Traffic fault warning device and method based on multi-UAV collaboration

Abstract

The present invention discloses a traffic fault warning device and method based on multi-UAV cooperation. The warning device includes a fuselage, a body, an arm, a GPS locator, a camera, a voice system, an infrared rangefinder, a single-chip controller, a main warning light, a secondary warning light, a direction indicator and a battery panel. When the present invention is used for warning, when a fault occurs on an urban road, the first UAV collects road conditions and vehicle information of the accident section, and the remaining UAVs obtain data of the entire traffic area and perform signal intersection delay T on the path. ij , Actual speed of vehicles on the road section Q ij 、Real-time road conditions ij When a traffic accident occurs on a highway, a swarm of drones takes off and calculates the maximum traffic volume, diverting the following vehicles based on the calculated results. This invention monitors road traffic in real time, identifies and warns of traffic failures, and takes emergency measures, ensuring safe and smooth road traffic.

Description

Technical Field

[0001] The present invention relates to a traffic fault warning device, and in particular to a traffic fault warning device and method based on multi-UAV collaboration. Background Art

[0002] With the continuous increase in the number of cars on the road, the incidence of traffic accidents on urban roads and highways is on the rise. When a vehicle needs to stop for an emergency, a warning triangle must be placed as soon as possible at a safe distance behind the accident vehicle to avoid secondary or tertiary accidents at the scene. This ensures the timely arrival of rescue organizations on urban roads and the diversion of vehicles behind on highways.

[0003] Due to the varying road conditions at accident scenes, the required safe placement of tripods also varies. Conventional tripod warning sign placement presents significant risks, requiring manual placement 50 to 100 meters behind the accident vehicle. On highways, warning signs are typically placed 150 meters behind the vehicle. In rainy or foggy weather, the distance is increased to 200 meters. However, the precise placement of tripod warning signs during an accident is difficult to control, as vehicle speeds on urban roads are relatively low. However, tripod warning signs must be placed for oncoming vehicles on highways.

[0004] The current automatic warning algorithm has false alarms in areas with large pedestrian flow and dense traffic. How to combine drones with warning algorithms and apply them to traffic fault warnings so that the warning device can move independently, help evacuate traffic, and guide rescue routes is a technical issue worth exploring. Summary of the Invention

[0005] Purpose of the invention: In response to the shortcomings of the existing technology, the present invention proposes a traffic fault warning device and method based on multi-UAV collaboration. By adopting multiple UAVs for collaborative monitoring, based on the Dijkstra algorithm and the Expressway Traffic Distribution Algorithm, accidents on urban roads and highways are analyzed and processed, thereby improving the accuracy and speed of fault identification; achieving all-round monitoring of road traffic conditions, avoiding monitoring blind spots and blind spots, improving monitoring coverage and accuracy, and thus effectively warning of traffic at the accident site and providing traffic evacuation and route guidance.

[0006] Technical solution: The traffic fault warning device based on multi-UAV collaboration of the present invention includes a fuselage, a body, an arm, a GPS locator, a camera, a voice system, an infrared rangefinder, a single-chip controller, a main warning light, a secondary warning light, a direction indicator light and a battery panel;

[0007] The camera and infrared rangefinder are located at the bottom of the fuselage, and the voice system is located on the fuselage; the GPS locator and single-chip microcomputer controller are located inside the fuselage, the main warning light is located at the top of the fuselage, the auxiliary warning light is located at the bottom of the fuselage arm, and the direction indicator light is located above the main warning light.

[0008] The GPS locator includes a positioning system and a communication system. The positioning system receives signals from satellites in the sky and calculates the geographic coordinates of the locator; the communication system transmits the coordinate information calculated by the positioning system to the server.

[0009] The traffic fault warning method based on multi-UAV collaboration of the present invention comprises the following steps:

[0010] (1) When an accident occurs on a city road, multiple drones are activated. After the first drone is launched, a warning sign is placed behind the faulty vehicle.

[0011] (2) Turn off the infrared rangefinder and turn on the voice system to warn vehicles coming from behind;

[0012] (3) The camera collects road condition information. If the vehicle is on the left side of the road, the right running light is on; if the vehicle is in the middle of the road, the left and right running lights are on at the same time; if the vehicle is on the right side of the road, the left running light is on;

[0013] (4) The first drone collects road conditions and vehicle information on the accident section, and the remaining drones obtain data for the entire traffic area;

[0014] (5) After the drone completes data collection, it transmits the data to the microcontroller to calculate the signal intersection delay T of the path. ij , Actual speed of vehicles on the road section Q ij 、Real-time road conditions ij Calculation of three data;

[0015] (6) Substitute the calculated results of the three data into the path difficulty Z ij =w1T ij +w2Q ij +w3L ij Calculate the path difficulty of each path;

[0016] (7) Z ij Substitute the Dijkstra algorithm to calculate the optimal path, and send the optimal path to the accident site to the rescue organization.

