Urine detection sensor for baby diapers and method for realizing detection

By integrating sensors and artificial intelligence technology in baby diapers, urine detection and reminders are achieved, which solves the problem of time-consuming, labor-intensive and negligible traditional detection methods, and improves the timeliness of diapers replacement and baby's comfort .

CN119023085AInactive Publication Date: 2024-11-26CHUZHOU FORTUNE COLLECTION IMPT & EXPT CO LTD
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202410896744.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-04
Publication Date
2024-11-26
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional baby diapers and urine detection methods rely on parents' vision and smell, which are time-consuming and labor-intensive and prone to inadvertent changes in time due to negligence, resulting in baby discomfort.

Method used

Design a baby diapers urine detection sensor, including data acquisition module, main control module, data processing module, urine analysis module, communication module and alarm prompt module, to detect urine conditions in real time through flexible sensor chip, using K- The means algorithm and artificial intelligence neural network model perform data analysis, and transmit detection results to parents through the Wi-Fi wireless communication module to promptly remind them to change diapers.

Benefits of technology

Automatic detection and timely reminder of baby diapers is realized, reducing the burden on parents, improving the baby's comfort and intelligent care level, ensuring timely replacement of diapers, and avoiding baby discomfort and skin problems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119023085A_ABST
    Figure CN119023085A_ABST
Patent Text Reader

Abstract

The invention discloses a diaper urine detection sensor and a method for realizing detection. The diaper urine detection sensor is characterized by comprising a data acquisition module, a main control module, a data processing module, a urine analysis module, an alarm prompt module and a communication module, the method for realizing detection through urine detection of the baby diapers comprises the following steps: step 1, carrying out data acquisition by utilizing the flexible sensor chip; step 2, carrying out data processing on the collected urine data information; step 3, analyzing the collected urine data information; step 4, transmitting a urine detection result to a mobile phone through a Wi-Fi wireless communication module; step 5, performing alarm prompt by using sound and light; the automatic detection of the urine of the baby diaper can be realized, and the intelligent degree is high.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention is based on the field of diaper urine detection, and more specifically relates to a baby diaper urine detection sensor and a method for realizing detection. Background Art

[0002] With the advancement of technology, more and more families are looking for more convenient and intelligent ways to care for their babies to reduce the burden on parents and improve the quality of life of their babies. Among them, baby diaper urine detection sensors, as a smart baby care product, have received widespread attention.

[0003] The traditional method of detecting urine in baby diapers mainly relies on the visual and olfactory judgment of parents. This traditional method is not only time-consuming and labor-intensive, but also prone to physical discomfort of babies due to parents' negligence and failure to change diapers in time. Therefore, the development of a sensor technology that can automatically and accurately detect the urine volume of baby diapers and promptly remind parents to change diapers is of great significance to improving the intelligent level of baby care. Summary of the invention

[0004] In view of the shortcomings of the above-mentioned technology, traditional diapers cannot automatically detect urine to promptly remind parents to change diapers. The present invention discloses a baby diaper urine detection sensor and a method for realizing detection, which can realize the detection of baby diaper urine, is simple and quick to operate, and meets user needs.

[0005] A baby diaper urine detection sensor and a method for realizing detection, comprising:

[0006] The data acquisition module uses the built-in flexible sensor chip of the diaper to detect the internal conditions of the diaper in real time and collect data;

[0007] The main control module controls the sensor to collect data through the microprocessor;

[0008] The data processing module receives data information from each sensor and processes the signal through the K-means algorithm so as to subsequently determine whether each urine data is within the threshold range; the data processing module includes a data receiving module, a data preprocessing module and a feature extraction module; the data receiving module is responsible for receiving the electrical signal from the sensor and converting the electrical signal into a digital format that can be processed by a computer; the data preprocessing module performs preliminary processing on the data to eliminate noise and smooth the signal; the feature extraction module extracts feature values ​​from the original data for subsequent data analysis;

[0009] Urine analysis module, which performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model;

[0010] The communication module transmits the urine test results to the mobile phone through the Wi-Fi wireless communication module;

[0011] An alarm prompt module receives the analysis result of the urine analysis module, and when the urine is detected and reaches a preset condition, it will give an audible and visual alarm prompt;

[0012] The signal output end of the data acquisition module is connected to the signal input end of the main control module, the signal output end of the main control module is connected to the signal input end of the data processing module, the signal output end of the data processing module is connected to the signal input end of the urine analysis, the signal output end of the urine analysis module is connected to the signal input end of the alarm prompt module, and the signal output end of the urine analysis module is connected to the signal input end of the communication module.

