A highway health code recognition system and working method based on drone

Abstract

The highway health code identification system based on drones described in the present invention includes a drone group, a dispatching center and a prevention and control subsystem, wherein the drone group includes a flow drone, a survey drone and a patrol drone, wherein the flow drone and the survey drone are used to collect information for the jurisdiction division system, and the patrol drone is used to detect the health code; the dispatching center includes a jurisdiction division system and a drone dispatching system, wherein the jurisdiction division system divides the detection area according to the collected information, and the drone dispatching system is used to dispatch drones to execute corresponding instructions; the prevention and control subsystem is used to collect risk information transmitted back by the patrol drone, so that the epidemic prevention system can take corresponding measures. The highway health code is identified through this system to improve the rigor and accuracy of prevention and control.

Description

Technical Field

[0001] The present invention relates to the fields of health protection, drone scheduling and edge computing, and in particular to a highway health code recognition system and working method based on drones. Background Art

[0002] At present, for health prevention and control, when taking a car, train, subway and other means of transportation, you need to show your health code to the staff. Only when the health code is displayed as green can you be allowed to enter. As a way for people to move quickly, the current measures taken by highways are to set up checkpoints at toll stations to check the health codes and other related information of the people in the car one by one. This requires a lot of manpower and time, and increases the risk of contact; and some sections of highways are without toll stations. They are connected to the fast-pass roads in the city, and there are blind spots for detection, resulting in missed inspections; if checkpoints are set up one by one in this situation, the inspection cost will be further increased. In view of this, a contactless, convenient and fast inspection method needs to be considered. Drone technology has become increasingly mature, and it can be intelligently dispatched according to the set tasks and flight routes. At present, it is completely possible to obtain health codes through face recognition technology. Therefore, how to use drone technology to check health codes and other related information, further improve prevention and control, and reduce contact risks is a problem that needs to be studied at present. Summary of the invention

[0003] In order to address the spread caused by the movement of people during prevention and control, and human errors in health code queries, and to improve the rigor and accuracy of prevention and control, the present invention provides a highway health code recognition system and working method based on drones.

[0004] The highway health code recognition system based on drones of the present invention comprises a drone group, a dispatching center and a prevention and control subsystem.

[0005] The drone group includes traffic drones, survey drones and inspection drones, each of which is equipped with a GPS positioning system;

[0006] The drone dispatching system dispatches drones to execute corresponding instructions, and the jurisdiction division system divides the detection area according to the information collected by the traffic drone and the survey drone;

[0007] The traffic drone is equipped with a license plate number recognition system and a face recognition system. The traffic drone receives instructions from the drone dispatching system, hovers in the air at its designated location, collects the traffic volume and pedestrian volume of the road section at each time period through the license plate number recognition system and the face recognition system, and transmits the collected information to the jurisdiction division system;

[0008] The survey drone is provided with an aerial photography device, an electronic speedometer and a rangefinder. The survey drone receives instructions from the drone dispatching system, flies along the expressway at a designated location, collects a panoramic view of the expressway through the aerial photography device, measures the length of each section of the trunk road through the rangefinder, measures the speed of vehicles on the expressway through the electronic speedometer, and then transmits the collected information to the jurisdiction division system;

[0009] The inspection drone is provided with an intelligent control flight system, a face recognition system, a health code detection system, a wireless communication system, and an information computing and storage system, wherein the intelligent control system plans the flight path according to the received information; the face recognition system is connected to the health code database, uses the face recognition system to perform face recognition on the person in the vehicle, and uses the health code detection system to query the health code status of the identified person; the information computing and storage system stores the current situation of the person detected in the vehicle and calculates the abnormal risk value of the vehicle; the wireless communication system is used for communication between inspection drones, and uploads the detected abnormal information to the prevention and control subsystem;

[0010] The prevention and control subsystem receives risk information transmitted by the inspection drone and takes corresponding epidemic prevention measures.

