Intelligent monitoring system and method based on paper diaper

Abstract

The invention provides an intelligent monitoring system and method based on a paper diaper, relates to the field of intelligent monitoring, and adopts a data analysis technology based on artificial intelligence to understand the state of the paper diaper by performing time sequence analysis on paper diaper humidity data measured by a humidity sensor and paper diaper smell data measured by a smell sensor. And whether to give an alarm or not is determined based on the state of the paper diaper so as to remind a guardian to replace the paper diaper. In this way, the use state of the paper diaper can be accurately judged, and therefore unnecessary waste and potential health risks are avoided.

Description

Technical Field

[0001] The present application relates to the field of intelligent monitoring, and more specifically, to an intelligent monitoring system and method based on diapers. Background Art

[0002] Baby diapers refer to a kind of daily necessities commonly used by babies, commonly known as "diapers", which are generally made of non-woven fabrics and super absorbent resin (SAP) as the main raw materials. People are easily troubled by a problem when using diapers: how often to change diapers and when to change them. The unreasonable use of diapers leads to two possible results: (1) The method of "changing diapers as soon as they are urinated" leads to a lot of waste. According to a sample survey, up to one-third of diapers are wasted because they do not achieve their proper functions. (2) Changing diapers for a long time or changing them at regular intervals will have an adverse effect on the physical health of infants and young children, especially during the long period of time at night. Parents and caregivers cannot spend the whole night to accompany their children and change diapers in time. This phenomenon will cause infants and young children to suffer from diaper rash, skin itching and other symptoms in the long run.

[0003] Therefore, a diaper-based intelligent monitoring solution is needed. Summary of the invention

[0004] In order to solve the above technical problems, the present application is proposed. The embodiments of the present application provide a diaper-based intelligent monitoring system and method.

[0005] According to one aspect of the present application, there is provided a diaper-based intelligent monitoring system, comprising:

[0006] A diaper monitoring-related data acquisition module is used to obtain the humidity value and odor data of diapers at multiple predetermined time points within a predetermined time period;

[0007] A diaper monitoring related data encoding module, used for performing association encoding on the humidity values ​​and odor data of the diapers at the plurality of predetermined time points to obtain a diaper monitoring time series feature vector;

[0008] A diaper monitoring time series feature enhancement module, used for performing feature enhancement on the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector;

[0009] The diaper changing alarm module is used to determine whether to issue an alarm to remind the guardian to change diapers based on the diaper monitoring time series feature enhancement vector.

[0010] According to another aspect of the present application, there is provided a diaper-based intelligent monitoring method, comprising:

[0011] Obtaining humidity values ​​and odor data of diapers at multiple predetermined time points within a predetermined time period;

[0012] Performing association coding on the humidity values ​​and odor data of the diapers at the plurality of predetermined time points to obtain a diaper monitoring time series feature vector;

[0013] Performing feature enhancement on the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector;

[0014] Based on the diaper monitoring time series feature enhancement vector, a result of whether to issue an alarm to remind the guardian to change the diaper is obtained.

[0015] Compared with the prior art, the diaper-based intelligent monitoring system and method provided by the present application adopts artificial intelligence-based data analysis technology to understand the state of the diaper by performing time series analysis on the diaper humidity data measured by the humidity sensor and the diaper odor data measured by the odor sensor, and decides whether to issue an alarm based on the state of the diaper to remind the guardian to change the diaper. In this way, the use status of the diaper can be accurately judged, thereby avoiding unnecessary waste and potential health risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0017] Figure 1 4 is a system block diagram of a diaper-based intelligent monitoring system according to an embodiment of the present application.

[0018] Figure 2 It is a block diagram of a diaper monitoring related data encoding module in a diaper-based intelligent monitoring system according to an embodiment of the present application.

[0019] Figure 3 4 is a block diagram of a diaper moisture monitoring time series feature extraction unit in a diaper-based intelligent monitoring system according to an embodiment of the present application.

[0020] Figure 4 4 is a block diagram of a diaper odor monitoring time series feature extraction unit in a diaper-based intelligent monitoring system according to an embodiment of the present application.

[0021] Figure 5 4 is a block diagram of a diaper changing alarm module in a diaper-based intelligent monitoring system according to an embodiment of the present application.