[0013] As a further embodiment of the present invention, the flexible sensor chip detects urine temperature, diaper surface humidity and diaper weight through a thermistor, a capacitance value and a pressure sensor; the flexible sensor chip converts the collected data information into electrical signals and transmits them to a data processing module.

[0014] As a further embodiment of the present invention, the feature extraction module uses a K-means algorithm to extract features, and the K-means algorithm output function is:

[0015]

[0016] In formula (1), A, G1, G2, T, and S are the output characteristic value, weight before water absorption, weight after water absorption, temperature, and humidity, respectively;

[0017] Randomly select n main points and divide them according to the distance between all points and the main points. The division calculation function is:

[0018]

[0019] In formula (2), X n is the divided cluster, N is the number of data in the nth cluster, y is the i-th element in the nth cluster, and then the mean of each cluster is obtained to update the cluster points. The process ends when the same data is found, and the original group is divided into n groups of clusters with similar data.

[0020] In formula (2), y is assumed to be a function of the cluster centroid, and the calculation function is:

[0021]

[0022] The adjustment calculation function for the cluster centroid is:

[0023] y(t)=bK(y(t-1))+(1-b)y0(t-1)+c ⑷

[0024] In formula (3), b is a random coefficient between 0 and 1, y0(t-1) is the optimal cluster centroid, K(y(t-1)) is the K-means clustering transformation feature at t-1 optimized by the centroid, K is the clustering feature, c is the feature constant term, and the best evaluation solution is obtained by combining formula (4). The probability feature output function is:

[0025] y(t)=K(y(t-1))+c⑸

[0026] The collected feature values ​​are transmitted to the data analysis module for processing and analysis.

[0027] As a further embodiment of the present invention, the method for establishing the artificial intelligence neural network model targeting urine volume is:

[0028] (1) Select the layers and neurons of the neural network,

[0029] (2) Collect a large amount of data on urine volume scenarios, including the humidity, temperature, and weight of diapers;

[0030] (3) Inputting the data into the model for forward deduction and back propagation error feedback training; the forward deduction is the process of multi-layer perceptron processing information, starting from the input layer, passing through the hidden layer, and finally reaching the output layer. In each layer, the output of the neuron is the weighted sum of the outputs of all neurons in the previous layer, plus a bias term, and converted through an activation function; the back propagation is used to calculate the gradient of the loss function for each weight, and update the weight to minimize the loss function.

[0031] (4) When the recognition accuracy reaches above 95%, the trigger weight parameters of each neuron are determined, thereby finally establishing the model.

[0032] As a further embodiment of the present invention, the establishment of the artificial intelligence neural network model targeting urine volume is based on MLP, and the MLP consists of an input layer, a hidden layer and an output layer;

[0033] First, perform forward deduction and weighted summation, and the calculation function is:

[0034]

[0035] The calculation function of the neuron output value is:

[0036]

[0037] The calculation function of the output value of the output layer is:

[0038]

[0039] In formulas (6), (7), and (8), t j represents the weighted sum of all inputs of the current node, i is the subscript of the previous layer of neurons or the input layer node, j is the subscript of the current neuron or the hidden layer neuron, k is the subscript of the next layer of neurons or the output layer neuron, L ij Indicates the weight of each neuron in the previous layer to the current neuron, L jk Represents the weight of the current neuron and the neurons in the next layer, e j represents the output value of the hidden layer neuron, s() represents the activation function, L represents the weight, X represents the input, L 0j X 0j represents the offset node, y represents the value of the output layer, t k represents the weighted sum of the inputs to neuron k in the output layer.

[0040] Back propagation uses gradient descent to find the optimal solution, and the calculation function is:

[0041]

[0042] The output value of the previous layer of neurons is:

[0043]

[0044] In formula (9), R is the loss function.