[0011] Furthermore, the intelligent control flight system of the inspection drone is set as follows: according to the six parameters of vehicle size, type, running speed, number of lanes, distance from the fork in the road, and distance from surrounding vehicles, the vehicle t is predicted by the probability graph model. n+1 Position at the moment The future position is predicted by historical trajectory, and the calculation formula is: To predict t n+1 The vector position at the time; x(t n ) is t n The real position at the moment; Δx(t n-1 ) is t n-1 to n The change in position during the time interval Δx(t n-1 )=x(t n )-x(t n-1 ); is time t n The correction factor within where ∑ω i =1,ω i+1 ≤ω i ; The predicted vehicle position in the next second λ∈[0,1];ω iand λ are coefficients; the value of λ is related to the type of vehicle. The value of λ for private cars approaches 1, and the value of λ for trucks and buses approaches 0. Cluster the vehicles according to their locations, find the center of vehicle distribution, and use the direction from the center to the direction with the largest number of undetected vehicles as the main search direction. Within the range allowed by the calculation time, use the vehicle type, size, location, density, and total vehicle risk value as input parameters, and obtain another search direction based on the inductive classification algorithm. Find a location within the communicable range where the most people can be detected as the flight point of the inspection drone in the next second to generate a planned path.

[0012] The method for identifying the highway health code using the above system includes the following steps:

[0013] Step 1: The dispatch center receives the highway health code recognition request, and the drone dispatch system dispatches traffic drones and survey drones to collect information;

[0014] Step 2: The jurisdiction division system divides the highway detection area according to the collected information and forms a detection plan;

[0015] Step 3: The UAV dispatching system dispatches inspection UAVs to the designated location for inspection according to the inspection plan; the inspection UAVs cooperate with each other and conduct flight inspections according to the road conditions under the control of the intelligent control flight system;

[0016] Step 4: The inspection drone identifies the license plate number. If the safety threshold is not reached, go to step 5; otherwise, go to step 8;

[0017] Step 5: The inspection drone performs anomaly detection to determine whether an abnormality is detected; if so, execute step 6; if not, execute step 7;

[0018] Step 6: According to the different anomalies set by the system, determine the anomaly type, add risk values, forward to the next drone and control subsystem, and execute step 8;

[0019] Step 7: The inspection drone determines whether to increase the safety threshold;

[0020] Step 8: Determine whether the detection task is completed. If so, execute step 9; otherwise, execute step 4 according to the planned path;

[0021] Step 9. Mission completed, recover the drone.

[0022] Furthermore, the steps for the drone dispatching system in step 1 to dispatch traffic drones and survey drones to collect information are as follows:

[0023] Step 101: A survey drone and a traffic drone take off at the same time and head to a designated location on the highway;

[0024] Step 102: The traffic drone hovers at a designated location on the highway and records the number of vehicles passing through and the number of people on the vehicles at each moment through a license plate number recognition device; after the information collection is completed, the traffic drone directly flies back to the dispatch center according to the straight-line distance;

[0025] Step 103, the survey drone flies along the main road path of the expressway. When encountering a fork in the road, the survey drone flies along the original road and transmits the coordinate position of the fork in the road to the dispatch center. The dispatch center dispatches a survey drone and a traffic drone on each main road according to the situation of the fork in the road, and executes step 101;

[0026] Step 104: After completing the information collection task, the survey drone directly flies back to the dispatch center.

[0027] Furthermore, the information collected in step 1 includes a panoramic view of the expressway, the length and width of each trunk road of the expressway, the traffic volume of each trunk road, the passenger volume of each trunk road, and the speed of vehicles at different points on the expressway.

[0028] Furthermore, in step 2, the detection scheme described above is based on the information collected in step 1. The jurisdiction division system performs fitting analysis based on the area, traffic flow, pedestrian flow, vehicle speed, and speed of the inspection drone for face recognition and health code detection system on the highway to obtain an area S with a relatively consistent detection task density. r ={s1, s2, ..., s n ,s r}, Inspection UAV r = {u1, u2, ..., u n ,u r},s i and u i One-to-one correspondence, a patrol drone detects vehicles and personnel in its corresponding area.