[0022] Figure 6 The figure is a flow chart of a diaper-based intelligent monitoring method according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0024] Baby diapers, as an indispensable daily necessity in modern parenting, are widely welcomed for their convenience and comfort. They are usually made of materials such as non-woven fabrics and super absorbent polymers (SAP), which can quickly absorb and lock urine to keep baby's skin dry. However, parents often face some confusion about the frequency of using diapers.

[0025] First of all, if diapers are used in a simple "change as soon as you urinate" manner, although this practice seems hygienic, it may actually lead to a huge waste of resources. According to relevant sampling surveys, due to frequent changes, about one-third of diapers are discarded without fully utilizing their absorption capacity. This waste not only increases the economic burden on families, but also creates unnecessary pressure on the environment. On the other hand, if diapers are not changed in time, especially for a long time at night, it may have adverse effects on the health of babies. Long-term contact with a humid environment can easily cause discomfort symptoms such as diaper rash and skin itching on the baby's sensitive skin, which may even lead to skin infections in severe cases. Therefore, a smart monitoring solution based on diapers is expected.

[0026] In recent years, deep learning and neural networks have been widely used in computer vision, natural language processing, text signal processing and other fields. In addition, deep learning and neural networks have also shown a level close to or even beyond that of humans in image classification, object detection, semantic segmentation, text translation and other fields. The development of deep learning and neural networks provides new solutions and solutions for intelligent monitoring based on diapers.

[0027] Figure 1 FIG. 1 is a system block diagram of a diaper-based intelligent monitoring system according to an embodiment of the present application. Figure 1As shown, in the diaper-based intelligent monitoring system 100, it includes: a diaper monitoring related data acquisition module 110, which is used to obtain the humidity value and odor data of the diaper at multiple predetermined time points within a predetermined time period; a diaper monitoring related data encoding module 120, which is used to associate and encode the humidity value and odor data of the diaper at the multiple predetermined time points to obtain a diaper monitoring time series feature vector; a diaper monitoring time series feature enhancement module 130, which is used to feature enhance the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector; a diaper changing alarm module 140, which is used to obtain the result of whether to issue an alarm to remind the guardian to change diapers based on the diaper monitoring time series feature enhancement vector.

[0028] Specifically, in the technical solution of the present application, it is first necessary to obtain the humidity value and odor data of the diaper at multiple predetermined time points within a predetermined time period. It should be understood that the humidity data can directly reflect the absorption and saturation of urine by the diaper. As urine accumulates, the humidity of the diaper will gradually increase. By monitoring the humidity change, it can be determined whether the diaper is close to the saturation state, thereby determining whether it needs to be replaced. When the humidity of the diaper is too high, leakage is prone to occur, which not only affects the comfort of the baby, but also may irritate the baby's skin. Monitoring humidity can remind the guardian to change it in time, thereby protecting the baby's skin health. Although the main function of diapers is to absorb urine, babies also use diapers when they defecate subconsciously. The presence of feces will produce a specific odor, and these odor data can indirectly reflect whether there is feces residue in the diaper. Bacteria and digestive enzymes in feces may irritate the baby's skin and cause diaper rash and other problems. By monitoring the odor data, the problem of feces residue can be discovered in time and replaced. In general, the combination of humidity and odor data can more comprehensively evaluate the current state of the diaper. Even if the humidity is not high, if there is a strong odor, it may mean that the diaper needs to be changed. In particular, in one embodiment of the present application, the humidity value data can be obtained by a humidity sensor, and the odor value data can be obtained by an odor sensor such as an NH3 sensor. These two sensors can be integrated into a monitor, and the monitor is mounted on the body of the diaper by an elastic clip.

[0029] Figure 2 FIG. 1 is a block diagram of a diaper monitoring-related data encoding module in a diaper-based intelligent monitoring system according to an embodiment of the present application. Figure 2As shown, the diaper monitoring related data encoding module 120 includes: a diaper humidity monitoring time series feature extraction unit 121, used for performing time series feature extraction on the humidity values ​​of the diapers at the multiple predetermined time points to obtain a diaper humidity monitoring time series feature vector; a diaper odor monitoring time series feature extraction unit 122, used for performing time series feature extraction on the odor data of the diapers at the multiple predetermined time points to obtain a diaper odor monitoring time series feature vector; a diaper monitoring feature fusion unit 123, used for fusing the diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector to obtain the diaper monitoring time series feature vector.