[0045] As a further embodiment of the present invention, the method for establishing the urine volume calculation model is:

[0046] (4) obtaining the weight of the diaper before absorbing water according to the pressure sensor;

[0047] (5) detecting the weight of the diaper at all times, and obtaining the weight after absorbing water according to the pressure sensor;

[0048] (6) Calculate urine volume based on the difference in weight before and after. The mathematical expression for calculating urine volume is:

[0049] W=(G1-G2) / k*1000⑾

[0050] In formula (11), W is the amount of urine, G1 is the weight of the diaper before absorbing water, G2 is the weight of the diaper after absorbing water, and k is the water absorption coefficient.

[0051] As a further embodiment of the present invention, the communication module is connected to the home Wi-Fi wireless communication network through the Wi-Fi wireless communication module, and transmits the data information to the guardian's mobile phone application. The guardian can view the diaper urine test results on the mobile phone application.

[0052] The alarm prompt module includes a sound alarm module and a light source alarm module; when the urine volume reaches a threshold value, the sound alarm module emits a specific sound to remind the guardian, and the light source alarm module uses an LED light to perform a light source alarm; the alarm prompt module has a delay function, and when the sensor detects that the urine exceeds the threshold value, the alarm will not be triggered immediately, ensuring that the baby has enough time to complete the urination action.

[0053] manual

[0054] As a further embodiment of the present invention, the baby diaper urine detection sensor comprises the following steps:

[0055] Step 1: using the flexible sensor chip to collect data;

[0056] The internal condition of the diaper is detected in real time by a flexible sensor chip built into the diaper, and data information collected by the sensor chip is received by a data acquisition module;

[0057] Step 2: Processing the collected urine data information;

[0058] The data processing module receives data information from various sensors and processes the signals through a K-means algorithm so as to subsequently determine whether various urine data are within a threshold range;

[0059] Step 3: Analyze the collected urine data information;

[0060] The urine analysis module performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model;

[0061] Step 4: Use the Wi-Fi wireless communication module to transmit the urine test results to the mobile phone;

[0062] Step 5: Alarm prompt: Use sound and light to give an alarm;

[0063] When the presence of urine is detected and reaches the preset conditions, an audible and visual alarm will be used to indicate the presence of urine.

[0064] Positive beneficial effects

[0065] A baby diaper urine detection sensor automatically detects the presence of urine, which not only greatly reduces the parents' constant attention and inspection burden on the baby's diaper status, but also improves the baby's comfort. The sensor detects urine in real time. Once the diaper is full, it reminds parents to change it, avoiding long-term wearing of diapers that may cause discomfort and skin problems for the baby. The diaper urine detection sensor is combined with smart devices to achieve remote monitoring and early warning. Even if the parents are not with the baby, they can receive urine detection information through the smart device and take corresponding measures in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] 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 or the description of the prior art are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work, among which:

[0067] Figure 1 This is a schematic diagram of the overall structure of a baby diaper urine detection sensor of the present invention;

[0068] Figure 2 A specific flow chart of a method for realizing detection by a baby diaper urine detection sensor of the present invention;

[0069] Figure 3 This is a diagram of an artificial intelligence neural model of a baby diaper urine detection sensor targeting urine volume of the present invention;

[0070] Figure 4 This is a workflow diagram of a K-means algorithm for a baby diaper urine detection sensor of the present invention;

[0071] Figure 5 This is a working principle diagram of a Wi-Fi wireless communication module of a baby diaper urine detection sensor of the present invention. DETAILED DESCRIPTION

[0072] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0073] A baby diaper urine detection sensor and a method for realizing detection, wherein the sensor comprises:

[0074] The data acquisition module uses the built-in flexible sensor chip of the diaper to detect the internal conditions of the diaper in real time and collect data;

[0075] The main control module controls the sensor to collect data through the microprocessor;

