Portable droplet detection system

Abstract

The invention discloses a portable droplet detection system, which belongs to the technical field of respiratory tract droplet detection and comprises a droplet detection device and a data processing platform. The droplet detection device comprises a mounting shell. And a wearable part is arranged on the back surface of the mounting shell. A window and a baffle plate are arranged on the front surface of the mounting shell; a droplet sensor is arranged in the window, a protective net is arranged on the window, and the droplet sensor is arranged in the protective net. A droplet detection alarm circuit board, a charging module and a power supply module which are electrically connected with one another are arranged in the mounting shell; an FPGA logic control module, an analog-to-digital conversion module, a posture movement module, an alarm prompt module, a wireless communication module, a signal input end and a power input end are integrated on the droplet detection alarm circuit board. And a droplet identification and detection model and a visual interface are arranged in the data processing platform.

Description

Technical Field

[0001] The invention belongs to the technical field of respiratory droplet detection, and in particular relates to a portable droplet detection system. Background Art

[0002] A large number of studies have shown that droplet transmission is widely considered to be the main mode of transmission of respiratory pathogens. When pathogens are released from the respiratory tract of an infected person and spread from person to person through droplets, the respiratory tract of exposed and susceptible people within the transmission range will be infected, causing illness or death. Therefore, if droplets can be detected, it will reduce the risk of infection, disease and mortality of respiratory infectious diseases to a certain extent, and save more public health resources, time and cost for respiratory epidemiological investigation and tracing.

[0003] So far, the mainstream methods used by researchers at home and abroad to detect droplets can generally be divided into sampling and collection methods and observation methods. Simha et al. used Schlieren photography technology to visualize droplets in the document Universal trends in human cough airflows at large distance, thereby exploring the propagation trend of droplet airflows generated by coughing. However, the above method has many shortcomings, such as complex detection process, high detection cost, large and expensive detection equipment, high environmental requirements, and is not easy to carry. In addition, Chinese patent CN112683753A discloses an automatic detector and system for microbial droplet aerosols, which pumps droplet aerosols through an air pump, and the droplet aerosols are counted by a photon counter in a ventilation detection device. However, the automatic detector and system for droplet aerosols are large in size, not easy to carry, and difficult to apply to the droplet detection process for daily prevention. Therefore, there is an urgent need for a droplet detection device to detect droplets in this field. Summary of the invention

[0004] In view of the deficiencies in the above-mentioned background technology, the present invention aims to provide a portable droplet detection system, which solves the problems that the droplet detection system in the prior art cannot detect accurately in real time and is not easy to carry.

[0005] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is as follows:

[0006] A portable droplet detection system is provided, which includes a droplet detection device and a data processing platform.

[0007] The droplet detection device includes an installation shell, a wearable component is arranged on the back of the installation shell, and a window and a baffle are arranged on the front of the installation shell; a droplet sensor is arranged in the window, a protective net is arranged on the window, and the droplet sensor is arranged in the protective net; a droplet detection alarm circuit board, a charging module and a power supply module electrically connected to each other are arranged inside the installation shell; an FPGA logic control module, an analog-to-digital conversion module, a posture motion module, an alarm prompt module, a wireless communication module and a signal input terminal are integrated on the droplet detection alarm circuit board; the signal input terminal is connected to the droplet sensor via a signal line; and a starting signal extraction double threshold algorithm is integrated in the FPGA logic control module.

[0008] The data processing platform is electrically connected to the wireless communication module in the droplet detection device. The data processing platform is equipped with a droplet recognition detection model and a visualization interface for determining whether droplets are detected and displaying the detection results; the droplet recognition detection model includes a data preprocessing module, a feature extraction module, a classification module and a model optimization module.