[0030] Figure 3 FIG. 1 is a block diagram of a diaper humidity monitoring time series feature extraction unit in a diaper-based intelligent monitoring system according to an embodiment of the present application. Figure 3 As shown, the diaper humidity monitoring time series feature extraction unit 121 includes: a diaper humidity data structuring subunit 1211, used to arrange the humidity values ​​of the diapers at the multiple predetermined time points according to the time dimension into a diaper humidity monitoring time series input vector; a diaper humidity time series feature generation subunit 1212, used to pass the diaper humidity monitoring time series input vector through a diaper humidity time series feature encoder to obtain the diaper humidity monitoring time series feature vector.

[0031] Considering that the humidity value data of diapers changes dynamically, the humidity of diapers will continue to change over time. In order to clearly capture this change, the humidity values ​​of diapers at the multiple predetermined time points need to be arranged according to the time dimension as a diaper humidity monitoring time series input vector in the technical solution of this application. The time series data arranged according to the time dimension can help identify the trend of humidity changes, such as whether the humidity is rising or remaining stable. This trend analysis is crucial for determining the usage status of diapers and deciding whether they need to be replaced.

[0032] The original diaper humidity time series input vector contains a lot of redundant information. In order to extract the most important features for judging the timing of diaper replacement from the original time series input data, such as the rate of change of humidity, peak value, etc., it is necessary to pass the diaper humidity monitoring time series input vector through the diaper humidity time series feature encoder to obtain the diaper humidity monitoring time series feature vector. In this application, the diaper humidity time series feature encoder is a convolutional neural network model including a fully connected layer and a one-dimensional convolutional layer. Those of ordinary skill in the art should understand that the one-dimensional convolutional layer in the convolutional neural network model can capture the local features in the input time series data, that is, the relationship between the humidity changes at adjacent time points, which is conducive to understanding the short-term trend and pattern of diaper humidity changes; and the fully connected layer can integrate the humidity local time series features extracted by the one-dimensional convolutional layer to form a global feature representation, thereby understanding the overall trend of diaper humidity changes. Combining the one-dimensional convolutional layer and the fully connected layer, multi-level feature extraction can be achieved, from local to global, to fully understand the changes in diaper humidity data, thereby providing a scientific basis for diaper replacement decisions.

[0033] In an embodiment of the present application, one possible implementation method of passing the diaper humidity monitoring time series input vector through a diaper humidity time series feature encoder to obtain a diaper humidity monitoring time series feature vector may be: using the fully connected layer of the diaper humidity time series feature encoder to perform full connection encoding on the diaper humidity monitoring time series input vector using the following fully connected encoding formula to extract high-dimensional implicit features of the eigenvalues ​​at each position in the input vector; wherein the fully connected encoding formula is:

[0034]

[0035] Where X is the diaper humidity monitoring time series input vector, Y is the output vector, W is the weight matrix, B is the bias vector, Represents matrix multiplication; using the one-dimensional convolution layer of the diaper humidity time series feature encoder to perform one-dimensional convolution encoding on the diaper humidity monitoring time series input vector using the following one-dimensional convolution encoding formula to extract the high-dimensional implicit correlation features between the eigenvalues ​​at each position in the input vector; wherein the one-dimensional convolution encoding formula is:

[0036]

[0037] Among them, a is the width of the convolution kernel in the x direction, F(a) is the convolution kernel parameter vector, G(xa) is the local vector matrix operated with the convolution kernel function, w is the size of the convolution kernel, X represents the diaper humidity monitoring time series input vector, and Cov(X) represents one-dimensional convolution encoding of the diaper humidity monitoring time series input vector.

[0038] Figure 4FIG. 1 is a block diagram of a diaper odor monitoring time series feature extraction unit in a diaper-based intelligent monitoring system according to an embodiment of the present application. Figure 4 As shown, the diaper odor monitoring time series feature extraction unit 122 includes: a diaper odor data structuring subunit 1221, which is used to arrange the odor data of the diapers at the multiple predetermined time points into a diaper odor monitoring time series input vector according to the time dimension; a diaper odor time series feature generation subunit 1222, which is used to pass the diaper odor monitoring time series input vector through a diaper odor time series feature encoder to obtain the diaper odor monitoring time series feature vector.