[0029] Furthermore, in step 3, the process of mutual cooperation between the inspection drones is as follows: the inspection drone needs to predict whether it can detect all the people in the same car in the area. If t 车 ≥t 传 +t 处 , that is, the inspection can be carried out on the inspection drone of this section, where t 车 is the time required for the UAV to start receiving the information of the person in vehicle i and drive out of this area. Assume that the vehicle is moving at a constant speed in each detection area, and the speed is V i , t 车 =L i / V i , L iThe distance that the UAV needs to travel from the time it starts receiving the information of the person in vehicle i to the time it drives out of this area; t 传 =O i / R i , t 传 represents the transmission time from vehicle i to the UAV for sending the information of all people in the vehicle, O i represents the size of the health code information of all persons in vehicle i, R i represents the transmission rate of the health code information of all persons in vehicle i, R i =W*log2(1+(P i *d i，u -α )(N0+∑ s∈N，s≠i P s *d s，u -α ))W represents the channel bandwidth, P i represents the transmission power of vehicle i, d i，u represents the distance between vehicle i and drone u, α represents the channel loss parameter, N0 represents the noise power, ∑ s∈N，s≠i P s *d s，u -α represents the impact value of other vehicles except vehicle i transmitting detection information to the drone at the same time; t 处 Indicates the time required to process the health code of the person on the vehicle; otherwise, t 车 <t 传 +t 处 , the inspection drone does not detect the people in the vehicle, but acts as a relay node to directly forward the information about the vehicle and all the people in the vehicle sent by the inspection drone on the previous route to all the inspection drones on the next route that the vehicle may enter; after receiving the information, the inspection drone on the next route will, according to the planned path, 传 +t 处 +Δt time to detect the vehicle information, continue to predict according to the above process, and the rest in t 传 +t 处 If the inspection drone does not detect the vehicle information within +Δt, the vehicle information will be deleted, where the value of Δt is the maximum time required for a vehicle to pass the next road section, Δt = L / Vmin, L is the distance the vehicle needs to travel through the road section, V min It is the minimum speed of vehicles passing through this road section.

[0030] Furthermore, in step 5, the abnormal conditions include:

[0031] (1) The inspection drone identifies the person through the face recognition system, and the health code detection system detects that the health code is yellow;

[0032] (2) The inspection drone identifies the person through the face recognition system, and the health code detection system detects that the health code is red;

[0033] (3) Inspection drones can identify people through the facial recognition system, but cannot detect health code status through the health code detection system;

[0034] (4) The inspection drone cannot identify people through the facial recognition system.

[0035] Furthermore, in step 6, the risk value refers to the severity of the abnormal situation in the vehicle. The larger the risk value, the higher the degree of danger; vice versa. Different risk values ​​are set for different abnormal types according to the degree of risk. When the abnormality is detected, the risk values ​​are superimposed and forwarded to the next drone and the prevention and control subsystem. The specific process is as follows: each car corresponds to a total risk value, and different abnormal types correspond to a risk value. Whenever the inspection drone detects one of the abnormal types in a car, the risk value of the corresponding abnormal situation of the vehicle is superimposed, and the total risk value is superimposed at the same time. After the inspection of all personnel in the vehicle is completed, the updated result value is immediately forwarded to the inspection drone and the prevention and control subsystem of the next section.

[0036] Furthermore, in step 7, the inspection drone determines whether to increase the safety threshold as follows:

[0037] Step 701: The inspection drone determines whether all persons in the vehicle have been inspected. If so, execute step 702; if not, execute step 5;

[0038] Step 702: The inspection drone superimposes the safety value of the vehicle to determine whether the safety threshold is reached. If so, execute step 703; if not, execute step 704;

[0039] Step 703: forward the vehicle information to all inspection drones on the next road section. After the inspection drones on the next road section identify the vehicle, there is no need to inspect the vehicle.

[0040] Step 704: forward the vehicle and personnel information to all inspection drones on the next road section. The inspection drones on the next road section need to continue to inspect the personnel on the vehicle according to the set planned path. If an abnormality is detected, the safety value is reset and step 6 is executed; otherwise, step 701 is executed.

[0041] The beneficial effects of the present invention are:

[0042] 1. The present invention sets up an intelligent control flight system for the inspection drone, plans a path that can detect the most personnel information in the area, and avoids contactless verification of personnel who are negligent or preparing to leave the area through other means;

[0043] 2. The present invention divides the detection area by fitting through the jurisdiction division system, designs the process of how drones in multiple areas cooperate with each other, can make full use of existing drone resources, achieve more reasonable division, no omissions in personnel detection, and conduct multiple inspections of personnel to reduce errors. At the same time, the operation through drones greatly reduces human resources;

[0044] 3. The present invention performs detection according to the priority of abnormalities, and can submit abnormal situations as soon as possible to carry out prevention and control with the greatest effectiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a schematic flow chart of the working method of the present invention.