[0009] The basic principle of the portable droplet detection system in the present invention is: the charging module supplements the power supply module with electric energy, and the power supply module is composed of a battery, a voltage-stabilizing chip and a voltage-stabilizing circuit, and is used to provide the required rated working voltage to each module of the droplet detection alarm circuit board. The FPGA logic control module is used for driving each working module and collecting, calculating, processing, integrating, transmitting and receiving data; the starting signal extraction double threshold algorithm integrated in the FPGA logic control module is used to obtain the effective starting signal perceived by the droplet sensor and the posture motion module; the analog-to-digital conversion module is used to convert and transmit the signal change of the droplet sensor; the posture motion module is used to obtain the human body movement state information to determine the current airflow environment, and then assist in the detection and judgment of droplets; the droplet recognition detection model and the visual interface arranged in the data processing platform are used to determine whether droplets are detected and to display the detection results; when in use, the user can wear the entire droplet detection device on the body through the wearable component, and after the data processing platform recognizes and detects the droplets in real time, it will send a command to the droplet detection device, so that the droplet detection device sends an alarm prompt sound and a flashing alarm indicator light, thereby reminding the wearer to maintain a safe distance for the spread of respiratory diseases and take disinfection avoidance measures.

[0010] Furthermore, the droplet detection alarm circuit board is also integrated with an image acquisition module electrically connected to the FPGA logic control module; specifically, the image acquisition method is micro camera acquisition, which is used to obtain the scene in which the wearer of the device is located.

[0011] Furthermore, the droplet detection alarm circuit board is also integrated with an infrared temperature measurement module electrically connected to the FPGA logic control module; specifically, the temperature is collected by a non-contact infrared thermometer to obtain the temperature of the droplet source.

[0012] Furthermore, two data acquisition holes are provided on the front side of the installation shell; the image acquisition module and the infrared temperature measurement module obtain image signals and temperature signals respectively through the two data acquisition holes.

[0013] Furthermore, the droplet detection alarm circuit board also integrates a data storage module that is electrically connected to the FPGA logic control module. The data storage module is stored in an SD card offline storage mode, which is used to store the sensor data of each working module for later traceability.

[0014] Furthermore, the alarm prompt module on the droplet detection alarm circuit board is electrically connected to the FPGA logic control module, including a sound alarm device and a visual light flashing alarm device.

[0015] Furthermore, a protective net is provided on the window, and the droplet sensor is arranged in the protective net, the protective net includes a plurality of criss-crossing strip structures, and the diameter of the strip structure is 0.8mm to 1.2mm; the length and width of a single grid in the protective net are not less than 5mm×5mm. The protective net is intended to protect the droplet sensor placed in the protective net from contamination, while ensuring good contact between the droplets and the droplet sensor.

[0016] Furthermore, the extension width of the baffle installed on the shell is in the range of 1 cm to 3 cm, which is used to reduce the interference of the oral and nasal breathing of the wearer of the device on the droplet detection result.

[0017] Furthermore, the droplet sensor is a capacitive sensor with a response recovery time of milliseconds; the droplet sensor has an interdigitated electrode structure with a droplet sensitive film attached to the surface for sensing droplets.

[0018] Furthermore, the posture motion module is a six-axis sensor, including a three-axis accelerometer and a three-axis angular velocity meter, which is electrically connected to the FPGA logic control module on the droplet detection alarm circuit board, and is used to sense the motion state of the wearer of the device and evaluate the current airflow environment.

[0019] Furthermore, the wearable component is a collar clip type structure, a lanyard type structure or a magnetic structure, and the wearable component can be flexibly selected according to needs.

[0020] Furthermore, the starting signal extraction double threshold algorithm includes:

[0021] Step 1: Set the sampling frequency f of the FPGA logic control module req , set the collected signal stable sensor sequence as the baseline sequence {x(m), 1≤m}, and set the baseline sequence buffer depth H bl ; Set a sliding window for real-time sampling and set the sliding window buffer depth H sw, the sliding window sequence is {y(n), 1≤n}, and the relative mean square error threshold of the sliding window sequence is set to T σ , set the comparison number factor C of the first-order change rate of the sliding window sequence;

[0022] Step 2: Calculate and store the baseline average value of the baseline sequence x(m), that is:

[0023]

[0024] Step 3: Calculate the relative mean square error of the current sliding window sequence y(n) relative to the baseline sequence mean baseline_av And deposit, that is:

[0025]

[0026] Step 4: The relative mean square error of the current sliding window sequence y(n) With threshold T σ Compare; if the relative mean square error Less than the threshold T σ , then recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; if the relative mean square error Greater than the threshold T σ , then keep the current sliding window sequence y(n) and go to step 5;