[0039] The change of odor data over time can reflect the change of the environment inside the diaper, such as fecal residue, bacterial growth, etc. In order to understand the trend and pattern of diaper odor change and capture the important odor time series characteristics of diapers, it is first necessary to arrange the odor data of the diapers at the multiple predetermined time points according to the time dimension as a diaper odor monitoring time series input vector. By arranging the odor data according to the time dimension, the time series analysis algorithm can be used to process the odor data of the diaper, so as to capture the key features in the original data, and the arrangement of the time dimension ensures the consistency of the odor data on the time axis, so as to facilitate the fusion analysis and comprehensive judgment with the humidity data of the diaper.

[0040] Next, the diaper odor monitoring time series input vector is passed through a diaper odor time series feature encoder to obtain a diaper odor monitoring time series feature vector. In the present application, the diaper odor time series feature encoder is a convolutional neural network model containing a fully connected layer and a one-dimensional convolutional layer. The one-dimensional convolutional layer can capture local features in the data and gradually extract the information of the entire time series through a sliding window. For diaper odor data, the change of odor is often continuous and correlated. The local feature extraction method of the one-dimensional convolutional layer can well capture the time series change characteristics of the odor, thereby providing an important basis for changing diapers. The fully connected layer can enhance the nonlinear modeling ability of the model by introducing nonlinear activation functions (such as ReLU), which is conducive to processing complex and changeable diaper odor data. Since the change of odor is often affected by multiple factors, it often presents a nonlinear relationship.

[0041] Humidity and odor are two important indicators that reflect the state of diapers. Humidity mainly reflects the saturation of urine absorption by diapers, while odor may reflect more complex environmental changes such as fecal residue and bacterial growth. These two indicators have different focuses when monitoring the state of diapers, but they complement each other. In order to comprehensively utilize information from two different sources and improve the comprehensiveness and accuracy of diaper monitoring, the technical solution of this application needs to fuse the diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector to obtain the diaper monitoring time series feature vector. It should be understood that the data of a single sensor will be affected by many factors, such as sensor accuracy, environmental interference, etc. Fusion of data from multiple sensors can reduce the impact of these factors and improve the robustness of diaper state monitoring. Even if the data of a certain sensor is abnormal or biased, the data of other sensors can still provide useful information, thereby avoiding misjudgment or missed judgment.

[0042] In an embodiment of the present application, a possible implementation method of fusing the diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector to obtain the diaper monitoring time series feature vector may be: fusing the diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector using the following fusion formula to obtain the diaper monitoring time series feature vector; wherein the fusion formula is:

[0043] V c =Concat[V1,V2]

[0044] Wherein, V1 represents the time series feature vector of the diaper humidity monitoring, V2 represents the time series feature vector of the diaper odor monitoring, Concat[·] represents the cascade function, V c Represents the diaper monitoring time series feature vector.

[0045] Although a comprehensive diaper monitoring time series feature vector has been obtained through feature fusion, there will still be redundant information in this vector, and some high-level features that are critical to determining the timing of replacement may be lacking. The fused vector needs to be reprocessed to further extract and enhance these key features, that is, the diaper monitoring time series feature vector is passed through a multi-scale diaper monitoring time series feature enhancer to obtain a diaper monitoring time series feature enhancement vector. The multi-scale diaper monitoring time series feature enhancer described in the present application is a convolutional neural network model comprising a first convolutional layer and a second convolutional layer, wherein the first convolutional layer uses a one-dimensional convolutional kernel of a first scale, and the second convolutional layer uses a one-dimensional convolutional kernel of a second scale, and the first scale is different from the second scale. Time series information of different scales has different values ​​for judging the state of diapers and determining whether they need to be replaced. For example, changes in humidity and odor in a short period of time may reflect the immediate state of the diaper, while the trend of changes over a long period of time may provide a more comprehensive assessment. Convolution kernels of different scales can extract feature information at different levels, which can complement and enhance each other in the subsequent processing process, so as to better capture and integrate the information and more accurately judge whether the diaper needs to be changed.

[0046] In an embodiment of the present application, one possible implementation method of passing the diaper monitoring time series feature vector through a multi-scale diaper monitoring time series feature enhancer to obtain a diaper monitoring time series feature enhancement vector may be: inputting the diaper monitoring time series feature vector into the first convolution layer of the multi-scale diaper monitoring time series feature enhancer to obtain a first-scale diaper monitoring time series feature enhancement vector, wherein the first convolution layer has a one-dimensional convolution kernel of a first scale; inputting the diaper monitoring time series feature vector into the second convolution layer of the multi-scale diaper monitoring time series feature enhancer to obtain a second-scale diaper monitoring time series feature enhancement vector, wherein the second convolution layer has a one-dimensional convolution kernel of a second scale, and the first scale is different from the second scale; cascading the first-scale diaper monitoring time series feature enhancement vector and the second-scale diaper monitoring time series feature enhancement vector to obtain the diaper monitoring time series feature enhancement vector.