[0046] Figure 2 This is a picture of the inspection drone working in this inspection area.

[0047] Figure 3 It is a communication diagram of the inspection drone during the vehicle movement process. DETAILED DESCRIPTION

[0048] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings of the specification. This embodiment is only used to explain the present invention and does not constitute a limitation on the protection scope of the present invention.

[0049] The highway health code recognition system based on drones described in the present invention includes a drone group, a dispatching center and a prevention and control subsystem. The drone group includes a traffic drone, a survey drone and an inspection drone. The traffic drone is provided with a license plate recognition system and a face recognition system; the survey drone is provided with aerial photography equipment, an electronic speedometer, and a rangefinder; the inspection drone is provided with an intelligent control flight system, a face recognition system, a health code detection system, a wireless communication system, and an information computing and storage system; the dispatching center includes a jurisdiction division system and a drone dispatching system.

[0050] Among them, the traffic drone hovers at a fixed point in the air, obtains the traffic volume of the section through the license plate recognition device, and obtains the pedestrian flow through the portrait recognition system.

[0051] The survey drones have aerial photography equipment, which is used to take panoramic views of the highway; electronic speedometers, which are used to measure the speed of vehicles on the highway and record the information on the panoramic view obtained by the aerial photography equipment; and rangefinders, which are used by the aerial photography equipment to measure the length of various parts of the highway when taking panoramic views;

[0052] On the inspection drone, the intelligent control flight system predicts the vehicle t based on the vehicle size, type, running speed, number of lanes, distance from the fork, distance from surrounding vehicles and other parameters through the Markov random field. n+1 Vector position at time The future position is predicted by historical trajectory, and the calculation formula is: To predict t n+1 The vector position at the time; x(t n ) is t n The real position at the moment; Δx(t n-1 ) is t n-1 to n The change in position during the time interval Δx(t n-1 )=x(t n )-x(t n-1 ); is time t n The correction factor within where ∑ω i =1,ω i+1 ≤ω i ; The predicted vehicle position in the next second λ∈[0,1], the value of λ depends on the type of car. The value of λ for private cars approaches 1, while the value of λ for trucks and buses approaches 0. The K-Means method is used for clustering to find the center point of vehicle distribution. The direction from the center point to the direction with the largest number of undetected vehicles is used as the main search direction. Within the scope allowed by the calculation time, the vehicle type, size, location, density, total vehicle risk value, etc. are used as input parameters, and another search direction is obtained according to the decision tree algorithm. A location where the most people can be detected within the communication range is found as the flight point of the drone in the next second to generate a planned path. Among them, the decision tree algorithm is only one of the inductive classification algorithms, but it is not limited to this inductive classification algorithm. The face recognition system is used for drones to perform face recognition on people in the vehicle. The health code detection system works in conjunction with the face recognition system to retrieve the health code of the passengers with face recognition and query their health code status. The wireless communication system is used for error-free and stable communication between drones, and is also used for drones to upload tracked abnormal information to the prevention and control subsystem. The information computing and storage system is used to store the current situation of the people monitored in the vehicle and calculate the abnormal risk value of the vehicle. Each drone is equipped with a GPS positioning system.

[0053] The dispatch center includes a jurisdiction division system and a drone dispatch system. The jurisdiction division system divides the detection area according to the traffic volume, pedestrian flow, vehicle speed, road width, and road length obtained by traffic drones and survey drones; the drone dispatch system assigns traffic drones to arrive at designated locations to obtain vehicle and pedestrian flow; assigns survey drones to take panoramic photos of the highways that need to be inspected, collect vehicle speeds, and measure various attributes of the highways; assigns inspection drones to fly to their corresponding monitoring areas and maintain communication with the drones. The prevention and control subsystem is responsible for receiving risk information on highways transmitted by drones, so that the epidemic prevention system can take corresponding measures.