[0027] Step 5: Calculate the first-order rate of change of the signal of the current sliding window sequence y(n) Δx n ,Right now:

[0028] Δx n =f req (x n -x n-1 )

[0029] Step 6: The first-order change rate Δx of the signal of the current sliding window sequence y(n) n The positive and negative values ​​are counted, that is:

[0030] Count + =Count + +1

[0031] Count - =Count - +1

[0032] Step 7: Compare the number of positive change rates Count + and the number of negative change rates Count -, compare the size relationship of the number factor C; if the number of positive change rates is less than the sum of the number of negative change rates and the comparison number factor, recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; otherwise, determine that the current sliding window sequence y(n) is the starting signal of the valid signal, and then move the data of the starting signal and the subsequent collected signals to the valid signal sending buffer in a queue manner.

[0033] In the above algorithm, the dual features of relative mean square error and first-order rate of change are mainly integrated to realize the extraction of effective starting signals of droplet sensors and posture motion modules, aiming to improve transmission bandwidth utilization and transmission efficiency, reduce power consumption, and provide a reliable data basis for the judgment and identification of droplet signals and non-droplet signals.

[0034] Furthermore, the droplet recognition detection model includes:

[0035] Step 1: Perform mean filtering on the droplet sensor signal received by the data processing platform to obtain x(n); perform Kalman filtering on the three-axis acceleration signal and the three-axis angular velocity signal received by the data processing platform to obtain A x (n), A y (n), A z (n),ω x (n),ω y (n),ω z (n);

[0036] Step 2: Calculate the resultant acceleration and resultant angular velocity signals according to the filtered three-axis acceleration signals and the three-axis angular velocity signals, that is;

[0037]

[0038] Step 3: Combine the filtered droplet sensor signal, the combined acceleration signal, and the combined angular velocity signal to establish a three-channel data set, and perform data set standardization, labeling, and tensorization;

[0039] Step 4: Create a data loader and divide the training set, validation set, and test set in proportion;

[0040] Step 5: Create a bidirectional gated recurrent unit neural network model and a 1D convolutional neural network model for the training set to extract signal features, and set and adjust hyperparameters;

[0041] Step 6: Perform feature fusion, PCA dimension reduction and standardization on the features extracted in parallel;

[0042] Step 7: Input the processed features into the SVM classifier for classification;

[0043] Step 8: Train the droplet recognition detection model, perform forward propagation and loss calculation; determine whether the model has converged; if the model has not converged, perform back propagation and use the Adam optimizer to optimize the droplet recognition detection model; if the model has converged, complete the training and save the model and parameters;

[0044] Step 9: Input the validation set to evaluate the model, and introduce the learning rate decay strategy and early stopping mechanism to improve the training effect;

[0045] Step 10: Input the test set into the droplet recognition detection model to obtain the output droplet determination result.

[0046] The above model mainly integrates mean filtering, Kalman filtering, gated recurrent unit neural network, 1D convolutional neural network, support vector machine and other algorithms to achieve the screening of droplet signals and the elimination of interference signals. The above model has good classification effect and high recognition accuracy.

[0047] The beneficial effects of the present invention are:

[0048] 1. A portable droplet detection system in the present invention can detect droplets in real time and accurately by integrating a dual threshold algorithm for starting signal extraction and a droplet recognition detection model, and has the advantages of long battery life, stable operation and strong anti-interference ability.

[0049] 2. A portable droplet detection system in the present invention is based on the parallel processing of the FPGA logic control module and the millisecond-level response recovery time of the droplet sensor. The response time of the portable droplet detection system is within 1s. Once droplets are detected, an alarm response is rapid. The detection is sensitive and droplets can be detected at a long distance of more than 2m. It is low in cost, compact in size, and can be portable and worn in a variety of ways.

[0050] 3. A portable droplet detection system in the present invention supports visual display of droplet signal changes by setting a data processing terminal. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A three-dimensional explosion of a droplet detection device Figure 1 .

[0052] Figure 2 A three-dimensional explosion of a droplet detection device Figure 2 .

[0053] Figure 3 This is a schematic diagram of the structure of the droplet detection host computer platform.

[0054] Figure 4 The figure is a schematic diagram of a droplet detection device.

[0055] Figure 5 This is a schematic diagram of the droplet detection host computer platform interface.