[0047] Figure 5 FIG. 1 is a block diagram of a diaper replacement alarm module in a diaper-based intelligent monitoring system according to an embodiment of the present application. Figure 5As shown, the diaper changing alarm module 140 includes: a feature vector optimization unit 141, which is used to perform probabilistic identification of feature distribution deviation of the diaper monitoring time series feature enhancement vector to obtain an optimized diaper monitoring time series feature enhancement vector; a feature vector classification unit 142, which is used to pass the optimized diaper monitoring time series feature enhancement vector through an alarm classifier to obtain a classification result, and the classification result is used to indicate whether to issue an alarm to remind the guardian to change diapers.

[0048] In particular, in the technical solution of the present application, the fused feature vector is processed by a multi-scale diaper monitoring time series feature enhancer to extract more significant features and form an enhanced diaper monitoring time series feature vector. However, considering that the feature enhancer may give too high weights to certain time series features when extracting features, the distribution of these features in the enhanced vector is shifted. Multi-scale processing may overemphasize features on certain time scales and ignore other equally important time scale features, resulting in feature imbalance. The shift deviation of feature distribution may cause the classifier to rely too much on certain features during training, resulting in extreme deviations in actual classification, such as erroneously issuing a change alarm or delaying the issuance of an alarm. If the classifier learns the wrong pattern, it may not be able to accurately distinguish when the diaper needs to be changed, resulting in a decrease in the accuracy of the classification results. Biases in the feature enhancement process may cause the classifier to be unable to generalize to new or unseen data, affecting its performance in an actual monitoring environment. In order to solve this problem, in the technical solution of the present application, the feature distribution deviation of the diaper monitoring time series feature enhancement vector is probabilistically identified to obtain an optimized diaper monitoring time series feature enhancement vector.

[0049] Specifically, in an embodiment of the present application, a possible implementation method for probabilistically identifying the characteristic distribution deviation of the diaper monitoring time series feature enhancement vector to obtain an optimized diaper monitoring time series feature enhancement vector may be: calculating the shift difference representation vector between the diaper monitoring time series feature vector and the diaper monitoring time series feature enhancement vector; probabilizing each eigenvalue of the shift difference representation vector based on a probability activation function to obtain a probabilistic shift difference representation vector; inputting the diaper monitoring time series feature enhancement vector into the alarm classifier to obtain an initial classification probability value; and The backpropagation compensation representation vector is determined by comparing the eigenvalues ​​of each position of the shifted differential representation vector with the initial classification probability value and the weight matrix of the alarm classifier; the natural exponential function value of each eigenvalue of the backpropagation compensation representation vector is calculated to obtain the backpropagation compensation class representation vector; the positional point multiplication between the backpropagation compensation class representation vector and the diaper monitoring timing feature enhancement vector is calculated to obtain the point product result vector, and the weighted sum between the point product result vector and the diaper monitoring timing feature enhancement vector is calculated to obtain the optimized diaper monitoring timing feature enhancement vector.

[0050] In an embodiment of the present application, another expression of performing probabilistic identification of feature distribution deviation on the diaper monitoring time series feature enhancement vector to obtain an optimized diaper monitoring time series feature enhancement vector may be: processing the diaper monitoring time series feature enhancement vector using the following formula to obtain the optimized diaper monitoring time series feature enhancement vector; wherein the formula is:

[0051]

[0052]

[0053] Among them, Sigmoid represents the sigmoid activation function, x 1i represents the eigenvalue of the ith position in the diaper monitoring time series eigenvector, x 2i represents the eigenvalue of the ith position in the diaper monitoring time series feature enhancement vector, sgn represents the sign function, X2 represents the diaper monitoring time series feature enhancement vector, and γ represent the first weighted hyperparameter and the second weighted hyperparameter respectively, M c represents the weight matrix of the alarm classifier, p represents the initial classification probability value, ⊙ represents the point multiplication by position, represents matrix multiplication, exp represents the natural exponential function, and X2' represents the optimized diaper monitoring time series feature enhancement vector.