[0054] Combination Figure 1 The specific working method of using the above system to identify the health code on the highway is as follows:

[0055] Step 1: The dispatch center receives the highway health code recognition request, and the drone dispatch system dispatches traffic drones and survey drones to collect information; the collected information includes a panoramic view of the highway, the length and width of each highway, the traffic volume of each highway, the passenger flow of each highway, and the speed of vehicles at different points on the highway;

[0056] Step 2: The jurisdiction division system divides the highway detection area according to the collected information to form a detection plan; the detection plan, based on the information collected in step 1, the jurisdiction division system performs fitting analysis based on the area, traffic flow, pedestrian flow, vehicle speed, and speed of the drone face recognition and health code detection system of the highway to obtain an area S with a relatively consistent detection task density. r ={s1,s2,…,s n ,s r}, Inspection UAV r ={u1,u2,…,u n ,u r},s i and u i One-to-one correspondence, one inspection drone detects vehicles and personnel in its corresponding area;

[0057] Step 3: The UAV dispatching system dispatches inspection UAVs to the designated location for inspection according to the inspection plan; the inspection UAVs cooperate with each other to conduct flight inspections according to the road conditions under the control of the intelligent flight control system, such as Figure 2 As shown;

[0058] Step 4: The inspection drone identifies the license plate number. If the safety threshold is not reached, go to step 5; otherwise, go to step 8;

[0059] Step 5: The inspection drone performs anomaly detection to determine whether an abnormality is detected. If so, execute step 6; if not, execute step 7;

[0060] Step 6: According to the different anomalies set by the system, determine the anomaly type, add risk values, forward to the next drone and control subsystem, and execute step 8;

[0061] Step 7: The inspection drone determines whether to increase the safety threshold;

[0062] Step 8: Determine whether the detection task is completed. If so, execute step 9; otherwise, execute step 4 according to the planned path;

[0063] Step 9. Mission completed, recover the drone.

[0064] In step 1, the steps for the drone dispatching system to dispatch traffic drones and survey drones to collect information are as follows:

[0065] Step 101: A survey drone and a traffic drone take off at the same time and head to a designated location on the highway;

[0066] Step 102: The traffic drone hovers at a designated location on the highway and records the number of vehicles passing through and the number of people on the vehicles at each moment through a license plate number recognition device; after the information collection is completed, the traffic drone directly flies back to the dispatch center according to the straight-line distance;

[0067] Step 103: The survey drone flies along the main road path of the expressway. When encountering a fork in the road, the survey drone flies along the original road and transmits the coordinate position of the fork in the road to the dispatch center. The dispatch center dispatches a survey drone and a traffic drone on a main road according to the situation of the fork in the road, and executes step 101.

[0068] Step 104: After completing the information collection task, the survey drone directly flies back to the dispatch center.

[0069] In step 3, the process of mutual cooperation between inspection drones is as follows: the inspection drone needs to predict whether it can detect all the people in the same car in the area. If t 车 ≥t 传 +t 处 , that is, the inspection can be carried out on the inspection drone of this section, where t 车 is the time required for the UAV to start receiving the information of the person in vehicle i and drive out of this area. Assume that the vehicle is moving at a constant speed in each detection area, and the speed is V i , t 车 =L i / V i , L iThe distance that the UAV needs to travel from the time it starts receiving the information of the person in vehicle i to the time it drives out of this area; t 传 =O i / R i , t 传 represents the transmission time from vehicle i to the UAV for sending the information of all people in the vehicle, O i represents the size of the health code information of all persons in vehicle i, R i represents the transmission rate of the health code information of all persons in vehicle i, R i =W*log2(1+(P i *d i，u -ɑ ) / (N0+∑ s∈N，s≠i P s *d s，u -ɑ )), where W represents the channel bandwidth, P i represents the transmission power of vehicle i, d i，u represents the distance between vehicle i and drone u, α represents the channel loss parameter, N0 represents the noise power, ∑ s∈N，s≠i P s *d s，u -ɑ represents the impact value of other vehicles except vehicle i transmitting detection information to the drone at the same time; t 处 Indicates the time required to process the health code of the person on the vehicle. Otherwise, t 车 <t 传 +t 处 , the inspection drone does not detect the people in the vehicle, but acts as a relay node to directly forward the information about the vehicle and all the people in the vehicle sent by the inspection drone on the previous route to all the inspection drones on the next route that the vehicle may enter; after receiving the information, the inspection drone on the next route will, according to the planned path, 传 +t 处 +Δt time to detect the vehicle information, continue to predict according to the above process, and the rest in t 传 +t 处 If the inspection drone does not detect the vehicle information within +Δt, the vehicle information will be deleted, where the value of Δt is the maximum time required for a vehicle to pass the next road section, Δt = L / V min , is the distance that the vehicle needs to travel through the road section, V min It is the minimum speed of vehicles passing through this road section.