[0076] The data processing module receives data information from each sensor and processes the signal through the K-means algorithm so as to subsequently determine whether each urine data is within the threshold range; the data processing module includes a data receiving module, a data preprocessing module and a feature extraction module; the data receiving module is responsible for receiving the electrical signal from the sensor and converting the electrical signal into a digital format that can be processed by a computer; the data preprocessing module performs preliminary processing on the data to eliminate noise and smooth the signal; the feature extraction module extracts feature values ​​from the original data for subsequent data analysis;

[0077] Urine analysis module, which performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model;

[0078] The communication module transmits the urine test results to the mobile phone through the Wi-Fi wireless communication module;

[0079] An alarm prompt module receives the analysis result of the urine analysis module, and when the urine is detected and reaches a preset condition, it will give an audible and visual alarm prompt;

[0080] The signal output end of the data acquisition module is connected to the signal input end of the main control module, the signal output end of the main control module is connected to the signal input end of the data processing module, the signal output end of the data processing module is connected to the signal input end of the urine analysis, the signal output end of the urine analysis module is connected to the signal input end of the alarm prompt module, and the signal output end of the urine analysis module is connected to the signal input end of the communication module.

[0081] In a specific embodiment, the working process of a baby diaper urine detection sensor is as follows:

[0082] (1) Initialization and calibration: Before starting work, the sensor will initialize the working parameters and calibration thresholds. Calibration ensures that the sensor can accurately identify the presence of urine and avoid false positives or negative positives.

[0083] (2) Urine detection: When the urine in the baby diaper comes into contact with the sensor, the sensor performs urine detection to detect the humidity and weight of the diaper;

[0084] (3) Data conversion and processing: The sensor converts the detected urine information into electrical signals. After the electrical signals are amplified and filtered by the circuit, they are transmitted to the data processing module to extract characteristic values ​​for subsequent analysis.

[0085] (4) Data analysis and alarm: Data judgment and analysis are performed by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model. If the threshold is exceeded, an alarm mechanism is triggered;

[0086] (5) Continuous monitoring: During the entire use process, the sensor will continuously monitor the status of the diaper to ensure that any new urine is discharged can be detected immediately.

[0087] In the present invention, the flexible sensor chip detects urine temperature, diaper surface humidity and diaper weight through thermistors, capacitance values ​​and pressure sensors; the flexible sensor chip converts the collected data information into electrical signals and transmits them to the data processing module.

[0088] In the present invention, the feature extraction module uses the K-means algorithm to extract features, and the K-means algorithm output function is:

[0089]

[0090] In formula (1), A, G1, G2, T, and S are the output characteristic value, weight before water absorption, weight after water absorption, temperature, and humidity, respectively;

[0091] Randomly select n main points and divide them according to the distance between all points and the main points. The division calculation function is:

[0092]

[0093] In formula (2), X n is the divided cluster, N is the number of data in the nth cluster, y is the i-th element in the nth cluster, and then the mean of each cluster is obtained to update the cluster points. The process ends when the same data is found, and the original group is divided into n groups of clusters with similar data.

[0094] In formula (2), y is assumed to be a function of the cluster centroid, and the calculation function is:

[0095]

[0096] The adjusted calculation function for the cluster centroid is:

[0097] y(t)=bK(y(t-1))+(1-b)y0(t-1)+c ⑷

[0098] In formula (3), b is a random coefficient between 0 and 1, y0(t-1) is the optimal cluster centroid, K(y(t-1)) is the K-means clustering transformation feature at t-1 optimized by the centroid, k is the clustering feature, c is the feature constant term, and the best evaluation solution is obtained by combining formula (4). The probability feature output function is:

[0099] y(t)=K(y(t-1))+c⑸

[0100] The collected feature values ​​are transmitted to the data analysis module for processing and analysis.

[0101] In a specific embodiment, since the K-means algorithm is sensitive to the initial cluster center, a binary K-means algorithm is used. The binary K-means algorithm is an algorithm that weakens the initial centroid. The specific steps are as follows:

[0102] Step (1), put all sample data into a queue as a cluster;

[0103] Step (2), select a cluster from the queue and perform K-means algorithm partitioning into two sub-clusters, and add the sub-clusters to the queue;

[0104] Step (3), loop and iterate the second step until the termination condition is met, and the cluster in the queue is the final set of classified clusters;

[0105] In a specific embodiment, the basic working mode of the K-means algorithm includes:

[0106] (1) Data initialization module, which divides the data into four clusters and randomly selects four data points as the initial cluster centers;

[0107] (2) Data point assignment module, which calculates the distance between each data point and each cluster center, and each data point is assigned to the cluster with the closest cluster center;

[0108] (3) Cluster center update module, which calculates the average value of all data points in each cluster. The average value is used as the new cluster center, representing the new position of the cluster.