[0054] That is, in view of the distribution shift deviation problem between feature vectors, the diaper monitoring time series feature enhancement vector is probabilistically identified based on the shift information between the diaper monitoring time series feature vector and the diaper monitoring time series feature enhancement vector. The shift difference representation vector is obtained by calculating the difference between the diaper monitoring time series feature vector and the diaper monitoring time series feature enhancement vector, so as to quantify the deviation of feature distribution. The shift difference representation vector reveals the alignment of the two feature vectors in the feature space, which provides a basis for subsequent compensation. Then, the probability activation function is applied to each eigenvalue of the shift difference representation vector for probabilistic processing, so that the difference value is converted into a probability distribution, so that the model can understand the deviation between features in a probabilistic form. Then, the original diaper monitoring time series feature enhancement vector is input into the classifier to obtain the initial classification probability value, wherein the initial classification probability value provides the model with the basic classification performance index before feature compensation. Then, the back propagation compensation representation vector is determined by combining the probabilistic shift difference representation vector and the initial classification probability value, as well as the weight matrix of the classifier. That is, the model's own weight and bias information is used to compensate the feature vector to reduce the classification error. Then, the natural exponential function value is calculated for each eigenvalue of the back-propagation compensation representation vector, thereby enhancing the expression range of the eigenvalue and providing a mechanism for enlarging or reducing the feature compensation. Finally, the optimized diaper monitoring time series feature enhancement vector is obtained by calculating the dot product result between the back-propagation compensation class representation vector and the diaper monitoring time series feature enhancement vector, and then calculating the weighted sum with the original diaper monitoring time series feature enhancement vector. It should be understood that the dot product and weighted sum operations combine the compensation information and the original feature information to generate an adjusted feature vector, aiming to reduce the extreme deviation of the category and improve the accuracy of classification. The entire data processing process achieves the goal of optimizing the feature representation from identifying the feature distribution shift deviation to compensating for these deviations. This process not only improves the adaptability of the alarm classifier to the change of feature distribution, but also enhances the model's discrimination of different categories. Through the probabilistic identification mechanism of feature distribution deviation, the model can more accurately capture the subtle differences between feature vectors, thereby achieving higher accuracy and better generalization performance in classification tasks.

[0055] Finally, the optimized diaper monitoring time series feature enhancement vector is passed through the alarm classifier to obtain a classification result, and the classification result is used to indicate whether to issue an alarm to remind the guardian to change diapers. It should be understood that the alarm classifier is a model that maps the extracted and enhanced feature vector to a specific decision result, that is, whether an alarm needs to be issued to remind the guardian to change diapers. Through the trained alarm classifier, judgments can be made automatically and quickly based on the input feature vector, avoiding the subjectivity and uncertainty of manual judgment. When the classification result is "need to issue an alarm", the alarm mechanism will be triggered immediately to remind the guardian to pay attention through sound, light signal or other means, which helps to ensure that the baby receives timely care and attention, and avoids discomfort or health problems caused by long-term non-change of diapers. In this way, the timeliness and accuracy of care can be significantly improved through intelligent monitoring and automatic alarms, which helps to reduce the burden on guardians and improve the efficiency and quality of care.

[0056] In an embodiment of the present application, the optimized diaper monitoring timing feature enhancement vector is passed through an alarm classifier to obtain a classification result. The classification result is used to indicate whether an alarm is issued to remind the guardian to change diapers. One possible implementation method may be: using the fully connected layer of the alarm classifier to fully connect encode the optimized diaper monitoring timing feature enhancement vector to obtain an optimized diaper monitoring coding classification feature vector; inputting the optimized diaper monitoring coding classification feature vector into the Softmax classification function of the alarm classifier to obtain the classification result.

[0057] In summary, the diaper-based intelligent monitoring system 100 according to the embodiment of the present application is explained, which uses artificial intelligence-based data analysis technology to understand the state of the diaper by performing time series analysis on the diaper humidity data measured by the humidity sensor and the diaper odor data measured by the odor sensor, and decides whether to issue an alarm based on the state of the diaper to remind the guardian to change the diaper. In this way, the use status of the diaper can be accurately judged, thereby avoiding unnecessary waste and potential health risks.

