Close contact person traceability device based on droplet detection and traceability method thereof

Abstract

The invention discloses an intimate contact person traceability device based on droplet detection and a traceability method thereof, and belongs to the circuit system field, the human respiration field and the respiratory tract disease traceability technical field, the intimate contact person traceability device comprises a droplet detection module, an intimate contact person traceability module and a rear end system which are electrically connected with each other, the spittle detection module detects spittle signals to judge physical contact conditions among individuals, the close contact person traceability module records geographic positions and time when the spittle signals are obtained, spittle spreading conditions of main spreading modes of respiratory diseases can be obviously embodied, and a rear-end system has a function of filtering a large amount of redundant information. According to the method, the search range can be accurately limited at a specific time point and a space position, close contact persons and secondary close contact persons can be found out more quickly, and an epidemic prevention department can lock high-risk crowds and regions in the shortest time and effectively control the illness state within the minimum range.

Description

Technical Field

[0001] The present invention belongs to the fields of circuit systems, human respiration, and respiratory disease tracing technology, and in particular relates to a close contact tracing device based on droplet detection and a tracing method thereof. Background Art

[0002] In recent decades, respiratory diseases have posed a challenge to global public health. Therefore, accurate disease prevention and control is crucial. However, traditional close contact tracing is limited by inaccurate memory, prompting scholars and institutions to widely study digital tracing methods.

[0003] At present, common digital methods for tracing close contacts include QR codes and WIFI. Shixiong Yao et al. proposed a blockchain-based solution in the document A multi-dimension traceable privacy-preserving prevention and control scheme of the COVID-19 epidemic based on blockchain. Sorena et al. proposed a conceptual passive contact tracing system for commercial buildings based on Wi-Fi indoor positioning in the document Developing A Conceptual Passive Contact Tracing System for Commercial Buildings Using WiFi Indoor Positioning. The close contact tracing system based on QR codes focuses on regional close contact tracing rather than individual close contact tracing; the close contact tracing system based on Wi-Fi mainly compares the information in the database. Even if there is an obstacle between an individual and the infected person, they will be regarded as close contacts, which increases the scope of investigation and is not conducive to accurate tracing. It can be seen that the existing technology is not accurate enough in processing and judging physical distance, and the world is in urgent need of an efficient and accurate technology to help trace the source of respiratory diseases. Summary of the invention

[0004] In view of the deficiencies in the above-mentioned background technology, the present invention aims to provide a close contact tracing device and a tracing method based on droplet detection, which solves the defects of the current mainstream digital close contact tracing technology in that the physical distance is not judged accurately enough and erroneous data is easily generated.

[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 close contact tracing device based on droplet detection is provided, which includes a droplet detection module, a close contact tracing module and a backend system electrically connected to each other;

[0007] The droplet detection module includes a wearable carrier, a droplet sensor is arranged on the outer front of the wearable carrier, a droplet detection circuit board is arranged on the inner side of the wearable carrier, a main control module and a capacitance conversion chip are arranged on the droplet detection circuit board, and the droplet sensor is electrically connected to the main control module through the capacitance conversion chip;

[0008] The close contact tracing module includes a shell, a display screen is arranged on the outside of the shell, and a tracing circuit board is arranged inside the shell, on which a wireless communication module, an indoor positioning module, an FPGA logic control module, an RTC module, a GPS module and a 4G module are integrated that are electrically connected to each other; the wireless communication module is electrically connected to the main control module; the indoor positioning module collects RSSI data of indoor Beacon devices; the FPGA logic control module is used to identify the droplet signals collected by the droplet sensor and control the GPS module and the RTC module to obtain geographic location information data and time information data; the 4G module uploads packaged data to the back-end system;

[0009] The back-end system is used to receive, store and visualize geographic location information data, time information data and personal identification information. The back-end system integrates indoor positioning algorithms and close contact tracing algorithms and combines the condition of the patient to achieve spatiotemporal tracing of patients with respiratory diseases.