[0109] (4) Data iterative optimization module, repeating the steps of data point allocation and cluster center update until the stopping condition is met;

[0110] (5) Data output module, outputs 3 clusters and the location of the cluster center of each cluster.

[0111] In the present invention, the method for establishing the artificial intelligence neural network model targeting urine volume is:

[0112] (1) Select the layers and neurons of the neural network,

[0113] (2) Collect a large amount of data on urine volume scenarios, including the humidity, temperature, and weight of diapers;

[0114] (3) Inputting the data into the model for forward deduction and back propagation error feedback training; the forward deduction is the process of multi-layer perceptron processing information, starting from the input layer, passing through the hidden layer, and finally reaching the output layer. In each layer, the output of the neuron is the weighted sum of the outputs of all neurons in the previous layer, plus a bias term, and converted through an activation function; the back propagation is used to calculate the gradient of the loss function for each weight, and update the weight to minimize the loss function.

[0115] (4) When the recognition accuracy reaches above 95%, the trigger weight parameters of each neuron are determined, thereby finally establishing the model.

[0116] In a specific embodiment, the number of layers and neurons of the neural network are first selected according to the machine training learning time and the target recognition accuracy, a large amount of data is collected through the sensor, and then the data is input into the model for forward deduction and back propagation error feedback training, and the trigger weight parameters of each neuron are determined when the recognition accuracy reaches more than 95%, thereby establishing the final model. On the basis of the above-mentioned mechanical training, the quantization interval of the identified urine volume data is finally determined, and the quantization scale is evenly distributed to indicate the corresponding amount of urine; according to the pre-established model and algorithm, the urine volume can be quickly calculated and identified in real time; in addition to calculating the urine volume, the sensor can also use the characteristics of the diaper's water-absorbing layer to detect the process of urine from the beginning to being completely absorbed to saturated and unable to be absorbed, thereby issuing an alarm.

[0117] In a specific embodiment, the MLP includes an input layer, a hidden layer and an output layer, each layer is composed of multiple neurons; the input layer is the first layer of the MLP, and the input layer receives raw data or features as input; the hidden layer is composed of one or more layers, each hidden layer is composed of multiple neurons, the neurons of the hidden layer receive the output of the previous layer, and perform weighted summation on the data, and perform nonlinear transformation through an activation function; the output layer is the last layer of the MLP, and the output layer determines the number of neurons according to the type of problem, and uses neurons with a softmax activation function for output.

[0118] In the present invention, the establishment of the artificial intelligence neural network model targeting urine volume is based on MLP, and the MLP consists of an input layer, a hidden layer and an output layer;

[0119] First, perform forward deduction and weighted summation, and the calculation function is:

[0120]

[0121] The calculation function of the neuron output value is:

[0122]

[0123] The calculation function of the output value of the output layer is:

[0124]

[0125] In formulas (6), (7), and (8), t j represents the weighted sum of all inputs of the current node, i is the subscript of the previous layer of neurons or the input layer node, j is the subscript of the current neuron or the hidden layer neuron, k is the subscript of the next layer of neurons or the output layer neuron, L ij Indicates the weight of each neuron in the previous layer to the current neuron, L jk Represents the weight of the current neuron and the neurons in the next layer, e j represents the output value of the hidden layer neuron, s() represents the activation function, L represents the weight, X represents the input, L 0j X 0j represents the offset node, y represents the value of the output layer, t k represents the weighted sum of the inputs to neuron k in the output layer.

[0126] Back propagation uses gradient descent to find the optimal solution, and the calculation function is:

[0127]

[0128] The output value of the previous layer of neurons is:

[0129]

[0130] In formula (9), R is the loss function.