[0056] Figure 6 The flowchart is for processing the start frame of the effective signal of the droplet sensor and the gesture motion module.

[0057] Figure 7 The flowchart is for identifying and processing droplet signals and non-droplet signals.

[0058] Among them, 1. Installation shell; 2. Droplet detection alarm circuit board; 3. Data processing platform; 4. Partition; 5. Protective net; 6. Droplet sensor; 8. Analog-to-digital conversion module; 9. Posture motion module; 10. Infrared temperature measurement module; 11. Image acquisition module; 12. FPGA logic control module; 13. Alarm prompt module; 14. Data storage module; 15. Wireless communication module; 16. Charging module; 17. Signal input terminal; 18. Power cord; 19. Power module; 20. Power input terminal; 21. Wearable parts; 22. Baffle. DETAILED DESCRIPTION

[0059] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0060] like Figure 1 and Figure 2 As shown, the present invention provides a droplet detection device, including a droplet detection device and a data processing platform.

[0061] The droplet detection device includes a mounting shell 1, and a window and a baffle 22 are arranged on the front of the mounting shell 1. The baffle 22 is located above the window, and a droplet sensor 6 is arranged in the window for detecting droplets generated by breathing, coughing, sneezing, and blowing the nose. Preferably, the extension width of the baffle 22 ranges from 1 cm to 3 cm, which is used to reduce the interference of the oral and nasal breathing of the wearer of the device on the droplet detection result. Preferably, a protective net 5 is arranged on the window, and the droplet sensor 6 is arranged in the protective net 5. The protective net 5 includes a plurality of criss-cross strip structures, and the diameter of the strip structure is 0.8 mm to 1.2 mm; the length and width of a single grid in the protective net 5 is greater than 5 mm×5 mm, and the protective net 5 is intended to protect the droplet sensor 6 placed in the protective net 5 from contamination, while ensuring that the droplets are in good contact with the droplet sensor 6. Further, the droplet sensor 6 is a cross-finger electrode structure, the response recovery time is in the millisecond level, and a droplet sensitive film is attached to the surface for sensing droplets.

[0062] A wearable component 21 is disposed on the back of the mounting housing 1; the wearable component 21 is a collar clip structure, a lanyard structure or a magnetic structure, and the wearable component 21 can be flexibly selected according to needs. In this embodiment, the wearable component 21 is a clip, which can be conveniently worn by the user to the collar clip or chest of the human body.

[0063] The interior of the installation shell 1 is provided with a droplet detection alarm circuit board 2, a charging module 16 and a power module 19 which are electrically connected to each other; specifically, a partition 4 is detachably provided in the installation shell 1, and the partition 4 can separate the interior of the installation shell 1 into a control compartment and a power compartment, the droplet detection alarm circuit board 2 is installed in the control compartment, and the charging module 16 and the power module 19 are installed in the power compartment.

[0064] Specifically, the droplet detection alarm circuit board 2 is integrated with an FPGA logic control module 12, an analog-to-digital conversion module 8, a posture motion module 9, an alarm prompt module 13, a wireless communication module 15, a signal input terminal 17 and a power input terminal 20; the signal input terminal 17 is connected to the droplet sensor 6 via a signal line; the power input terminal 20 is electrically connected to the power module 19 via a power line 18; and the FPGA logic control module 12 integrates a dual threshold algorithm for starting signal extraction.

[0065] The charging module 16 replenishes the power supply module 19 with electric energy. The power supply module 19 is composed of a battery, a voltage-stabilizing chip and a voltage-stabilizing circuit, and is used to provide the required rated working voltage to each module of the droplet detection alarm circuit board 2. The FPGA logic control module 12 is used for driving each working module and collecting, calculating, processing, integrating, transmitting and receiving data; the analog-to-digital conversion module 8 is used to convert and transmit the signal changes of the droplet sensor 6; the posture movement module 9 is used to obtain the human body movement state information to determine the current airflow environment, and then assist in the detection and judgment of droplets; the starting signal extraction double threshold algorithm integrated in the FPGA logic control module 12 is used to obtain the effective starting signal sensed by the droplet sensor 8 and the posture movement module 9.

