A method and system for detecting humidity of absorbent pad polymer water-absorbing material

Abstract

The invention discloses a method and system for detecting humidity of a water-absorbing pad polymer water-absorbing material, and relates to the field of water-absorbing pad polymer water-absorbing material. The method comprises the following steps: pre-processing a sample; generating a humidity detection plan, notifying a detection personnel and updating the humidity detection plan according to feedback; based on a feature fusion model, performing humidity detection on a water-absorbing pad polymer water-absorbing material sample placed in a humidity detection area through a humidity detection device, and in the humidity detection process, recording humidity data of each area of ​​the water-absorbing pad polymer water-absorbing material sample, and obtaining whether there is an abnormal phenomenon; based on the obtained humidity data, using a statistical analysis algorithm to describe the humidity characteristics and laws of the water-absorbing pad polymer water-absorbing material sample; the system comprises a pre-processing module, a detection planning module, a humidity detection module and an analysis module. The humidity state of the material detected by the invention will not be affected by the material area, thereby reducing hardware cost and maintenance cost.

Description

Technical Field

[0001] The invention relates to the field of water-absorbing pad polymer water-absorbing materials, and in particular to a method and a system for detecting humidity of water-absorbing pad polymer water-absorbing materials. Background Art

[0002] Absorbent pads usually use polymer absorbent materials, which can absorb and retain a large amount of liquid. Polymer absorbent materials are materials with high specific surface area and large porosity. They can absorb a large amount of water, thus having certain wettability and breathability.

[0003] Humidity testing of absorbent pad polymer water-absorbing materials can ensure that the produced absorbent pad polymer water-absorbing materials meet the design specifications, thereby ensuring the quality and performance of the product. Detecting humidity helps monitor the water absorption performance and stability of polymer water-absorbing materials to ensure the consistency of products under different environmental conditions. Through humidity testing, it can be ensured that the absorbent pad polymer water-absorbing materials show the expected water absorption speed and amount during use, thereby improving the user experience. Therefore, a method and system for humidity detection of absorbent pad polymer water-absorbing materials are urgently needed.

[0004] Among the commonly used humidity detection methods in the prior art, for example, Chinese Patent 202210643050.6 discloses a humidity detection method, system and device for textiles, which calculates the humidity difference between the first humidity collection point and the second humidity collection point, and compares it with the preset humidity difference to obtain the current humidity value of the textile. However, the above method still has the following shortcomings: when using a humidity detector to collect humidity from the material, it will be affected by the area of ​​the material, that is, the humidity detector is not suitable for humidity detection of a large range of absorbent pads. In specific applications, an additional number of humidity detectors may be required, which increases hardware costs and maintenance costs.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of the problems in the related art, the present invention proposes a method and system for detecting the humidity of a polymer water-absorbing material of an absorbent pad, so as to overcome the above technical problems existing in the existing related art.

[0007] To this end, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, a method for detecting the humidity of a polymer water-absorbing material of a water-absorbing pad is provided. The method for detecting the humidity of a polymer water-absorbing material of a water-absorbing pad comprises the following steps:

[0009] S1. Add additives to the absorbent pad polymer water-absorbing material sample to be tested, and pre-treat the sample to remove impurities.

[0010] S2. According to the emergency level of humidity detection, obtain the sample information of the absorbent pad polymer water-absorbing material and the information of the detection personnel, match and generate the humidity detection plan, notify the detection personnel and update the humidity detection plan according to the feedback.

[0011] S3. According to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection through a humidity detection device. During the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded to obtain whether there is any abnormality.

[0012] S4. Based on the acquired humidity data, a statistical analysis algorithm is used to describe the humidity characteristics and laws of the absorbent pad polymer water-absorbing material samples.

[0013] Furthermore, a humidity detection plan is generated to optimize resource allocation:

[0014] According to the emergency level of humidity detection, the following steps are involved: obtaining the sample information of the absorbent pad polymer absorbent material and the information of the test personnel, matching and generating the humidity detection plan, notifying the test personnel and updating the humidity detection plan based on the feedback:

[0015] S21, obtaining information about the polymer water-absorbing material of the water-absorbing pad and information about the testing personnel;

[0016] S22. Determine the number of humidity detection equipment and detection personnel required according to the emergency level of humidity detection;

[0017] S23, matching the ID of the inspector with the ID of the humidity detection device, and generating a humidity detection plan according to the matching result;

[0018] S24, notifying the test personnel of the humidity test plan and obtaining feedback from the test personnel;

[0019] S25. Update the humidity detection plan according to the feedback information from the detection personnel.