[0131] In a specific embodiment, the training process of the MLP includes two steps:

[0132] (1) Forward propagation: During the forward propagation process, the values ​​of the weights and the type of activation function in the network structure are known. By inputting their data into the input layer, the input signal is propagated layer by layer in the network structure, and then an output result y is obtained through the final output layer. This y value is the final output result of the forward propagation and is called the expected response.

[0133] (2) Back propagation: It compares the error signal between the expected response and the actual response of the network, and propagates the error signal back to each layer of the model. Considering that the error signal is only in the output layer and the hidden layer is not directly connected to the error signal, the chain derivative rule is used to correct the parameters in the model during the parameter update process. Compared with the single-layer perceptron, the multi-layer perceptron introduces more neurons, so that the network can better fit a large amount of data, and introduces nonlinear activation functions, so that the network has the ability to fit nonlinear data.

[0134] In a specific embodiment, the learning process of the back propagation algorithm is as follows:

[0135] (1) Select a set of training examples, each of which consists of two parts: input information and expected output results;

[0136] (2) Take an example from the training example set and input the input information into the network;

[0137] (3) Calculate the output of each layer of nodes after neuron processing;

[0138] (4) Calculate the error between the actual output and the expected output of the network;

[0139] (5) Calculate backward from the output layer to the first hidden layer, and adjust the connection weights of each neuron in the network according to a principle that can reduce the error;

[0140] (6) Repeat steps (3) to (5) for each example in the training sample set until the error for the entire training sample set reaches the requirement.

[0141] In the present invention, the method for establishing the urine volume calculation model is:

[0142] (1) obtaining the weight of the diaper before absorbing water according to the pressure sensor;

[0143] (2) detecting the weight of the diaper at all times, and obtaining the weight after absorbing water according to the pressure sensor;

[0144] (3) Calculate urine volume based on the difference in weight before and after. The mathematical expression for calculating urine volume is:

[0145] W=(G1-G2) / k*1000⑾

[0146] In formula (11), W is the amount of urine, G1 is the weight of the diaper before absorbing water, G2 is the weight of the diaper after absorbing water, and k is the water absorption coefficient.

[0147] In a specific embodiment, five diapers are tested continuously, and the usage of the diapers is shown in Table 1. When the total amount of the diapers exceeds 300g and the humidity exceeds 70%RH, an alarm is automatically sounded and it is recommended to change the diapers.

[0148] Table 1 Diaper usage

[0149]

[0150] In the present invention, the communication module is connected to the home Wi-Fi wireless communication network through the Wi-Fi wireless communication module, and transmits data information to the guardian's mobile phone application. The guardian can view the diaper urine test results on the mobile phone application.

[0151] In a specific embodiment, the Wi-Fi wireless communication module is based on the transmission technology of radio waves, and utilizes a wireless router and a wireless network card to transmit data between devices through wireless signals; the working principle of the Wi-Fi wireless communication module is divided into three steps, signal sending, signal receiving and data processing; first, the wireless router sends a wireless signal, the signal propagates in the air and is received by a nearby wireless network card, and the wireless network card that receives the signal converts it into a digital signal and transmits it to the processor of the device, the processor decodes and processes the received data, and then displays the data information on the device, completing the process of data transmission and sharing.

[0152] In the present invention, the alarm prompt module includes a sound alarm module and a light source alarm module; when the urine volume reaches a threshold value, the sound alarm module emits a specific sound to remind the guardian, and the light source alarm module uses an LED light to perform a light source alarm; the alarm prompt module has a delay function, and when the sensor detects that the urine exceeds the threshold value, the alarm will not be triggered immediately, ensuring that the baby has enough time to complete the urination action.

[0153] In the present invention, the baby diaper urine detection sensor comprises the following steps:

[0154] Step 1: using the flexible sensor chip to collect data;

[0155] The internal condition of the diaper is detected in real time by a flexible sensor chip built into the diaper, and data information collected by the sensor chip is received by a data acquisition module;

[0156] Step 2: Processing the collected urine data information;

[0157] The data processing module receives data information from various sensors and processes the signals through a K-means algorithm so as to subsequently determine whether various urine data are within a threshold range;

[0158] Step 3: Analyze the collected urine data information;

[0159] The urine analysis module performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model;

[0160] Step 4: Use the Wi-Fi wireless communication module to transmit the urine test results to the mobile phone;

[0161] Step 5: Alarm prompt: Use sound and light to give an alarm;

[0162] When the presence of urine is detected and reaches the preset conditions, an audible and visual alarm will be used to indicate the presence of urine.