[0066] like Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown, when in use, the user can wear the entire droplet detection device on the body through the wearable component 21. After the data processing platform 3 recognizes and detects the droplets in real time, it will send a command to the droplet detection device, causing the alarm prompt module 13 of the droplet detection device to emit an alarm prompt sound and a flashing alarm indicator light, thereby reminding the wearer to maintain a safe distance from the spread of respiratory diseases and take disinfection avoidance measures.

[0067] like Figure 3 and Figure 5As shown, the data processing platform 3 is electrically connected to the wireless communication module 15 on the droplet detection alarm circuit board to realize information interaction; a droplet recognition detection model and a visualization interface are arranged in the data processing platform 3 to determine whether droplets are detected and to display the detection results; the droplet recognition detection model includes a data preprocessing module, a feature extraction module, a classification module and a model optimization module.

[0068] Preferably, but not limited to, the droplet detection alarm circuit board 2 is also integrated with a data storage module 14 electrically connected to the FPGA logic control module 12. The data storage module 14 is stored in an SD card offline storage mode, which is used to store the sensor data of each working module to facilitate later traceability.

[0069] Preferably, but not limited to, the droplet detection alarm circuit board 2 is also integrated with an infrared temperature measurement module 10 electrically connected to the FPGA logic control module 12 for obtaining the temperature of the droplet source; specifically, the temperature is collected by a non-contact infrared thermometer.

[0070] Preferably, but not limited to, the droplet detection alarm circuit board 2 is also integrated with an image acquisition module 11 electrically connected to the FPGA logic control module 12, which is used to obtain the scene in which the wearer of the device is located; the image acquisition method is micro camera acquisition.

[0071] Preferably, but not limited to, two data acquisition holes are opened on the front of the mounting housing 1; the infrared temperature measurement module 10 and the image acquisition module 11 obtain the temperature signal and the image signal respectively through the two data acquisition holes.

[0072] like Figure 6 As shown, the dual threshold algorithm for starting signal extraction includes:

[0073] Step 1: Set the sampling frequency f of the FPGA logic control module req , set the collected signal stable sensor sequence as the baseline sequence {x(m), 1≤m}, and set the baseline sequence buffer depth H bl ; Set a sliding window for real-time sampling and set the sliding window buffer depth H sw , the sliding window sequence is {y(n), 1≤n}, and the relative mean square error threshold of the sliding window sequence is set to T σ , set the comparison number factor C of the first-order change rate of the sliding window sequence;

[0074] Step 2: Calculate and store the baseline average value of the baseline sequence x(m), that is:

[0075]

[0076] Step 3: Calculate the relative mean square error of the current sliding window sequence y(n) relative to the baseline sequence mean baseline_av And deposit, that is:

[0077]

[0078] Step 4: The relative mean square error of the current sliding window sequence y(n) With threshold T σ Compare; if the relative mean square error Less than the threshold T σ , then recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; if the relative mean square error Greater than the threshold T σ , then keep the current sliding window sequence y(n) and go to step 5;

[0079] Step 5: Calculate the first-order rate of change of the signal of the current sliding window sequence y(n) Δx n ,Right now:

[0080] Δx n =f req (x n -x n-1 )

[0081] Step 6: The first-order change rate Δx of the signal of the current sliding window sequence y(n) n The positive and negative values ​​are counted, that is:

[0082] Count + =Count + +1

[0083] Count - =Count - +1

[0084] Step 7: Compare the number of positive change rates Count + and the number of negative change rates Count - , compare the size relationship of the number factor C; if the number of positive change rates is less than the sum of the number of negative change rates and the comparison number factor, recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; otherwise, determine that the current sliding window sequence y(n) is the starting signal of the valid signal, and then move the data of the starting signal and the subsequent collected signals to the valid signal sending buffer in a queue manner.

[0085] In the above algorithm, the dual features of relative mean square error and first-order rate of change are mainly integrated to realize the extraction of effective starting signals of droplet sensors and posture motion modules, aiming to improve transmission bandwidth utilization and transmission efficiency, reduce power consumption, and provide a reliable data basis for the judgment and identification of droplet signals and non-droplet signals.