[0020] Furthermore, in order to use the feature fusion model and classifier, the humidity state of the absorbent polymer material of the absorbent pad is obtained:

[0021] According to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection by the humidity detection equipment, and during the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded, and whether there is any abnormal phenomenon is obtained, including the following steps:

[0022] S31, using a humidity detection device to obtain a number of image data of a sample of a water-absorbing pad polymer water-absorbing material;

[0023] S32, preprocessing the collected image data;

[0024] S33, extracting features of all image data, and arranging the image features of consecutive time points into a time series;

[0025] S34, constructing a feature fusion model, and outputting fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model;

[0026] S35, predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier;

[0027] S36, detecting abnormalities in humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample.

[0028] Furthermore, constructing a feature fusion model and outputting the fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model includes the following steps:

[0029] S341, capturing the historical information of the time series through forward convolution, and capturing the future information of the time series through backward convolution;

[0030] S342. Linearly transform the features obtained by forward convolution and backward convolution, and concatenate them according to the dimensions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

[0031] Furthermore, capturing the historical information of the time series through forward convolution and capturing the future information of the time series through backward convolution includes the following steps:

[0032] S3411, forward convolution: perform one-dimensional dilated convolution calculation on the input time series from left to right, and the output at each moment is based on the current input and the input before this moment. After each layer of convolution, perform layer normalization and nonlinear activation, and perform residual connection with the original input;

[0033] S3412, backward convolution: Flip the input time series, and perform convolution calculation on the flipped time series in the same way as forward convolution. After each layer of convolution, perform layer normalization, nonlinear activation and residual connection.

[0034] Furthermore, the features obtained by the forward convolution and the backward convolution are linearly transformed and spliced ​​according to the dimension to obtain the fusion features of the absorbent pad polymer water-absorbing material sample, including the following steps:

[0035] S3421, linearly transform the image feature vectors at the last moment obtained by the forward convolution and the backward convolution respectively;

[0036] S3422. Concatenate the image feature vectors after linear transformation in two directions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

[0037] Further, predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier includes the following steps:

[0038] S351, using the water-absorbing pad polymer water-absorbing material data containing the mark as a data set, and dividing the data set into a training set and a validation set, and the data set contains fusion features and corresponding humidity state labels;

[0039] S352, training a classifier using the training set, and evaluating the performance of the classifier using the validation set;

[0040] S353, deploying the trained and verified classifier, inputting the fusion features of the new sample into the trained and verified classifier, and predicting the humidity state of the absorbent pad polymer water-absorbing material sample;

[0041] Among them, the formula for predicting the humidity state of the absorbent pad polymer water-absorbing material sample is:

[0042]

[0043] In the formula, P A vector representing the probability of humidity states;

[0044] W represents the weight matrix, X The fused feature vector representing the sample of absorbent pad polymer water-absorbing material;

[0045] b Represents the bias term vector.

[0046] Further, detecting the abnormality of humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample includes the following steps:

[0047] S361, dividing the water-absorbing pad polymer water-absorbing material sample into a number of regions, and normalizing the humidity state data of each region;

[0048] S362, constructing features and labels of humidity state data based on the sliding window;

[0049] S363, selecting a support vector regression model, and presetting parameters of the support vector regression model;

[0050] S364, using historical data to train a support vector regression model to obtain an anomaly detection model;

[0051] S365. Use root mean square error and mean absolute percentage error to evaluate the prediction performance of anomaly detection models.

[0052] S366. Detect anomalies using the difference between the predicted value and the actual value of the anomaly detection model. If the difference exceeds a preset threshold, an anomaly occurs in the corresponding area of ​​the absorbent pad polymer water-absorbing material sample at that time point.