[0163] Although the specific embodiments of the present invention are described above, it should be understood by those skilled in the art that these specific embodiments are only illustrative, and those skilled in the art may omit, replace, and change the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, merging the above method steps so as to perform substantially the same functions in substantially the same manner to achieve substantially the same results is within the scope of the present invention. Therefore, the scope of the present invention is limited only by the appended claims.

Claims

1. A baby diaper urine detection sensor, characterized by: Include: The data acquisition module uses the flexible sensor chip built into the diaper to detect the internal conditions of the diaper in real time and collect data; The main control module controls the sensor to collect data through the microprocessor; The data processing module receives data from each sensor and processes the signal using the K-means algorithm to subsequently determine whether each urine data is within a threshold range. The data processing module includes a data receiving module, a data preprocessing module, and a feature extraction module. The data receiving module is responsible for receiving electrical signals from the sensor and converting the electrical signals into a digital format that can be processed by a computer. The data preprocessing module performs preliminary processing on the data to eliminate noise and smooth the signal; the feature extraction module extracts feature values ​​from the original data for subsequent data analysis; The urine analysis module performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model; The communication module transmits the urine test results to the mobile phone via the Wi-Fi wireless communication module; An alarm prompt module receives the analysis result of the urine analysis module and generates an audible and visual alarm when the presence of urine is detected and the preset conditions are met; The signal output end of the data acquisition module is connected to the signal input end of the main control module, the signal output end of the main control module is connected to the signal input end of the data processing module, the signal output end of the data processing module is connected to the signal input end of the urine analysis module, the signal output end of the urine analysis module is connected to the signal input end of the alarm prompt module, and the signal output end of the urine analysis module is connected to the signal input end of the communication module.

2. The baby diaper urine detection sensor according to claim 1, characterized in that: The flexible sensor chip detects urine temperature, diaper surface humidity and diaper weight through a thermistor, a capacitance value and a pressure sensor; the flexible sensor chip converts the collected data information into an electrical signal and transmits it to a data processing module.

3. The baby diaper urine detection sensor according to claim 1, characterized in that: The feature extraction module uses the K-means algorithm to extract features. The output function of the K-means algorithm is: In formula (1), A, G1, G2, T, and S are the output characteristic value, weight before water absorption, weight after water absorption, temperature, and humidity, respectively; Randomly select n main points and divide all points according to their distance from the main points. The division calculation function is: In formula (2), X n is the divided cluster, N is the number of data in the nth cluster, y is the i-th element in the nth cluster, then get the mean of each cluster to update the cluster point, and end when the same data is found, dividing the original group into n groups of clusters with similar data; In formula (2), y is set as a function of the cluster centroid, and the calculation function is: The calculation function for adjusting the cluster centroid is: y(t)=bK(y(t-1))+(1-b)y0(t-1)+c ⑷ In formula (3), b is a random coefficient between 0 and 1, y0(t-1) is the optimal cluster centroid, K(y(t-1)) is the K-means clustering transformation feature at t-1 optimized by the centroid, K is the clustering feature, c is the feature constant term, and the best evaluation solution is obtained by combining formula (4). The probability feature output function is: y(t)=K(y(t-1))+c⑸ The collected feature values ​​are sent to the data analysis module for processing and analysis.

4. The baby diaper urine detection sensor according to claim 1, characterized in that: The method for establishing the artificial intelligence neural network model targeting urine volume is: (1) Select the layers and neurons of the neural network, (2) Collect a large amount of data on urine volume scenarios, including diaper humidity, temperature, and weight; (3) inputting the data into the model for forward deduction and back propagation error feedback training; the forward deduction is the process of multi-layer perceptron processing information, starting from the input layer, passing through the hidden layer, and finally reaching the output layer. In each layer, the output of the neuron is the weighted sum of the outputs of all neurons in the previous layer, plus a bias term, and converted by an activation function; The backpropagation is used to calculate the gradient of the loss function with respect to each weight and update the weights to minimize the loss function. (4) When the recognition accuracy reaches above 95%, the trigger weight parameters of each neuron are determined, thereby finally establishing the model.