[0017] The calculation formula of step (5) is as follows:

[0018]

[0019]

[0020] L ij =α1U1+α2(U2+U3+U4) (2)

[0021] Where d is the green light time; t is the signal cycle duration; λ is the green-to-signal ratio; c is the number of vehicles passing through the exit per hour; q is the traffic volume of the entrance per hour; t1 is the driving time of a vehicle; x is the displacement of the front or rear of the vehicle within time t1; v0 is the maximum speed allowed on the road section when there are no vehicles on the road; U1 is the road surface abnormality caused by special weather, U2 is the road surface construction, U3 is the road section with damaged road surface that has not yet been repaired, U4 is the peak section, α1 is the weather impact factor, and α2 is the road surface condition factor.

[0022] In step (5), α1+α2=1.

[0023] The green-to-signal ratio is the ratio of the green light time to the signal cycle length.

[0024] The present invention is based on a multi-UAV collaboration traffic fault warning method, comprising the following steps:

[0025] (1) When a traffic accident occurs on a highway, a swarm of drones takes off and hovers at intervals behind the accident scene. The infrared rangefinder of the first drone is turned on. When the vehicle behind is several meters away from the first drone, the warning light turns on. Each drone detects the frequency of vehicles entering and exiting and changing lanes in its own lane.

[0026] (2) In the hour t1, suppose all vehicles change lanes n times, and the lane-changing frequency During this period, a total of a vehicle left the road monitored by the drone group, and b vehicles entered the area monitored by the drone group, then the frequency of leaving Entry frequency

[0027] (3) Upload the data from step (2) to the microcontroller and calculate the traffic composition correction coefficient f according to the following formula HV and slope correction factor f LG :

[0028]

[0029]

[0030] The theoretical multi-lane capacity calculation formula is as follows:

[0031] N 多 =N i ·∑K n

[0032] The leftmost lane of the expressway is designed for a speed of 120 km / h.

[0033] f SW =1

[0034] f W =1

[0035] This leads to the following three formulas:

[0036]

[0037]

[0038]

[0039] The maximum traffic volume is calculated, where f HV The correction factor representing the presence of traffic components, Represents the actual traffic capacity of the leftmost road of the highway, Represents the actual traffic capacity of the middle road of the highway, Represents the actual traffic capacity of the rightmost road of the highway;

[0040] Where p i is the proportion of each model; E i is the small amount conversion factor; P T is the proportion of trucks and buses; P R The proportion of trucks and tourist cars; E i E is the equivalent value of trucks and buses converted into passenger cars; R is the equivalent value of trucks and tourist buses converted into passenger cars; f LG is the longitudinal slope correction coefficient; f SW f is the correction factor for hard shoulder width; W is the lane width correction factor; N i K is the capacity of the leftmost lane of the highway; n is the reduction coefficient corresponding to the lane; t1 is the time; n is the total number of lane changes by all vehicles; a is the number of vehicles leaving the road monitored by the drone swarm; b is the number of vehicles entering the area monitored by the drone swarm;

[0041] (4) Divert the vehicles behind based on the calculation results.

[0042] In step (1), after the first drone is started, a warning sign is placed 50 to 100 meters behind the faulty vehicle.

[0043] In step (1), the UAV obtains the road network plan and establishes a labeled mathematical model to represent the planned path with a road network structure diagram:

[0044] G=(V ij , E ij , Z ij)

[0045] where node u ij The set of V ij , E ij is the weight set path set, Z ij The difficulty of the path.

[0046] Working Principle: The drone automatically activates when a traffic accident occurs. If the breakdown is caused solely by a vehicle malfunction, the drone can be activated with a single click. After the drone is activated, the GPS locator 2 located within the drone determines the vehicle's location, providing traffic control authorities with the precise location of the accident. Furthermore, if the vehicle is located on an urban road, the warning system's voice system is activated, the infrared rangefinder is deactivated, and the traffic warning system's warning lights are fully activated. Urban roads are densely populated and traffic is slow, so the warning lights are fully activated. Cameras are activated to collect ground information for road section determination, and the drone's direction lights guide the vehicle out of the accident lane. Simultaneously, other drones collect information on surrounding road conditions. Through algorithmic calculations, an optimal route is calculated and provided to rescue organizations.