[0053] Furthermore, constructing features and labels of humidity state data based on the sliding window includes the following steps:

[0054] S3621, obtaining historical humidity values ​​of samples during humidity testing;

[0055] S3622. Select a sliding window size, use the historical humidity values ​​within the sliding window size as features of the humidity status data, and use the humidity status data at the current moment as a label.

[0056] According to another aspect of the present invention, a system for detecting the humidity of a polymer water-absorbent material of an absorbent pad is provided, the system for detecting the humidity of a polymer water-absorbent material of an absorbent pad comprising a preprocessing module, a detection planning module, a humidity detection module and an analysis module; wherein the preprocessing module is connected to the detection planning module, the detection planning module is connected to the humidity detection module, and the humidity detection module is connected to the analysis module.

[0057] The pretreatment module is used to add additives to the absorbent pad polymer water-absorbing material sample to be tested, and pre-treat the sample to remove impurities.

[0058] The detection planning module is used to obtain the sample information of the absorbent pad polymer water-absorbing material and the detection personnel information according to the emergency level of the humidity detection, match and generate the humidity detection plan, notify the detection personnel and update the humidity detection plan according to the feedback.

[0059] The humidity detection module is used to perform humidity detection on the absorbent pad polymer water-absorbent material sample placed in the humidity detection area through a humidity detection device according to the humidity detection plan and based on the feature fusion model. During the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbent material sample is recorded to obtain whether there is any abnormal phenomenon.

[0060] The analysis module is used to describe the humidity characteristics and laws of the absorbent pad polymer water-absorbing material sample by using a statistical analysis algorithm based on the acquired humidity data.

[0061] The beneficial effects of the present invention are:

[0062] (1) The present invention provides a method and system for detecting humidity of a water-absorbing pad polymer water-absorbing material, which generates a humidity detection plan, optimizes resource allocation, and improves the response speed and efficiency of the detection process. The water-absorbing pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection by a humidity detection device, and the humidity state of the water-absorbing pad polymer water-absorbing material is obtained by using a feature fusion model and a classifier. In addition, a large range of materials can be covered by one or a few shots, rather than using multiple humidity detectors to perform humidity detection on each area of ​​the material separately. This will not be affected by the area of ​​the material, thereby reducing hardware costs and maintenance costs.

[0063] (2) The present invention strengthens the correlation between image features through the feature fusion model, improves the analysis efficiency, captures the comprehensive information of the time series through forward and backward convolution, enhances the model's ability to understand the data, strengthens the feature representation ability through linear transformation and splicing, and finally accurately judges the humidity state through classifier prediction.

[0064] (3) The present invention detects the abnormality of humidity data in each area of ​​the absorbent pad polymer water-absorbing material sample, so as to timely discover the deficiency of each area of ​​the absorbent pad polymer water-absorbing material, so as to analyze and optimize the material production process, reduce the output of unqualified materials, reduce material waste, and improve production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0066] Figure 1 is a flow chart of a method for detecting humidity of a polymer water-absorbing material of a water-absorbing pad according to an embodiment of the present invention;

[0067] Figure 2 The schematic diagram is a principle block diagram of a system for detecting humidity of a polymer water-absorbing material of a water-absorbing pad according to an embodiment of the present invention.

[0068] In the figure:

[0069] 1. Preprocessing module; 2. Detection planning module; 3. Humidity detection module; 4. Analysis module. DETAILED DESCRIPTION

[0070] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0071] According to an embodiment of the present invention, a method and system for detecting humidity of a polymer water-absorbing material of a water-absorbing pad are provided.

[0072] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to one embodiment of the present invention, a method for detecting the humidity of a polymer water-absorbing material of a water-absorbing pad is provided, and the method for detecting the humidity of a polymer water-absorbing material of a water-absorbing pad comprises the following steps:

[0073] S1. Add desiccant or other additives to the sample of the absorbent pad polymer water-absorbing material to be tested to make it have a certain uniformity and stability, and pre-treat the sample, such as drying, filtering, grading, etc., to remove impurities, ensure the purity of the sample, and improve the accuracy of the test results.

[0074] Specifically, select a suitable desiccant or other additive according to the type of sample and the desired treatment effect, calculate the amount of the required additive, and add it evenly to the sample.

[0075] Make sure the additive is thoroughly mixed with the sample to achieve an even distribution.