[0086] like Figure 7 As shown, the droplet recognition detection model includes:

[0087] Step 1: Perform mean filtering on the droplet sensor signal received by the data processing platform to obtain x(n); perform Kalman filtering on the three-axis acceleration signal and the three-axis angular velocity signal received by the data processing platform to obtain A x (n), A y (n), A z (n),ω x (n),ω y (n),ω z (n);

[0088] Step 2: Calculate the resultant acceleration and resultant angular velocity signals according to the filtered three-axis acceleration signals and the three-axis angular velocity signals, that is;

[0089]

[0090] Step 3: Combine the filtered droplet sensor signal, the combined acceleration signal, and the combined angular velocity signal to establish a three-channel data set, and perform data set standardization, labeling, and tensorization;

[0091] Step 4: Create a data loader and divide the training set, validation set, and test set in proportion;

[0092] Step 5: Create a parallel bidirectional gated recurrent unit neural network model and a 1D convolutional neural network model for the training set to extract signal features, and set and adjust hyperparameters;

[0093] Step 6: Perform feature fusion, PCA dimension reduction and standardization on the features extracted in parallel;

[0094] Step 7: Input the processed features into the SVM classifier for classification;

[0095] Step 8: Train the droplet recognition detection model, perform forward propagation and loss calculation; determine whether the model has converged; if the model has not converged, perform back propagation and use the Adam optimizer to optimize the droplet recognition detection model; if the model has converged, complete the training and save the model and parameters;

[0096] Step 9: Input the validation set to evaluate the model, and introduce the learning rate decay strategy and early stopping mechanism to improve the training effect;

[0097] Step 10: Input the test set into the droplet recognition detection model to obtain the output droplet determination result.

[0098] The above model mainly integrates mean filtering, Kalman filtering, gated recurrent unit neural network, 1D convolutional neural network, support vector machine and other algorithms to achieve the screening of droplet signals and the elimination of interference signals. The above model has good classification effect and high recognition accuracy.

Claims

1. A portable droplet detection system, characterized in that: Including droplet detection device and data processing platform; The droplet detection device comprises an installation shell; a wearable component is arranged on the back of the installation shell, and a window and a baffle are arranged on the front of the installation shell; the baffle is located above the window, a droplet sensor is arranged in the window, a protective net is arranged on the window, and the droplet sensor is arranged in the protective net; a droplet detection alarm circuit board, a charging module and a power supply module electrically connected to each other are arranged inside the installation shell; an FPGA logic control module, an analog-to-digital conversion module, a posture motion module, an alarm prompt module, a wireless communication module and a signal input terminal are integrated on the droplet detection alarm circuit board; the signal input terminal is connected to the droplet sensor via a signal line; a starting signal extraction double threshold algorithm is integrated in the FPGA logic control module; The data processing platform is electrically connected to the wireless communication module, and the data processing platform is provided with a droplet recognition detection model and a visualization interface for determining whether droplets are detected and displaying the detection results; the droplet recognition detection model includes a data preprocessing module, a feature extraction module, a classification module and a model optimization module.

2. The portable droplet detection system according to claim 1, characterized in that: The droplet detection alarm circuit board also integrates an image acquisition module and an infrared temperature measurement module that are electrically connected to the FPGA logic control module; two data acquisition holes are also opened on the front of the installation shell; the image acquisition module and the infrared temperature measurement module respectively obtain image signals and temperature signals through the two data acquisition holes.

3. The portable droplet detection system according to claim 1, characterized in that: The droplet detection alarm circuit board also integrates a data storage module electrically connected to the FPGA logic control module.

4. The portable droplet detection system according to claim 1, characterized in that: The analog-to-digital conversion module is electrically connected to the FPGA logic control module and the signal input end on the droplet detection alarm circuit board, and is used to convert and transmit the signal changes sensed by the droplet sensor; the posture motion module is electrically connected to the FPGA logic control module, and the posture motion module is a six-axis sensor, which includes a three-axis accelerometer and a three-axis angular velocity meter.

5. The portable droplet detection system according to claim 1, characterized in that: The alarm prompt module is electrically connected to the FPGA logic control module, and the alarm prompt module includes a sound alarm device and a visual light flashing alarm device.