[0010] Furthermore, as a specific setting method of installing the droplet sensor on a wearable carrier, the wearable carrier is a hollow structure, and an ear hanging rope is provided on the wearable carrier; a window is provided on the outer front side of the wearable carrier, a protective net is provided outside the window, and the droplet sensor is provided in the window.

[0011] In the above technical solution, the sensitive surface of the droplet sensor faces the external environment to ensure that the droplets generated by the individual can effectively contact the droplet sensor.

[0012] Furthermore, as a specific setting method of the protective net, the protective net is composed of a criss-cross strip structure. In the above technical solution, the protective net can prevent large particles from directly contacting and damaging the droplet sensor, while allowing droplets to pass through for easy detection.

[0013] Furthermore, the droplet sensor is a capacitive sensor, and a sensitive film is attached to its outer surface.

[0014] Furthermore, the main control module is a WIFI module; the capacitance conversion chip is used to convert the droplet signal collected by the droplet sensor into a digital signal.

[0015] Furthermore, the wireless communication module is a WIFI communication module and is electrically connected to the FPGA logic control module. The FPGA logic control module calculates and determines the droplet signal transmitted by the wireless communication module; the RTC module is a real-time clock module for providing accurate time information; the GPS module is a high-precision positioning module based on the global positioning system, which uses the global positioning system satellite network to determine the geographical location information; the 4G module is a high-speed wireless communication module based on the cellular network. Once the FPGA logic control module determines the droplet signal, it controls the RTC module and the GPS module to obtain the geographical location information and time when the droplets are detected, and packages the data and uploads it to the back-end system through the 4G module.

[0016] Furthermore, the present invention also provides a tracing method of a close contact tracing device based on droplet detection, which comprises:

[0017] Step 1: After wearing the droplet detection module and the close contact tracing module, start the droplet detection module and the close contact tracing module;

[0018] Step 2: When the droplet detection module detects the droplet signal, if it is in an indoor environment, the FPGA logic control module controls the indoor positioning module to collect the RSSI data of the indoor Beacon device and package the data and upload it to the back-end system, use the indoor positioning algorithm to locate the device, and store the located location information data in the back-end system;

[0019] Step 3: When the droplet detection module detects the droplet signal, if it is in an outdoor environment, the FPGA logic control module controls the RTC module and the GPS module to obtain the geographic location information data and time data when the droplets are detected, and the droplet signal and the geographic location information data and time data when the droplets are detected are packaged and uploaded to the back-end system through the 4G module;

[0020] Step 4: The back-end system stores the droplet signal, the indoor or outdoor geographic location information data and time data when the droplets are detected, and the personal identification information of the user wearing the tracing device;

[0021] Step 5: Input the personal identification information of the confirmed respiratory disease patient into the back-end system, use the KVST-DBSCAN algorithm to screen out the close contacts of the patient, and complete the spatiotemporal tracing of the respiratory disease patient. Furthermore, in step 2, the calculation steps of the indoor positioning algorithm include:

[0022] Step 2.1: The relationship between RSSI and the distance d from the device to the Beacon can be approximately calculated using the path loss model:

[0023] RSSI = -10nlog 10 (d)+A

[0024] Where n is the environmental attenuation factor, and A is the RSSI value at the reference point (usually measured at a distance of 1 meter);

[0025] Step 2.2: Kalman filtering is used to remove dynamic noise in RSSI signals, smooth RSSI values, and improve positioning accuracy:

[0026] First, predict the RSSI value and prediction error covariance at the next moment:

[0027]

[0028] in, is the RSSI prediction value at the kth moment, P k|k-1 is the prediction error covariance, A is the state transfer matrix, B is the control matrix, and Q is the process noise covariance matrix;

[0029] Update the data again:

[0030]

[0031] P k|k =(IK k H)P k|k-1

[0032] Among them, K k is the Kalman gain, z k is the current observation value (RSSI measurement value). H is the observation matrix, R is the observation noise covariance matrix;

[0033] Step 2.3: Gaussian filtering is used to further smooth the RSSI data and reduce the impact of extreme values:

[0034]

