strength, compressive strength, and bursting strength. The environmental parameter detection unit includes a high-precision temperature and humidity sensor and an online salt concentration sensor; The prediction model training unit collects information on the corrugation type, thickness, ring crush strength, corrugation density, and sizing amount of corrugated cardboard. After standardization, the data is input into the random forest regression model for training, and the model outputs predicted edge crush strength, predicted compressive strength, and predicted bursting strength.

4. The corrugated cardboard strength testing system according to claim 1, characterized in that: The operation process of the abnormal operating condition identification module includes: The residual between the predicted mechanical performance parameters and the actual mechanical performance parameters under each environment is calculated using the following formula: ; Calculate the mean μ and standard deviation σ of the residual sequence, and construct a normal distribution model of the residuals; Environments with residuals exceeding the range of [μ−3σ,μ+3σ] are identified as environments to be analyzed, i.e., abnormal operating conditions with significant synergistic effects of salt and humidity.

5. The corrugated cardboard strength testing system according to claim 1, characterized in that: The working process of the synergistic influence quantification module includes: standardizing multi-dimensional physical parameters, and extracting the first three principal components as core influencing factors through principal component analysis. Cumulative contribution rate ≥90%; The formula for calculating the attenuation rate of actual mechanical performance parameters compared to those under standard conditions is as follows: ; The correlation weights between each core influencing factor and the decay rate η were calculated using grey relational analysis. ,in The sum is 1; Internal structure damage index constructed based on ultrasonic propagation velocity and ultrasonic attenuation coefficient: ,in Ultrasonic parameters under standard conditions. These are the weighting coefficients. ; Overall impact , where α is the structural damage correction factor, with a value of 0.3 to 0.

5.

6. The corrugated cardboard strength testing system according to claim 5, characterized in that: The extraction process of the core influencing factors includes: constructing a raw data matrix of multi-dimensional physical parameters. Where m is the number of samples and n is the number of parameters; The original data matrix is ​​standardized to obtain the standardized matrix. ; Calculate the covariance matrix C of the normalized matrix and solve for the eigenvalues ​​of the covariance matrix. and the corresponding feature vectors ; Sort the eigenvalues ​​from largest to smallest, select the eigenvectors corresponding to the top 3 eigenvalues ​​to construct the principal component loading matrix, and obtain the core influence factor through linear combination. .

7. The corrugated cardboard strength testing system according to claim 1, characterized in that: The working process of the intelligent correction module includes: Calculate the partial correlation coefficients between the residual sequence of predicted mechanical performance parameters and each core influencing factor to obtain the correlation strength R between environmental parameters and predicted residuals, where R ranges from 0 to 1; Based on the predicted value Y output by the random forest regression model, combined with the association strength R and association weights, the prediction is made. And calculate the correction amount based on the overall impact level Q. ; Corrected mechanical property parameters .

8. The corrugated cardboard strength testing system according to claim 1, characterized in that: The closed-loop detection module includes: a real-time transmission unit, a data fusion unit, a qualification determination unit, and a model update unit; The real-time transmission unit adopts industrial Ethernet + LoRa dual-mode communication to achieve stable transmission of detection data with a latency of ≤50ms; The data fusion unit performs spatiotemporal alignment and redundancy removal on the multi-source data from the composite sensing detection module to generate a unified detection dataset. The pass / fail judgment unit compares the corrected mechanical performance parameters with the preset threshold and outputs the pass / fail judgment result and an analysis of the cause of the abnormality. Every 100 sets of valid detection data accumulated, the model update unit automatically performs incremental training on the random forest regression model and updates the model parameters to improve long-term detection accuracy.

9. The corrugated cardboard strength testing system according to claim 1, characterized in that: The composite sensing and detection module also includes a surface morphology detection unit, which uses a laser confocal microscope to detect the surface roughness of corrugated cardboard, in order to help determine the degree of surface structure damage caused by salt-humidity erosion.

10. The corrugated cardboard strength testing system according to claim 1, characterized in that: It also includes a full lifecycle data management module, which stores environmental parameters, multi-dimensional detection data, correction data, judgment results and model update logs, and supports data traceability, trend analysis and abnormal working condition playback.

Images