[0058] As described above, the diaper-based intelligent monitoring system 100 according to the embodiment of the present application can be implemented in various terminal devices, such as a server for intelligent monitoring of diapers. In one example, the diaper-based intelligent monitoring system 100 according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the diaper-based intelligent monitoring system 100 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the diaper-based intelligent monitoring system 100 can also be one of the many hardware modules of the terminal device.

[0059] Alternatively, in another example, the diaper-based smart monitoring system 100 and the terminal device may also be separate devices, and the diaper-based smart monitoring system 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0060] Figure 6 FIG. 1 is a flow chart of an intelligent monitoring method based on diapers according to an embodiment of the present application. Figure 6 As shown, in the diaper-based intelligent monitoring method, the following includes: S110, obtaining the humidity value and odor data of the diaper at multiple predetermined time points within a predetermined time period; S120, associating and encoding the humidity value and odor data of the diaper at the multiple predetermined time points to obtain a diaper monitoring time series feature vector; S130, performing feature enhancement on the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector; S140, based on the diaper monitoring time series feature enhancement vector, obtaining a result of whether to issue an alarm to remind the guardian to change diapers.

[0061] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned intelligent monitoring method based on diapers have been referred to above. Figures 1 to 5 The invention has been described in detail in the description of the diaper-based intelligent monitoring system, and therefore, its repeated description will be omitted.

[0062] In summary, the intelligent monitoring method based on diapers according to the embodiment of the present application is explained, which uses artificial intelligence-based data analysis technology to understand the state of diapers by performing time series analysis on the diaper humidity data measured by the humidity sensor and the diaper odor data measured by the odor sensor, and decides whether to issue an alarm based on the state of the diapers to remind the guardian to change the diapers. In this way, the use status of diapers can be accurately judged, thereby avoiding unnecessary waste and potential health risks.

[0063] It should be understood that, for the sake of clarity, embodiments of the present application are described with reference to different functional units. However, it will be apparent that, without departing from the present application, the functionality of each functional unit can be implemented in a single unit, in multiple units or as a part of other functional units. For example, the functionality that is described as being performed by a single unit can be performed by multiple different units. Therefore, reference to a specific functional unit is only considered as a reference to the appropriate unit for providing the described functionality, rather than indicating a strict logical or physical structure or organization. Therefore, the present application can be implemented in a single unit, or can be physically and functionally distributed between different units and circuits.

[0064] Although the present application has been described in conjunction with some embodiments, it is not intended to be limited to the specific form set forth herein. On the contrary, the scope of the present application is limited only by the appended claims. Additionally, although individual features may be included in different claims, these may possibly be advantageously combined, and the inclusion in different claims does not imply that the combination of features is not feasible and / or advantageous. The order of features in the claims does not imply any specific order in which the features must work. In addition, in the claims, the word "comprising" does not exclude other elements, and the term "one" or "an" does not exclude a plurality.

Claims

1. An intelligent monitoring system based on diapers, It is characterized in that include: A diaper monitoring-related data acquisition module is used to obtain the humidity value and odor data of diapers at multiple predetermined time points within a predetermined time period; A diaper monitoring related data encoding module, used for performing association encoding on the humidity values ​​and odor data of the diapers at the plurality of predetermined time points to obtain a diaper monitoring time series feature vector; A diaper monitoring time series feature enhancement module, used for performing feature enhancement on the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector; The diaper changing alarm module is used to determine whether to issue an alarm to remind the guardian to change diapers based on the diaper monitoring time series feature enhancement vector.

2. The diaper-based intelligent monitoring system according to claim 1, It is characterized in that The diaper monitoring related data encoding module includes: A diaper humidity monitoring time series feature extraction unit, used for extracting time series features of the humidity values ​​of the diapers at the plurality of predetermined time points to obtain a diaper humidity monitoring time series feature vector; A diaper odor monitoring time series feature extraction unit, used for performing time series feature extraction on the odor data of the diapers at the plurality of predetermined time points to obtain a diaper odor monitoring time series feature vector; The diaper monitoring feature fusion unit is used to fuse the diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector to obtain the diaper monitoring time series feature vector.

3. The diaper-based intelligent monitoring system according to claim 2, It is characterized in that The diaper humidity monitoring time series feature extraction unit comprises: A diaper humidity data structuring subunit, configured to arrange the humidity values ​​of the diapers at the plurality of predetermined time points according to the time dimension into a diaper humidity monitoring time series input vector; The diaper humidity time series feature generating subunit is used to pass the diaper humidity monitoring time series input vector through a diaper humidity time series feature encoder to obtain the diaper humidity monitoring time series feature vector.