[0076] S2. According to the emergency level of humidity detection, obtain the sample information of the absorbent pad polymer water-absorbing material and the information of the detection personnel, match and generate the humidity detection plan, notify the detection personnel and update the humidity detection plan according to the feedback.

[0077] In a further embodiment, according to the emergency level of humidity detection, obtaining the sample information of the absorbent pad polymer water-absorbing material and the detection personnel information, matching and generating the humidity detection plan, notifying the detection personnel and updating the humidity detection plan according to the feedback includes the following steps:

[0078] S21. Obtain information about the polymer absorbent material of the absorbent pad, including location, model, characteristics of the absorbent material, etc., as well as information about the inspector.

[0079] S22. Determine the number of humidity detection equipment and testers required according to the emergency level of humidity detection. It should be noted that the specific classification of emergency levels includes low, medium and high, and each level corresponds to different response times and resource allocations. For example, low emergency level: check humidity once a week; medium emergency level: check once a day; high emergency level: multiple testers are required to monitor humidity around the clock.

[0080] S23, matching the ID of the inspector with the ID of the humidity detection device, and generating a humidity detection plan according to the matching result, including task allocation, detection process, etc.

[0081] S24. Notify the inspectors of the humidity detection plan and obtain feedback from the inspectors, including equipment operation status.

[0082] S25. Update the humidity testing plan based on the feedback from the testers, including reallocating tasks, adjusting the testing plan, etc.

[0083] S3. According to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection through a humidity detection device. During the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded to obtain whether there is any abnormality.

[0084] In a further embodiment, according to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection by a humidity detection device, and during the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded, and whether there is an abnormal phenomenon is obtained, including the following steps:

[0085] S31. Use a humidity detection device (camera or other image acquisition device) to obtain a number of image data of the absorbent pad polymer water-absorbing material sample;

[0086] S32, preprocessing the collected image data;

[0087] S33, extracting features of all image data, and arranging the image features of consecutive time points into a time series, and the feature vector of each time point represents the state of the absorbent pad polymer water-absorbing material sample at that moment;

[0088] S34, constructing a feature fusion model, and outputting fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model;

[0089] S35, predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier;

[0090] S36, detecting abnormalities in humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample.

[0091] In a further embodiment, constructing a feature fusion model and outputting the fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model includes the following steps:

[0092] S341. Capture the historical information of the time series through forward convolution, and capture the future information of the time series through backward convolution (“future” is relative to a certain point in time, which is actually the reverse processing of historical data);

[0093] S342. Linearly transform the features obtained by forward convolution and backward convolution, and concatenate them according to the dimensions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

[0094] In a further embodiment, capturing historical information of the time series by forward convolution and capturing future information of the time series by backward convolution comprises the following steps:

[0095] S3411, forward convolution: perform one-dimensional dilated convolution calculation on the input time series from left to right, and the output at each moment is based on the current input and the input before this moment. After each layer of convolution, perform layer normalization and nonlinear activation, and perform residual connection with the original input to enhance the network's expressiveness and stability.

[0096] S3412, backward convolution: Flip the input time series and perform convolution calculation on the flipped time series in the same way as forward convolution to capture the impact of the latter context on the former context. After each layer of convolution, perform layer normalization, nonlinear activation and residual connection.

[0097] In a further embodiment, the features obtained by the forward convolution and the backward convolution are linearly transformed and spliced ​​according to the dimension to obtain the fusion features of the absorbent pad polymer water-absorbing material sample, including the following steps:

[0098] S3421, linearly transform the image feature vectors at the last moment obtained by the forward convolution and the backward convolution respectively;

[0099] S3422. Concatenate the image feature vectors after linear transformation in two directions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

[0100] In a further embodiment, predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier comprises the following steps:

[0101] S351. Use the water-absorbing pad polymer water-absorbing material data containing labels as a data set, and divide the data set into a training set and a validation set, and the data set contains fusion features and corresponding humidity state labels.

[0102] S352. Use the training set to train a classifier (e.g., a softmax classifier) ​​and use the validation set to evaluate the performance of the classifier. This includes forward propagation to calculate the loss, then adjusting the classifier parameters through backpropagation, and after each training cycle, calculating the loss and other performance indicators on the validation set to monitor the generalization ability of the classifier.