[0035] Among them, x i is the RSSI measurement value at each moment, and N is the number of measurements;

[0036] Step 2.4, BRB uses belief rules to reason about RSSI data to determine the possible location of the target device. In BRB reasoning, the possibility P(Location=L) of location L can be expressed as the weighted sum of the confidences of different RSSI values:

[0037]

[0038] Among them, β i is the confidence of each rule, P iis the prior probability of each location, and the location probability distribution obtained by BRB is used for rough positioning;

[0039] Step 2.5: The neural network model uses a multi-layer structure to input the processed RSSI value and output the coordinates of the device. (l) :

[0040] a (l) =f(W (l) -a (l-1) +b (l) )

[0041] Among them, W (l) is the weight matrix of the lth layer, b (l) is the bias, f is the activation function, and during the training process, the mean square error MSE is used as the loss function to calculate the deviation between the predicted position and the actual position:

[0042]

[0043] in, is the predicted position, y i is the actual position, m is the number of samples;

[0044] Step 2.6: Store the final obtained indoor geographic location information data in the back-end system.

[0045] Furthermore, in step 5, the calculation steps of the KVST-DBSCAN algorithm include:

[0046] Step 5.1: Determine the spatiotemporal distance matrix D of the range dataset D of a cluster n×n :

[0047] D n×n ={dist(i,j)|1≤i,j≤n}

[0048] Step 5.2: After rearranging the space-time distance matrix in ascending order, the Kth column vector is the K-nearest neighbor distance vector D of D K , for D K The elements in the vector are averaged to get the K average nearest neighbor distance D of D. K , generate the candidate parameter list D of parameter Eps Eps :

[0049]

[0050] Step 5.3: Use the mathematical expectation method and the generated parameter Eps candidate parameters to determine the corresponding MinPts:

[0051]

[0052] MinPts=min{MinPts1,MinPts2}

[0053] Among them, N Eps1 (p i ) and N Eps2 (p i ) represent the point p i The total number of points in the Eps1 and Eps2 domains, n is the total number of points contained in the dataset D;

[0054] Step 5.4: Determine the intra-cluster density of each object in the i-th cluster data. The intra-cluster density Den(ST i ) is expressed as:

[0055]

[0056] Among them, N(ST i ) represents the number of data points in the cluster, A(ST j ) indicates N(ST i ) in the j-th object, Mean(ST i ) indicates N(ST i ) the average variance of all objects;

[0057] Step 5.5: If Den(ST i ) is greater than the density threshold θ and is marked as a noise point. If a user wearing a traceability device is diagnosed as an infected person, other users in the cluster where the user belongs are marked as close contacts, and the close contacts of confirmed respiratory disease patients are finally output.

[0058] The beneficial effects of the present invention are:

[0059] 1. A close contact tracing device based on droplet detection and a tracing method thereof in the present invention determine the physical contact between individuals by setting a droplet detection module to detect droplet signals, and record the geographical location and time when the droplet signals are obtained through the close contact tracing module, which can clearly reflect the spread of droplets, the main transmission method of respiratory diseases, and the back-end system has the function of filtering a large amount of redundant information.

[0060] 2. The close contact tracing device and tracing method based on droplet detection in the present invention process and transmit data in parallel based on an FPGA logic control module, which is more real-time and efficient.

[0061] 3. The close contact tracing device and tracing method based on droplet detection in the present invention introduces Kalman filtering and Gaussian filtering to process data, and performs indoor positioning through an artificial neural network based on BRB, thus breaking the limitation that close contact tracing cannot be performed indoors and broadening the use scenarios of the equipment.

[0062] 4. The close contact tracing device and tracing method based on droplet detection in the present invention can solve the problem of over-reliance on manually selected parameters by introducing the KVST-DBSCAN algorithm, accurately limit the search range to a specific time point and spatial location, and more quickly find close contacts and secondary close contacts, so that the epidemic prevention department can lock in high-risk groups and areas in the shortest time and effectively control the disease within the smallest range. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a schematic diagram of the structure of a droplet detection module in a close contact tracing device based on droplet detection.