[0070] In step 5, there are various abnormal situations, including:

[0071] (1) The drone identifies the person through the face recognition system, and the health code detection system detects that the health code is yellow;

[0072] (2) The drone identifies the person through the face recognition system, and the health code detection system detects that the health code is red;

[0073] (3) Drones can identify people through the facial recognition system, but cannot detect health code status through the health code detection system;

[0074] (4) The drone cannot identify people through its facial recognition system.

[0075] In step 6, the risk value refers to the severity of the abnormal situation in the vehicle. The larger the risk value, the higher the danger level; vice versa. The above abnormal situations (1) to (4) are set with different risk values ​​according to the risk level.

[0076] Among them, in step 6, the abnormality is detected, the risk value is superimposed, and forwarded to the next drone and the prevention and control subsystem. The specific process is as follows: each car corresponds to a total risk value, and the abnormal situation (1) to (4) corresponds to a risk value respectively. Whenever the inspection drone detects one of the abnormal situations (1) to (4) in a car, the risk value of the corresponding abnormal situation of the vehicle is superimposed, and the total risk value is superimposed at the same time; after the inspection of all personnel in the vehicle is completed, the updated result value is immediately forwarded to the drone and the prevention and control subsystem of the next section, such as Figure 3 ;

[0077] In step 7, the inspection drone determines whether to increase the safety threshold as follows:

[0078] Step 701: The inspection drone determines whether all persons in the vehicle have been inspected. If so, execute step 702; if not, execute step 5;

[0079] Step 702: The inspection drone superimposes the safety value of the vehicle to determine whether the safety threshold is reached. If so, execute step 703; if not, execute step 704;

[0080] Step 703: forward the vehicle information to all inspection drones on the next road section. After the inspection drones on the next road section identify the vehicle, there is no need to inspect the vehicle.

[0081] Step 704: forward the vehicle and personnel information to all inspection drones on the next road section. The inspection drones on the next road section need to continue to inspect the personnel on the vehicle according to the planned path. If an abnormality is detected, the safety value is reset and step 6 is executed; otherwise, step 701 is executed.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to be a further limitation of the present invention. All equivalent changes made using the contents of the present specification and drawings are within the protection scope of the present invention.