5. The baby diaper urine detection sensor according to claim 4, characterized in that: The establishment of the artificial intelligence neural network model targeting urine volume is based on MLP, which consists of an input layer, a hidden layer, and an output layer. First, a forward deduction weighted summation is performed, and the calculation function is: The calculation function of the neuron output value is: The calculation function of the output value of the output layer is: In formulas (6), (7), and (8), t j Represents the weighted sum of all inputs of the current node, i is the subscript of the previous layer of neurons or the input layer node, j is the subscript of the current neuron or the hidden layer neuron, k is the subscript of the next layer of neurons or the output layer neuron, L ij Indicates the weight of each neuron in the previous layer to the current neuron, L jk Represents the weight of the current neuron and the neurons in the next layer, e j represents the output value of the hidden layer neuron, s() represents the activation function, L represents the weight, X represents the input, L 0j X 0j represents the offset node, y represents the value of the output layer, t k represents the weighted sum of the inputs to neuron k in the output layer. Back propagation uses gradient descent to find the optimal solution, and the calculation function is: The output value of the upper layer of neurons is: In formula (9), R is the loss function.

6. The baby diaper urine detection sensor according to claim 1, characterized in that: The method for establishing the urine volume calculation model is: (1) obtaining the weight of the diaper before absorbing water according to the pressure sensor; (2) Detecting the weight of the diaper every 5 minutes and obtaining the weight after absorbing water according to the pressure sensor; (3) Calculate urine volume based on the difference in weight before and after. The mathematical expression for calculating urine volume is: W=(G1-G2) / k*1000⑾ In formula (11), W is the amount of urine, G1 is the weight of the diaper before absorbing water, G2 is the weight of the diaper after absorbing water, and k is the water absorption coefficient.

7. The baby diaper urine detection sensor according to claim 1, characterized in that: The communication module is connected to the home Wi-Fi wireless communication network through the Wi-Fi wireless communication module, and transmits data information to the guardian's mobile phone application. The guardian can view the diaper urine test results on the mobile phone application.

8. The baby diaper urine detection sensor according to claim 1, characterized in that: The alarm prompt module includes a sound alarm module and a light alarm module; when the urine volume reaches a threshold, the sound alarm module emits a specific sound to remind the guardian, and the light alarm module uses an LED light to sound a light alarm; the alarm prompt module has a delay function. When the sensor detects that the urine exceeds the threshold, the alarm will not be triggered immediately, ensuring that the baby has enough time to complete the urination action.

9. A method for detecting urine in infant diapers, characterized in that: The application of the baby diaper urine detection sensor according to any one of claims 1 to 8 comprises the following steps: Step 1: using the flexible sensor chip to collect data; The internal condition of the diaper is detected in real time by a flexible sensor chip built into the diaper, and the data information collected by the sensor chip is received by a data acquisition module; Step 2: Processing the collected urine data; The data processing module receives data information from each sensor and processes the signal using the K-means algorithm to subsequently determine whether each urine data is within a threshold range; Step 3: Analyze the collected urine data information; The urine analysis module performs data judgment and analysis by establishing an artificial intelligence neural network model targeting urine volume and a urine volume calculation model; Step 4: Use the Wi-Fi wireless communication module to transmit the urine test results to the mobile phone; Step 5: Alarm prompt: Use sound and light to give an alarm; When the presence of urine is detected and reaches the preset conditions, an audible and visual alarm will be issued.

Citation Information

Patent Citations

  • Multifunctional baby diaper alarm system capable of achieving real time communications of mobile phone

    CN104504844A

  • Baby diaper monitoring system and monitoring method based on BP neural network

    CN108606872A

  • Monitoring alarm system for diapers

    CN109945927A

  • Wearable urine detection device and method

    CN110672681A

  • Intelligent recommendation system and recommendation method based on industrial park investment attraction

    CN113537728A