[0064] Figure 2 It is a structural schematic diagram of a close contact tracing module in a close contact tracing device based on droplet detection.

[0065] Figure 3 This is a query result diagram of the back-end system of a close contact tracing device based on droplet detection in an indoor environment.

[0066] Figure 4 This is a query result diagram of the back-end system of a close contact tracing device based on droplet detection in an outdoor environment.

[0067] Among them, 1. Wearable carrier; 2. Ear strap; 3. Droplet detection circuit board; 4. Droplet sensor; 5. Main control module; 6. Capacitance conversion chip; 7. Shell; 8. Display screen; 9. Tracing circuit board; 10. Wireless communication module; 11. FPGA logic control module; 12. 4G module; 13. RTC module; 14. GPS module; 15. Indoor positioning module. DETAILED DESCRIPTION

[0068] 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.

[0069] like Figure 1 and Figure 2As shown, the present invention provides a close contact tracing device based on droplet detection, which includes a droplet detection module, a close contact tracing module and a back-end system that are electrically connected to each other.

[0070] The droplet detection module includes a wearable carrier 1, and a droplet sensor 4 is arranged on the outer front of the wearable carrier 1. The droplet sensor 4 is a capacitive sensor, and a sensitive film is attached to the outer surface. A droplet detection circuit board 3 is arranged on the inner side of the wearable carrier 1, and a main control module 5 and a capacitive conversion chip 6 are arranged on the droplet detection circuit board 3. The droplet sensor 4 is electrically connected to the main control module 5 through the capacitive conversion chip 6. The droplet sensor 4 is used to collect droplet signals in real time and transmit the droplet signals to the close contact tracing module.

[0071] The close contact tracing module includes a shell 7, a display screen 8 is provided on the outside of the shell 7, a tracing circuit board 9 is provided on the inside of the shell 7, and a wireless communication module 10, an FPGA logic control module 11, a 4G module 12, an RTC module 13, a GPS module 14 and an indoor positioning module 15 which are electrically connected to each other are integrated on the tracing circuit board 9; the wireless communication module 10 is electrically connected to the main control module 5; in the indoor environment, the FPGA logic control module 11 is used to identify the droplet signals collected by the droplet sensor 4 and control the indoor positioning module 15 to collect the RSSI data of the indoor Beacon device, and package the data and upload it to the back-end system through the 4G module 12; in the outdoor environment, the FPGA logic control module 11 is used to identify the droplet signals collected by the droplet sensor 4 and control the RTC module 13 and the GPS module 14 to obtain the geographic location information data and time information data, and package the data and upload it to the back-end system through the 4G module 12. The FPGA logic control module 11 in the close contact tracing module identifies the droplet signal; after the FPGA logic control module 11 identifies the droplet signal, if it is in an indoor environment, the indoor positioning module 15 is controlled to collect the RSSI data of the indoor Beacon device; if it is in an outdoor environment, the RTC module 13 and the GPS module 14 are controlled to record the geographic location and time when the droplet signal is obtained.

[0072] In an indoor environment, the back-end system is used to receive and process the RSSI data of indoor Beacon devices collected by the indoor positioning module 15 for indoor positioning, tracing and managing the information of close contacts; in an outdoor environment, it receives, stores and visualizes the geographic location information data, time information data and personal identification information obtained by the RTC module 13 and the GPS module 14 for tracing and managing the information of close contacts. Specifically, the back-end system stores indoor or outdoor geographic location information data, time information data and personal identification information, draws activity trajectory maps, crowd aggregation analysis, etc., and realizes accurate spatiotemporal tracing of patients with respiratory diseases in combination with the disease, which is conducive to reducing the cost of disease prevention and control.

[0073] Preferably, as a specific arrangement mode in which the droplet sensor 4 is installed on the wearable carrier 1, the wearable carrier 1 is in the form of a mask, the wearable carrier 1 is a hollow structure, and an ear-hanging rope 2 is arranged on the wearable carrier 1; a window is arranged on the outer front of the wearable carrier 1, a protective net is arranged outside the window, and a droplet sensor 4 is arranged inside the window. The sensitive surface of the droplet sensor 4 faces the external environment to ensure that the droplets generated by the individual can effectively contact the droplet sensor 4.