[0103] S353, deploying the trained and verified classifier, inputting the fusion features of the new sample into the trained and verified classifier, and predicting the humidity state of the absorbent pad polymer water-absorbing material sample.

[0104] Among them, the formula for predicting the humidity state of the absorbent pad polymer water-absorbing material sample is:

[0105]

[0106] In the formula, P A vector representing the probabilities of humidity states.

[0107] W Represents the weight matrix, each row corresponds to a weight vector of humidity state, X Represents the fused feature vector of the absorbent pad polymer water-absorbing material samples.

[0108] b Represents the bias term vector.

[0109] In a further embodiment, detecting the abnormality of humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample comprises the following steps:

[0110] S361, dividing the absorbent pad polymer water-absorbing material sample into a number of regions, and normalizing the humidity state data of each region.

[0111] S362. Construct features and labels of humidity status data based on the sliding window.

[0112] S363: Select a support vector regression model and pre-set the parameters of the support vector regression model.

[0113] The support vector regression model can fit the regression model of data points in the feature space. When predicting, new data points are mapped to the feature space, and then the trained regression function is used for prediction.

[0114] S364: Use historical data to train the support vector regression model to obtain an anomaly detection model. In specific applications, the historical data is divided into a training set. The features and labels extracted from the training set are input into the support vector regression model for training.

[0115] S365. Use root mean square error (RMSE) and mean absolute percentage error (MAPE) to evaluate the predictive performance of the anomaly detection model.

[0116] S366. Detect anomalies using the difference between the predicted value and the actual value of the anomaly detection model. If the difference exceeds a preset threshold, an anomaly occurs in the corresponding area of ​​the absorbent pad polymer water-absorbing material sample at that time point.

[0117] In a further embodiment, constructing features and labels of humidity status data based on a sliding window includes the following steps:

[0118] S3621. Obtain historical humidity values ​​of samples during humidity testing, and collect data such as timestamp, temperature, and air pressure;

[0119] S3622. Select a sliding window size, use the historical humidity values ​​within the sliding window size as features of the humidity status data, and use the humidity status data at the current moment as a label.

[0120] For example, the following historical humidity value data is collected and normalized, as shown in Table 1;

[0121] Table 1 Historical humidity values ​​and related data

[0122] If the prediction is to predict the humidity at 12:00 on 2023-12-01, use the sliding window method to construct the following features and labels:

[0123] Feature Set:

[0124] Timestamp: morning (9-11 am).

[0125] Historical humidity values: [0.1, 0.4, 0.7].

[0126] Temperature: [20°C, 21°C, 22°C].

[0127] Air pressure: [1013hPa, 1012hPa, 1011hPa].

[0128] Label:

[0129] Humidity: 0.8.

[0130] S4. Based on the acquired humidity data, a statistical analysis algorithm, such as mean value, standard deviation, correlation coefficient, etc., is used to describe the humidity characteristics and rules of the absorbent pad polymer water-absorbing material sample.

[0131] It should be noted that the basic statistics of humidity data are calculated, including mean, median, standard deviation, etc. The results of statistical analysis are interpreted to determine humidity characteristics and patterns. Reports are written or visual charts are made to present the analysis results to stakeholders.

[0132] like Figure 2 As shown, according to another embodiment of the present invention, a system for detecting the humidity of a water-absorbing pad polymer water-absorbing material is provided, and the system for detecting the humidity of a water-absorbing pad polymer water-absorbing material includes a preprocessing module 1, a detection planning module 2, a humidity detection module 3 and an analysis module 4; wherein the preprocessing module 1 is connected to the detection planning module 2, the detection planning module 2 is connected to the humidity detection module 3, and the humidity detection module 3 is connected to the analysis module 4.

[0133] The pretreatment module 1 is used to add additives to the absorbent pad polymer water-absorbing material sample to be tested, and pretreat the sample to remove impurities.

[0134] The detection planning module 2 is used to obtain the sample information of the absorbent pad polymer water-absorbing material and the detection personnel information according to the emergency level of the humidity detection, match and generate the humidity detection plan, notify the detection personnel and update the humidity detection plan according to the feedback.