4. The diaper-based intelligent monitoring system according to claim 3, It is characterized in that The diaper odor monitoring time series feature extraction unit comprises: A diaper odor data structuring subunit, configured to arrange the diaper odor data at the plurality of predetermined time points into a diaper odor monitoring time series input vector according to a time dimension; The diaper odor time series feature generating subunit is used to pass the diaper odor monitoring time series input vector through a diaper odor time series feature encoder to obtain the diaper odor monitoring time series feature vector.

5. The diaper-based intelligent monitoring system according to claim 4, It is characterized in that The diaper monitoring time series feature enhancement module is used to: pass the diaper monitoring time series feature vector through a multi-scale diaper monitoring time series feature enhancer to obtain the diaper monitoring time series feature enhancement vector.

6. The diaper-based intelligent monitoring system according to claim 5, It is characterized in that The diaper humidity time series feature encoder is a first convolutional neural network model including a fully connected layer and a one-dimensional convolutional layer, the diaper odor time series feature encoder is a second convolutional neural network model including a fully connected layer and a one-dimensional convolutional layer, and the multi-scale diaper monitoring time series feature enhancer is a convolutional neural network model including a first convolutional layer and a second convolutional layer, wherein the first convolutional layer uses a one-dimensional convolutional kernel of a first scale, and the second convolutional layer uses a one-dimensional convolutional kernel of a second scale, and the first scale is different from the second scale.

7. The diaper-based intelligent monitoring system according to claim 6, It is characterized in that The diaper changing alarm module comprises: A feature vector optimization unit, used for performing a feature distribution deviation probabilistic identification on the diaper monitoring time series feature enhancement vector to obtain an optimized diaper monitoring time series feature enhancement vector; The feature vector classification unit is used to pass the optimized diaper monitoring time series feature enhancement vector through an alarm classifier to obtain a classification result, and the classification result is used to indicate whether to issue an alarm to remind the guardian to change the diaper.

8. The diaper-based intelligent monitoring system according to claim 7, It is characterized in that The feature vector optimization unit is used for: Calculating a shift difference representation vector between the diaper monitoring time series feature vector and the diaper monitoring time series feature enhancement vector; Probabilize each eigenvalue of the shifted differential representation vector based on a probability activation function to obtain a probabilized shifted differential representation vector; Inputting the diaper monitoring time series feature enhancement vector into the alarm classifier to obtain an initial classification probability value; Determining a back-propagation compensation representation vector based on a comparison between the feature values ​​of each position of the probabilistic shifted differential representation vector and the initial classification probability value and a weight matrix of the alarm classifier; Calculating the natural exponential function value of each eigenvalue of the back-propagation compensation representation vector to obtain a back-propagation compensation class representation vector; Calculate the positional point multiplication between the back-propagation compensation class representation vector and the diaper monitoring time series feature enhancement vector to obtain a point product result vector, and calculate the weighted sum between the point product result vector and the diaper monitoring time series feature enhancement vector to obtain the optimized diaper monitoring time series feature enhancement vector.

9. An intelligent monitoring method based on diapers, It is characterized in that include: Obtaining humidity values ​​and odor data of diapers at multiple predetermined time points within a predetermined time period; Performing association coding on the humidity values ​​and odor data of the diapers at the plurality of predetermined time points to obtain a diaper monitoring time series feature vector; Performing feature enhancement on the diaper monitoring time series feature vector to obtain a diaper monitoring time series feature enhancement vector; Based on the diaper monitoring time series feature enhancement vector, a result of whether to issue an alarm to remind the guardian to change the diaper is obtained.

10. The intelligent monitoring method based on diapers according to claim 9, It is characterized in that The humidity values ​​and odor data of the diapers at the plurality of predetermined time points are associated and encoded to obtain a diaper monitoring time series feature vector, including: Extracting time series features of the humidity values ​​of the diapers at the plurality of predetermined time points to obtain a time series feature vector for monitoring humidity of the diapers; Performing time series feature extraction on the odor data of the diapers at the plurality of predetermined time points to obtain a diaper odor monitoring time series feature vector; The diaper humidity monitoring time series feature vector and the diaper odor monitoring time series feature vector are fused to obtain the diaper monitoring time series feature vector.

Images