[0074] Specifically, as a specific setting mode of the protection net, the protection net is composed of a criss-cross strip structure. In the above technical solution, the protection net can prevent large particles from directly contacting and damaging the droplet sensor 4, while allowing the droplets to pass through for easy detection.

[0075] Specifically, the main control module 5 is a WIFI module; the capacitance conversion chip 6 is used to convert the droplet signal collected by the droplet sensor 4 into a digital signal; the wireless communication module 10 is a WIFI communication module and is electrically connected to the FPGA logic control module 11, and the FPGA logic control module 11 calculates and judges the droplet signal transmitted by the wireless communication module 10; the RTC module 13 is a real-time clock module, which is used to provide accurate time information; the GPS module 14 is a high-precision positioning module based on the global positioning system, which uses the global positioning system satellite network to determine the geographical location information; the 4G module 12 is a high-speed wireless communication module based on the cellular network; once the FPGA logic control module 11 determines the droplet signal, it controls the RTC module 13 and the GPS module 14 to obtain the geographical location information and time when the droplets are detected, and packages the data and uploads it to the back-end system through the 4G module 12.

[0076] Furthermore, the present invention also provides a tracing method of a close contact tracing device based on droplet detection, which comprises:

[0077] Step 1. After wearing the droplet detection module and the close contact tracing module, start the droplet detection module and the close contact tracing module; specifically, the droplet detection module is worn on the user's face through the ear hanging rope 2, and the close contact tracing module can be used by hand or placed in a portable container for carrying.

[0078] Step 2: When the droplet detection module detects the droplet signal, if it is in an indoor environment, the FPGA logic control module controls the indoor positioning module 15 to collect the RSSI data of the indoor Beacon device and package the data and upload it to the back-end system, use the indoor positioning algorithm to locate the device, and store the located location information data in the back-end system;

[0079] Step 3, when the droplet detection module detects the droplet signal, the FPGA logic control module 11 controls the RTC module 13 and the GPS module 14 to obtain the geographic location information data and time data when the droplets are detected, and the droplet signal and the geographic location information data and time data when the droplets are detected are packaged and uploaded to the back-end system through the 4G module 12;

[0080] Step 4: The back-end system stores the droplet signal, the indoor or outdoor geographic location information data and time data when the droplets are detected, and the personal identification information of the user wearing the tracing device;

[0081] Step 5: Input the personal identification information of the confirmed respiratory disease patient into the back-end system, use the KVST-DBSCAN algorithm to screen out the close contacts of the patient, and complete the spatiotemporal tracing of the respiratory disease patient. Furthermore, in step 2, the calculation steps of the indoor positioning algorithm include:

[0082] Step 2.1: The relationship between RSSI and the distance d from the device to the Beacon can be approximately calculated using the path loss model:

[0083] RSSI = -10nlog 10 (d)+A

[0084] Where n is the environmental attenuation factor, and A is the RSSI value at the reference point, usually measured at a distance of 1 meter;

[0085] Step 2.2: Kalman filtering is used to remove dynamic noise in RSSI signals, smooth RSSI values, and improve positioning accuracy:

[0086] First, predict the RSSI value and prediction error covariance at the next moment:

[0087]

[0088] P k|k-1 =AP k-1|k-1 A T +Q

[0089] in, is the RSSI prediction value at the kth moment, P k|k-1 is the prediction error covariance, A is the state transfer matrix, B is the control matrix, and Q is the process noise covariance matrix;

[0090] Update the data again:

[0091]

[0092] P k|k =(IK k H)Pk|k-1

[0093] Among them, K k is the Kalman gain, z k is the current observation value (RSSI measurement value). H is the observation matrix, R is the observation noise covariance matrix;

[0094] Step S2.3: Gaussian filtering is used to further smooth the RSSI data and reduce the impact of extreme values:

[0095]

[0096] Among them, x i is the RSSI measurement value at each moment, and N is the number of measurements;

[0097] Step 2.4, BRB uses belief rules to reason about RSSI data to determine the possible location of the target device. In BRB reasoning, the possibility P(Location=L) of location L can be expressed as the weighted sum of the confidences of different RSSI values:

[0098]

[0099] Among them, β i is the confidence of each rule, P i is the prior probability of each location, and the location probability distribution obtained by BRB is used for rough positioning;

[0100] Step 2.5: The neural network model uses a multi-layer structure to input the processed RSSI value and output the coordinates of the device. (l) :

[0101] a (l) =f(W (l) -a (l-1) +b (l) )

[0102] Among them, W (l) is the weight matrix of the lth layer, b (l) is the bias, f is the activation function, and during the training process, the mean square error MSE is used as the loss function to calculate the deviation between the predicted position and the actual position:

[0103]

[0104] in, is the predicted position, y i is the actual position, m is the number of samples;

[0105] Step S2.6, storing the final indoor geographic location information data in the back-end system for subsequent close contact tracing;

[0106] Furthermore, in step 5, the calculation steps of the KVST-DBSCAN algorithm include:

[0107] Step 5.1: Determine the spatiotemporal distance matrix D of the range dataset D of a cluster n×n :

[0108] D n×n ={dist(i,j)|1≤i,j≤n}

[0109] Step 5.2: After rearranging the space-time distance matrix in ascending order, the Kth column vector is the K-nearest neighbor distance vector D of D K , for D K The elements in the vector are averaged to get the K average nearest neighbor distance D of D. K , generate the candidate parameter list D of parameter Eps Eps :

[0110]

[0111] Step 5.3: Use the mathematical expectation method and the generated parameter Eps candidate parameters to determine the corresponding MinPts:

[0112]

[0113] MinPts=min{MinPts1,MinPts2}

[0114] Among them, N Eps1 (p i ) and N Eps2 (p i ) represent the point p i The total number of points in the Eps1 and Eps2 domains, n is the total number of points contained in the dataset D;

[0115] Step 5.4: Determine the intra-cluster density of each object in the i-th cluster data. The intra-cluster density Den(ST i ) is expressed as:

[0116]

[0117] Among them, N(ST i ) represents the number of data points in the cluster, A(ST j ) indicates N(ST i ) in the j-th object, Mean(ST i ) indicates N(ST i ) the average variance of all objects;

[0118] Step 5.5: If Den(ST i ) is greater than the density threshold θ and is marked as a noise point. If a user wearing a traceability device is diagnosed as an infected person, other users in the cluster where the user belongs are marked as close contacts, and the close contacts of confirmed respiratory disease patients are finally output.

[0119] Specifically, if Figure 3 As shown in the figure, in an indoor environment, taking the respiratory disease patient "bc20_2" as an example, its indoor geographic location information is visualized, and users who were in contact with individual droplets in the same time and space as "bc20_2" when he was exposed to droplets are screened. Combining the stored information, the close contacts of the three respiratory disease patients can be quickly and accurately screened. Finally, the list of close contacts of the patient "bc20_2" and the contact locations are returned as query results.

[0120] Specifically, if Figure 4 As shown in the figure, in an outdoor environment, taking respiratory disease patients "bc20_0" and "bc20_12" as an example, users who were in contact with individual droplets in the same time and space as "bc20_0" and "bc20_12" when they were in contact with droplets are screened. By entering the corresponding respiratory disease information of patients "bc20_0" and "bc20_12" in the search bar of the back-end system, and combining the stored information, the close contacts of the two respiratory disease patients can be quickly and accurately screened. Finally, the list of close contacts and contact locations of patients "bc20_0" and "bc20_12" are returned as query results. The staff of the CDC can carry out the next step of work based on the obtained close contact list, which can effectively reduce the workload and take good prevention and control measures.