6. The portable droplet detection system according to claim 1, characterized in that: The protective net includes a plurality of criss-cross strip structures, the diameter of the strip structures ranges from 0.8 mm to 1.2 mm; the length and width of a single grid in the protective net are not less than 5 mm×5 mm; the extended width of the baffle ranges from 1 cm to 3 cm, which is used to reduce the interference of the oral and nasal breathing of the wearer of the device on the droplet detection results.

7. The portable droplet detection system according to claim 1, characterized in that: The droplet sensor is a capacitive sensor with a cross-finger electrode structure and a droplet sensitive film attached to the surface for sensing droplets.

8. The portable droplet detection system according to claim 1, characterized in that: The wearable component is a collar clip type structure, a lanyard type structure or a magnetic structure.

9. The portable droplet detection system according to claim 2, characterized in that: The dual threshold algorithm for extracting the starting signal is used to extract the effective starting signal sensed by the droplet sensor and the gesture motion module. The dual threshold algorithm for extracting the starting signal includes: Step 1: Set the sampling frequency f of the FPGA logic control module req , set the collected signal stable sensor sequence as the baseline sequence {x(m), 1≤m}, and set the baseline sequence buffer depth H bl ; Set a sliding window for real-time sampling and set the sliding window buffer depth H sw , the sliding window sequence is {y(n), 1≤n}, and the relative mean square error threshold of the sliding window sequence is set to T σ , set the comparison number factor C of the first-order change rate of the sliding window sequence; Step 2: Calculate and store the baseline average value of the baseline sequence x(m), that is: Step 3: Calculate the relative mean square error of the current sliding window sequence y(n) relative to the baseline sequence mean baseline_av And deposit, that is: Step 4: The relative mean square error of the current sliding window sequence y(n) With threshold T σ Compare; if the relative mean square error Less than the threshold T σ , then recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; if the relative mean square error Greater than the threshold T σ , then keep the current sliding window sequence y(n) and go to step 5; Step 5: Calculate the first-order rate of change of the signal of the current sliding window sequence y(n) Δx n ,Right now: Δx n =f req (x n -x n-1 ) Step 6: The first-order change rate Δx of the signal of the current sliding window sequence y(n) n The positive and negative values ​​are counted, that is: Count + =Count + +1 Count - =Count - +1 Step 7: Compare the number of positive change rates Count + and the number of negative change rates Count - , compare the size relationship of the number factor C; if the number of positive change rates is less than the sum of the number of negative change rates and the comparison number factor, recursively push the new value to the sliding window, and update the sliding window sequence y(n) and return to step 3; otherwise, determine that the current sliding window sequence y(n) is the starting signal of the valid signal, and then move the data of the starting signal and the subsequent collected signals to the valid signal sending buffer in a queue manner.

10. The portable droplet detection system according to claim 1, characterized in that: The droplet recognition detection model includes: Step 1: Perform mean filtering on the droplet sensor signal received by the data processing platform to obtain x(n); perform Kalman filtering on the three-axis acceleration signal and the three-axis angular velocity signal received by the data processing platform to obtain A x (n), A y (n), A z (n),ω x (n),ω y (n),ω z (n); Step 2: Calculate the resultant acceleration and resultant angular velocity signals according to the filtered three-axis acceleration signals and the three-axis angular velocity signals, that is; Step 3: Combine the filtered droplet sensor signal, the combined acceleration signal, and the combined angular velocity signal to establish a three-channel data set, and perform data set standardization, labeling, and tensorization; Step 4: Create a data loader and divide the training set, validation set, and test set in proportion; Step 5: Create a parallel bidirectional gated recurrent unit neural network model and a 1D convolutional neural network model for the training set to extract signal features, and set and adjust hyperparameters; Step 6: Perform feature fusion, PCA dimension reduction and standardization on the features extracted from the parallel network; Step 7: Input the processed features into the SVM classifier for classification; Step 8: Train the droplet recognition detection model, perform forward propagation and loss calculation; determine whether the model has converged; if the model has not converged, perform back propagation and use the Adam optimizer to optimize the droplet recognition detection model; if the model has converged, complete the training and save the model and parameters; Step 9: Input the validation set to evaluate the model, and introduce the learning rate decay strategy and early stopping mechanism to improve the training effect; Step 10: Input the test set into the droplet recognition detection model to obtain the output droplet determination result.

Images