[0135] The humidity detection module 3 is used to perform humidity detection on the absorbent pad polymer water-absorbing material sample placed in the humidity detection area through a humidity detection device according to the humidity detection plan and based on the feature fusion model. During the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded to obtain whether there is any abnormal phenomenon.

[0136] The analysis module 4 is used to describe the humidity characteristics and laws of the absorbent pad polymer water-absorbing material sample by using a statistical analysis algorithm according to the acquired humidity data.

[0137] In summary, the present invention provides a method and system for detecting humidity of a water-absorbing pad polymer water-absorbing material, which generates a humidity detection plan, optimizes resource allocation, and improves the response speed and efficiency of the detection process. By passing a sample of a water-absorbing pad polymer water-absorbing material placed in a humidity detection area through a humidity detection device for humidity detection, and using a feature fusion model and a classifier, the humidity state of the water-absorbing pad polymer water-absorbing material is obtained, and a large range of materials is covered by one or a few shots, rather than using multiple humidity detectors to perform humidity detection on each area of ​​the material separately, which will not be affected by the area of ​​the material, and reduces hardware costs and maintenance costs. The present invention strengthens the correlation between image features through a feature fusion model, improves analysis efficiency, and captures comprehensive information of time series through forward and backward convolution, enhances the model's ability to understand data, strengthens the representation ability of features through linear transformation and splicing, and finally accurately judges the humidity state through classifier prediction. The present invention detects the abnormality of humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample, so as to timely discover the deficiency of each area of ​​the absorbent pad polymer water-absorbing material, so as to analyze and optimize the production process of the material, reduce the output of unqualified materials, reduce material waste and improve production efficiency.

[0138] 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 principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material, characterized in that: The method for detecting humidity of a water-absorbing pad polymer water-absorbing material comprises the following steps: S1, adding additives to the absorbent pad polymer water-absorbing material sample to be tested, and pre-treating the sample to remove impurities; S2. According to the emergency level of humidity detection, obtain the sample information of the absorbent pad polymer water-absorbing material and the information of the tester, match and generate a humidity detection plan, notify the tester and update the humidity detection plan according to the feedback; S3. According to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection by a humidity detection device, and during the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded, and whether there is any abnormality is obtained; S4. Based on the acquired humidity data, a statistical analysis algorithm is used to describe the humidity characteristics and laws of the absorbent pad polymer water-absorbing material sample; According to the humidity detection plan and based on the feature fusion model, the absorbent pad polymer water-absorbing material sample placed in the humidity detection area is subjected to humidity detection by a humidity detection device, and during the humidity detection process, the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample is recorded, and whether there is an abnormal phenomenon is obtained, including the following steps: S31, using a humidity detection device to obtain a number of image data of a sample of a water-absorbing pad polymer water-absorbing material; S32, preprocessing the collected image data; S33, extracting features of all image data, and arranging the image features of consecutive time points into a time series; S34, constructing a feature fusion model, and outputting fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model; S35, predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier; S36, detecting abnormality in humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample; The method of constructing a feature fusion model and outputting the fusion features of the absorbent pad polymer water-absorbing material sample through the feature fusion model comprises the following steps: S341, capturing the historical information of the time series through forward convolution, and capturing the future information of the time series through backward convolution; S342. Linearly transform the features obtained by forward convolution and backward convolution, and concatenate them according to the dimensions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

2. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 1, characterized in that: The method of obtaining the sample information of the absorbent pad polymer absorbent material and the information of the inspector according to the emergency level of the humidity detection, matching and generating the humidity detection plan, notifying the inspector and updating the humidity detection plan according to the feedback includes the following steps: S21, obtaining information about the polymer water-absorbing material of the water-absorbing pad and information about the testing personnel; S22. Determine the number of humidity detection equipment and detection personnel required according to the emergency level of humidity detection; S23, matching the ID of the inspector with the ID of the humidity detection device, and generating a humidity detection plan according to the matching result; S24, notifying the test personnel of the humidity test plan and obtaining feedback from the test personnel; S25. Update the humidity detection plan according to the feedback information from the detection personnel.

3. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 1, characterized in that: The method of capturing the historical information of the time series through forward convolution and capturing the future information of the time series through backward convolution includes the following steps: S3411, forward convolution: perform one-dimensional dilated convolution calculation on the input time series from left to right, and the output at each moment is based on the current input and the input before this moment. After each layer of convolution, perform layer normalization and nonlinear activation, and perform residual connection with the original input; S3412, backward convolution: Flip the input time series, and perform convolution calculation on the flipped time series in the same way as forward convolution. After each layer of convolution, perform layer normalization, nonlinear activation and residual connection.

4. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 3, characterized in that: The linear transformation of the features obtained by the forward convolution and the backward convolution and the splicing according to the dimension to obtain the fusion features of the absorbent pad polymer water-absorbing material sample includes the following steps: S3421, linearly transform the image feature vectors at the last moment obtained by the forward convolution and the backward convolution respectively; S3422. Concatenate the image feature vectors after linear transformation in two directions to obtain the fusion features of the absorbent pad polymer water-absorbing material sample.

5. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 4, characterized in that: The method of predicting the humidity state of the absorbent pad polymer water-absorbing material sample by a classifier comprises the following steps: S351, using the water-absorbing pad polymer water-absorbing material data containing the mark as a data set, and dividing the data set into a training set and a validation set, and the data set contains fusion features and corresponding humidity state labels; S352, training a classifier using the training set, and evaluating the performance of the classifier using the validation set; S353, deploying the trained and verified classifier, inputting the fusion features of the new sample into the trained and verified classifier, and predicting the humidity state of the absorbent pad polymer water-absorbing material sample; Among them, the formula for predicting the humidity state of the absorbent pad polymer water-absorbing material sample is: In the formula, P A vector representing the probability of humidity states; W represents the weight matrix, X The fused feature vector representing the sample of absorbent pad polymer water-absorbing material; b Represents the bias term vector.

6. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 5, characterized in that: The detection of abnormal humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample comprises the following steps: S361, dividing the water-absorbing pad polymer water-absorbing material sample into a number of regions, and normalizing the humidity state data of each region; S362, constructing features and labels of humidity state data based on the sliding window; S363, selecting a support vector regression model, and presetting parameters of the support vector regression model; S364, using historical data to train a support vector regression model to obtain an anomaly detection model; S365. Use root mean square error and mean absolute percentage error to evaluate the prediction performance of anomaly detection models. S366. Detect anomalies using the difference between the predicted value and the actual value of the anomaly detection model. If the difference exceeds a preset threshold, an anomaly occurs in the corresponding area of ​​the absorbent pad polymer water-absorbing material sample at that time point.

7. A method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to claim 6, characterized in that: The method of constructing the features and labels of humidity state data based on the sliding window includes the following steps: S3621, obtaining historical humidity values ​​of samples during humidity testing; S3622. Select a sliding window size, use the historical humidity values ​​within the sliding window size as features of the humidity status data, and use the humidity status data at the current moment as a label.

8. A system for detecting humidity of a water-absorbing pad polymer water-absorbing material, used to implement the method for detecting humidity of a water-absorbing pad polymer water-absorbing material according to any one of claims 1 to 7, characterized in that: The system for detecting humidity of a water-absorbing pad polymer water-absorbing material comprises a pre-processing module, a detection planning module, a humidity detection module and an analysis module; Wherein, the preprocessing module is connected to the detection planning module, the detection planning module is connected to the humidity detection module, and the humidity detection module is connected to the analysis module; The pretreatment module is used to add additives to the absorbent pad polymer water-absorbing material sample to be tested, and pre-treat the sample to remove impurities; The detection planning module is used to obtain the sample information of the absorbent pad polymer water-absorbing material and the detection personnel information according to the emergency level of the humidity detection, match and generate the humidity detection plan, notify the detection personnel and update the humidity detection plan according to the feedback; The humidity detection module is used to perform humidity detection on the absorbent pad polymer water-absorbing material sample placed in the humidity detection area through the humidity detection device according to the humidity detection plan and based on the feature fusion model, and in the humidity detection process, record the humidity data of each area of ​​the absorbent pad polymer water-absorbing material sample and obtain whether there is any abnormal phenomenon; The analysis module is used to describe the humidity characteristics and laws of the absorbent pad polymer water-absorbing material sample by using a statistical analysis algorithm based on the acquired humidity data.

Images