[0121] To sum up, a close contact tracing device based on droplet detection and a tracing method thereof in the present invention, by setting a droplet detection module to detect droplet signals to determine the physical contact between individuals, and recording the geographical location and time when the droplet signal is obtained through the close contact tracing module, can clearly reflect the spread of droplets, the main transmission mode of respiratory diseases. The back-end system has the function of filtering a large amount of redundant information, and can accurately limit the search range to a specific time point and spatial location, so as to find close contacts and secondary close contacts more quickly, so that the epidemic prevention department can lock in high-risk populations and areas in the shortest time, and effectively control the disease within the smallest range.

Claims

1. A close contact tracing device based on droplet detection, characterized in that: It includes a droplet detection module, a close contact tracing module and a back-end system that are electrically connected to each other; The droplet detection module includes a wearable carrier, a droplet sensor is arranged on the outer front of the wearable carrier, a droplet detection circuit board is arranged on the inner side of the wearable carrier, a main control module and a capacitance conversion chip are arranged on the droplet detection circuit board, and the droplet sensor is electrically connected to the main control module through the capacitance conversion chip; The close contact tracing module includes a shell, a display screen is arranged on the outside of the shell, and a tracing circuit board is arranged inside the shell, on which a wireless communication module, an indoor positioning module, an FPGA logic control module, an RTC module, a GPS module and a 4G module are integrated that are electrically connected to each other; the wireless communication module is electrically connected to the main control module; the indoor positioning module collects RSSI data of indoor Beacon devices; the FPGA logic control module is used to identify the droplet signals collected by the droplet sensor and control the GPS module and the RTC module to obtain geographic location information data and time information data; the 4G module uploads packaged data to the back-end system; The back-end system is used to receive, store and visualize geographic location information data, time information data and personal identification information. The back-end system integrates indoor positioning algorithms and close contact tracing algorithms and combines the condition of the patient to achieve spatiotemporal tracing of patients with respiratory diseases.

2. The close contact tracing device based on droplet detection according to claim 1 is characterized in that: The wearable carrier is of a hollow structure and is provided with an ear-hanging rope; a window is provided on the outer front of the wearable carrier, a protective net is provided on the outer side of the window, and the droplet sensor is provided in the window.

3. The close contact tracing device based on droplet detection according to claim 2 is characterized in that: The protective net is composed of a crisscross strip structure.

4. The close contact tracing device based on droplet detection according to claim 1 is characterized in that: The droplet sensor is a capacitive sensor, and a sensitive film is attached to the outer surface of the sensor.

5. The close contact tracing device based on droplet detection according to claim 4 is characterized in that: The main control module is a WIFI module; the capacitance conversion chip is used to convert the droplet signal collected by the droplet sensor into a digital signal.

6. The close contact tracing device based on droplet detection according to claim 2 is characterized in that: The wireless communication module is a WIFI communication module and is electrically connected to the FPGA logic control module. The indoor positioning module is a Bluetooth module and is electrically connected to the FPGA logic control module, and is used to collect RSSI data of indoor Beacon devices; the RTC module is a real-time clock module, which is used to provide accurate time information; the GPS module is a high-precision positioning module based on the global positioning system, which uses the global positioning system satellite network to determine the geographic location information; the 4G module is a high-speed wireless communication module based on a cellular network.

7. A tracing method for a close contact tracing device based on droplet detection according to any one of claims 1 to 6, characterized in that: include: Step 1: After wearing the droplet detection module and the close contact tracing module, start the droplet detection module and the close contact tracing module; Step 2: When the droplet detection module detects the droplet signal, if it is in an indoor environment, the FPGA logic control module controls the indoor positioning module to collect the RSSI data of the indoor Beacon device and package the data and upload it to the back-end system, use the indoor positioning algorithm to locate the device, and store the located location information data in the back-end system; Step 3: When the droplet detection module detects the droplet signal, if it is in an outdoor environment, the FPGA logic control module controls the RTC module and the GPS module to obtain the geographic location information data and time data when the droplets are detected, and the droplet signal and the geographic location information data and time data when the droplets are detected are packaged and uploaded to the back-end system through the 4G module; Step 4: The back-end system stores the droplet signal, the indoor or outdoor geographic location information data and time data when the droplets are detected, and the personal identification information of the user wearing the tracing device; Step 5: Enter the personal identification information of patients with confirmed respiratory diseases into the back-end system, use the KVST-DBSCAN algorithm to screen out the patient's close contacts, and complete the spatiotemporal tracing of patients with respiratory diseases.