Claims

1. The highway health code recognition system based on drones is characterized by: Including drone groups, dispatch centers and prevention and control subsystems, The drone group includes traffic drones, survey drones and inspection drones, each of which is equipped with a GPS positioning system; The dispatch center includes a jurisdiction division system and a drone dispatch system. The drone dispatch system dispatches drones to execute corresponding instructions. The jurisdiction division system divides the detection area according to the information collected by the traffic drone and the survey drone. The traffic drone is equipped with a license plate number recognition system and a face recognition system. The traffic drone receives instructions from the drone dispatching system, hovers in the air at its designated location, collects the traffic volume and pedestrian volume at each time period at the designated location through the license plate number recognition system and the face recognition system, and transmits the collected information to the jurisdiction division system; The survey drone is provided with an aerial photography device, an electronic speedometer and a rangefinder. The survey drone receives instructions from the drone dispatching system, flies along the expressway at a designated location, collects a panoramic view of the expressway through the aerial photography device, measures the length of each section of the trunk road through the rangefinder, measures the speed of vehicles on the expressway through the electronic speedometer, and then transmits the collected information to the jurisdiction division system; The inspection drone is provided with an intelligent control flight system, a face recognition system, a health code detection system, a wireless communication system, and an information computing and storage system, wherein the intelligent control system plans the flight path according to the received information; the face recognition system is connected to the health code database, and uses the face recognition system to perform face recognition on the people in the vehicle and uses the health code detection system to query the health code status of the people in the vehicle; the information computing and storage system stores the current detected situation of the people in the vehicle and calculates the abnormal risk value of the vehicle; The wireless communication system is used for communication between inspection drones and uploading detected abnormal information to the prevention and control subsystem; The prevention and control subsystem receives risk information transmitted by the inspection drone and takes corresponding epidemic prevention measures; The intelligent control flight system of the inspection drone is set as follows: according to the six parameters of vehicle size, type, running speed, number of lanes, distance from the fork in the road, and distance from surrounding vehicles, the vehicle t is predicted by the probability graph model. n+1 Vector position at time The future position is predicted by historical trajectory, and the calculation formula is: To predict t n+1 The vector position at the time; x(t n ) is t n The real position at the moment; Δx(t n-1 ) is t n-1 to n The change in position during the time interval Δx(t n-1 )=x(t n )-x(t n-1 ); is time t n The correction factor within where ∑ω i =1,ω i+1 ≤ω i ; The predicted vehicle position in the next second λ∈[0,1]; ω i and λ are coefficients; cluster the vehicles according to their positions, find the center of the vehicle distribution, and use the direction from the center to the direction with the largest number of undetected vehicles as the main search direction. Within the range allowed by the calculation time, use the vehicle type, size, position, density, and total vehicle risk value as input parameters, and obtain another search direction based on the inductive classification algorithm. Find a location where the most people are detected within the communication range as the flight point of the inspection drone in the next second, and generate a planned path.

2. The method for identifying the highway health code recognition system based on drone according to claim 1 is characterized in that: The following steps are involved: Step 1: The dispatch center receives the highway health code recognition request, and the drone dispatch system dispatches traffic drones and survey drones to collect information; Step 2: The jurisdiction division system divides the highway detection area according to the collected information and forms a detection plan; Step 3: The drone dispatching system dispatches the inspection drone to the designated location for inspection according to the inspection plan; Inspection drones work together to conduct flight inspections based on road conditions and under the control of the intelligent flight control system; Step 4: The inspection drone identifies the license plate number. If the safety threshold is not reached, go to step 5; otherwise, go to step 8; Step 5: The inspection drone performs anomaly detection to determine whether an abnormality is detected; if so, execute step 6; if not, execute step 7; Step 6: According to the different anomalies set by the system, determine the anomaly type, add risk values, forward to the next drone and control subsystem, and execute step 8; Step 7: The inspection drone determines whether to increase the safety threshold; Step 8: Determine whether the detection task is completed. If so, execute step 9; otherwise, execute step 4 according to the planned path; Step 9: Mission completed, drone recovered; In step 3, the process of mutual cooperation between inspection drones is as follows: the inspection drone needs to predict whether it can detect all the people in the same car in the area. If t 车 ≥t 传 +t 处 , the inspection is performed on the inspection drone of the road section, where t 车 is the time required for the UAV to start receiving the information of the person in vehicle i and drive out of this area. Assume that the vehicle is moving at a constant speed in each detection area, and the speed is V i , t 车 =L i / V i ,L i The distance that the UAV needs to travel from the time it starts receiving the information of the person in vehicle i to the time it drives out of this area; t 传 =O i / R i ,t 传 represents the transmission time from vehicle i to send the information of all people in the vehicle to the drone, O i represents the size of the health code information of all persons in vehicle i, R i represents the transmission rate of the health code information of all persons in vehicle i, R i =W*log2(1+(P i *d i,u -α ) / (N0+∑ s∈N,s≠i P s *d s,u -α )), where W represents the channel bandwidth, P i represents the transmission power of vehicle i, d i,u represents the distance between vehicle i and drone u, α represents the channel loss parameter, N0 represents the noise power, ∑ s∈N,s≠i P s *d s,u -α It represents the impact value of other vehicles except vehicle i transmitting detection information to the UAV at the same time; t 处 Indicates the time required to process the health code of the person on the vehicle; otherwise, t 车 <t 传 +t 处 The inspection drone does not detect the people in the vehicle, but acts as a relay node to directly forward the information about the vehicle and all the people in the vehicle transmitted by the inspection drone on the previous route to all the inspection drones on the next route that the vehicle enters; After receiving the information, the inspection drone of the next section will follow the planned path. 传 +t 处 +Δt time to detect the vehicle information, continue to predict according to the above process, and the rest in t 传 +t 处 If the inspection drone does not detect the vehicle information within +Δt, the vehicle information will be deleted, where the value of Δt is the maximum value of the time required for a vehicle to pass the next road section, Δt = L / V min , L is the distance that the vehicle needs to travel through the road section, V min It is the minimum speed of vehicles passing through this road section.

3. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: In step 1, the steps for the drone dispatching system to dispatch traffic drones and survey drones to collect information are as follows: Step 101: A survey drone and a traffic drone take off at the same time and head to a designated location on the highway; Step 102: The traffic drone hovers at a designated location on the highway and records the number of vehicles passing through and the number of people on the vehicles at each moment through a license plate number recognition device; after the information collection is completed, the traffic drone directly flies back to the dispatch center according to the straight-line distance; Step 103, the survey drone flies along the main road path of the expressway. When encountering a fork in the road, the survey drone flies along the original road and transmits the coordinate position of the fork in the road to the dispatch center. The dispatch center dispatches a survey drone and a traffic drone on each main road according to the situation of the fork in the road, and executes step 101; Step 104: After completing the information collection task, the survey drone directly flies back to the dispatch center.

4. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: The information collected in step 1 includes a panoramic view of the expressway, the length and width of each trunk road of the expressway, the traffic volume of each trunk road, the passenger volume of each trunk road, and the speed of vehicles at different points on the expressway.

5. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: In step 2, the detection scheme described above is based on the information collected in step 1. The jurisdiction division system performs fitting analysis based on the area, traffic flow, pedestrian flow, vehicle speed, and speed of the inspection drone for face recognition and health code detection system on the highway to obtain an area S with a relatively consistent detection task density. r ={s1,s2,…,s n ,s r }, Inspection UAV r ={u1,u2,…,u n ,u r },s i and u i One-to-one correspondence, a patrol drone detects vehicles and personnel in its corresponding area.

6. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: In step 5, the abnormal conditions include: (1) The drone identifies the person through the face recognition system, and the health code detection system detects that the health code is yellow; (2) The drone identifies the person through the face recognition system, and the health code detection system detects that the health code is red; (3) The drone can identify people through the facial recognition system, but cannot detect the health code status through the health code detection system; (4) The drone cannot identify people through its facial recognition system.

7. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: In step 6, the risk value refers to the severity of the abnormal situation in the vehicle. The larger the risk value, the higher the danger level; otherwise, the lower the risk value. Set different risk values ​​for different types of anomalies according to their risk levels; When an abnormality is detected, the risk values ​​are superimposed and forwarded to the next drone and prevention and control subsystem. The specific process is as follows: each car corresponds to a total risk value, and different abnormal types correspond to a risk value. Whenever the inspection drone detects one of the abnormal types in a car, the risk value of the corresponding abnormal situation of the vehicle is superimposed, and the total risk value is superimposed at the same time; after the inspection of all personnel in the vehicle is completed, the updated result value is immediately forwarded to the inspection drone and prevention and control subsystem of the next section.

8. The method for identifying highway health codes based on drones according to claim 2 is characterized in that: In step 7, the inspection drone determines whether to increase the safety threshold as follows: Step 701: The inspection drone determines whether all persons in the inspected vehicle have been inspected. If so, execute step 702; if not, execute step 5; Step 702: The inspection drone superimposes the safety value of the inspected vehicle to determine whether the safety threshold is reached. If so, execute step 703; If not, execute step 704; Step 703: forward the information of the inspected vehicle to all inspection drones of the next road section. After the inspection drones of the next road section identify the vehicle, there is no need to inspect the vehicle. Step 704: forward the inspected vehicle and personnel information to all inspection drones on the next road section. The inspection drones on the next road section need to continue inspecting the personnel on the vehicle according to the set planned path. If an abnormality is detected, the safety value is reset and step 6 is executed; otherwise, step 701 is executed.

Images