8. The tracing method of the close contact tracing device based on droplet detection according to claim 7 is characterized in that: In step 2, the calculation steps of the indoor positioning algorithm include: Step 2.1: The relationship between RSSI and the distance d from the device to the Beacon can be approximately calculated using the path loss model: RSSI=-10nlog 10 (d)+A Where n is the environmental attenuation factor, and A is the RSSI value at the reference point; Step 2.2: Kalman filtering is used to remove dynamic noise in RSSI signals, smooth RSSI values, and improve positioning accuracy: First, predict the RSSI value and prediction error covariance at the next moment: P k|k-1 =AP k-1|k-1 From T +Q in, is the RSSI prediction value at the kth moment, P k|k-1 is the prediction error covariance, A is the state transfer matrix, B is the control matrix, and Q is the process noise covariance matrix; Update the data again: P k|k =(I-K k H)P k|k-1 Among them, K k is the Kalman gain, z k is the current observation value (RSSI measurement value). H is the observation matrix, R is the observation noise covariance matrix; Step 2.3: Gaussian filtering is used to further smooth the RSSI data and reduce the impact of extreme values: Among them, x i is the RSSI measurement value at each moment, and N is the number of measurements; Step 2.4, BRB uses belief rules to reason about RSSI data to determine the possible location of the target device. In BRB reasoning, the possibility P(Location=L) of location L can be expressed as the weighted sum of the confidences of different RSSI values: Among them, β i is the confidence of each rule, P i is the prior probability of each location, and the location probability distribution obtained by BRB is used for rough positioning; Step 2.5: The neural network model uses a multi-layer structure to input the processed RSSI value and output the coordinates of the device. (l) : a (l) =f(W (l) -a (l-1) +b (l) ) Among them, W (l) is the weight matrix of the lth layer, b (l) is the bias, f is the activation function, and during the training process, the mean square error MSE is used as the loss function to calculate the deviation between the predicted position and the actual position: in, is the predicted position, y i is the actual position, m is the number of samples; Step 2.6: Store the final obtained indoor geographic location information data in the back-end system.

9. The tracing method of the close contact tracing device based on droplet detection according to claim 7 is characterized in that: In step 5, the calculation steps of the KVST-DBSCAN algorithm include: Step 5.1: Determine the spatiotemporal distance matrix D of the range dataset D of a cluster n×n : D n×n ={dist(i,j)|1≤i,j≤n} Step 5.2: After rearranging the space-time distance matrix in ascending order, the Kth column vector is the K-nearest neighbor distance vector D of D K , for D K The elements in the vector are averaged to get the K average nearest neighbor distance of D Generate candidate parameter list D of parameter Eps Eps : Step 5.3: Use the mathematical expectation method and the generated parameter Eps candidate parameters to determine the corresponding MinPts: MinPts=min{MinPts1,MinPts2} Among them, N Eps1 (p i ) and N Eps2 (p i ) represent the point p i The total number of points in the Eps1 and Eps2 domains, n is the total number of points contained in the dataset D; Step 5.4: Determine the intra-cluster density of each object in the i-th cluster data. The intra-cluster density Den(ST i ) is expressed as: Among them, N(ST i ) represents the number of data points in the cluster, A(ST j ) indicates N(ST i ) in the j-th object, Mean(ST i ) indicates N(ST i ) the average variance of all objects; Step 5.5: If Den(ST i ) is greater than the density threshold θ and is marked as a noise point. If a user wearing a traceability device is diagnosed as an infected person, other users in the cluster where the user belongs are marked as close contacts, and the close contacts of confirmed respiratory disease patients are finally output.

Images