[0047] If the vehicle is on a highway, the voice system of the warning device is turned off. Because the speed of vehicles on highways is high, the voice warning effect is not significant. Therefore, the first drone uses an infrared rangefinder to detect that the rear vehicle is 500 meters away from the drone, and the warning light turns on. The voice system is turned off at this time, and the warning light is turned on again when the rear vehicle is 500 meters away from the first drone. Both are to increase the endurance of the system. The camera is turned on to collect ground information to determine the road section and collect traffic information. By calculating the road's traffic capacity, multiple drones guide the direction of each section and perform staged evacuation to ensure the smooth flow of the road section. The specific concept of the present invention is as follows:

[0048] On urban roads, multiple drones collaborate to calculate the optimal route by collecting three data points: signal intersection delays, actual vehicle speeds, and real-time road conditions. Using three formulas, different values ​​are calculated for practical scenarios. The calculated influence coefficients w1, w2, and w3 are then used to calculate the difficulty of the road. The optimal route is then calculated using the Dijkstra algorithm, which calculates the shortest route and sends it to rescue organizations, ensuring that emergency vehicles can reach the accident scene as quickly as possible.

[0049] The drone first obtains the road network plan and establishes a labeled mathematical model to represent the planned path using a road network structure diagram:

[0050] G=(V ij , E ij , Z ij )

[0051] Among them, Vij is the u in the road network structure ij The node set and weight set path set are represented as E ij , path e ij Represents node v i To node v j An optional path is established if and only if two nodes are directly connected. ij The property is defined as (T ij , Q ij , L ij ). Z ij The difficulty of the path.

[0052] Then T ij Signalized intersection delay handling:

[0053] When dealing with signal intersection delays, the present invention uses an exponential function delay formula:

[0054]

[0055] d is the effective green light time (seconds); t is the signal cycle time (seconds); λ is the green-to-signal ratio, which is the ratio of the effective green light time to the signal cycle time; c is the exit lane capacity (vehicles / hour); q is the entrance lane traffic volume (vehicles / hour);

[0056] Q ij Processing of actual vehicle speed on road section:

[0057] The actual vehicle speed is divided into two parts. One part is the road following speed. The following speed data is collected by drones, which is the displacement distance x of the front (or rear) of a vehicle within a period of time t1. The other part is the road section with no vehicles, and the maximum speed v0 allowed on the section is implemented.

[0058]

[0059] Real-time road conditions include L ij : Road surface abnormalities caused by special weather conditions U1, road construction U2, road damage that has not yet been repaired U3, and peak-hour sections U4.

[0060] L ij =α1U1+α2(U2+U3+U4)

[0061] The above four situations only need the corresponding U iIf (i=1, 2, 3, 4), the corresponding value is 1. If no other conditions occur, the value is 0, and the influencing factor α1 + α2 = 1. When special weather conditions occur, the influencing factor α1 is determined based on the weather conditions. Therefore, α1 is divided into four levels: 0.25, 0.50, 0.75, and 1.00, corresponding to the severity of the weather: mild, moderate, severe, and severe. α2 is calculated by α1 + α2 = 1.

[0062] Among them, E n e in the road network structure ij The edge set of Z ij is the path difficulty of path e(i, j); T ij is the signal intersection stop delay of the path after range normalization, Q ij is the actual speed of the vehicle, L ij is the real-time road condition; w1, w2, and w3 are the influence coefficients of the three parameters. According to the size of the influence, the three parameters w1, w2, and w3 are assigned values ​​of 0.5, 0.3, and 0.2 respectively, and w1+w2+w3=1. The formula for comprehensively evaluating the difficulty of each path based on the attributes of the path is:

[0063] Z ij =w1T ij +w2Q ij +w3L ij

[0064] The Dijkstra algorithm used in the present invention calculates the shortest distance between directed graph paths through breadth-first search, that is, the shortest path from a source point to the remaining vertices. The idea is to calculate the shortest distance between the paths of the directed graph G=(V ij , E ij， z ij ), specify the source vertex V0, and then find the shortest path from this vertex to all other nodes in the graph, generating a shortest path tree.

[0065] The Expressway Traffic Distribution Algorithm (ETD) is a combination of the actual lane capacity and the drone's route guidance.

[0066] Using multi-drone multi-segment guidance on highways, we can determine the maximum traffic volume (capacity) for a single lane and ideal traffic conditions. This involves a single standard vehicle fleet, continuously traveling at the same speed in a single lane, maintaining a minimum headway distance appropriate to the speed, and without interference from any direction. Under these ideal conditions, the established traffic flow calculation model yields the maximum traffic throughput, or basic capacity c, derived from the following formula, which in turn yields the maximum traffic volume N:

[0067]

[0068] l0 Minimum headway (m), l 车 is the average length of the vehicle (m); l 安 is the safe distance between vehicles (m) (for high-speed driving, that is, when the speed is above 100 km / h, the safe distance is above 100 meters; for fast driving, that is, when the speed is above 60 km / h, the safe distance is numerically equal to the speed); v0 is the speed of the vehicle closest to the fault warning sign in the lane where the faulty vehicle is located; L is the distance between the faulty vehicle and the vehicle closest to the fault warning sign in the lane where the faulty vehicle is located; w1 is the driver's intuitive reaction time, w2 is the driver's reaction time, and w3 is the driver's actual operation time; j max represents the vehicle momentum performance. The parameters in the formula are adjusted according to the vehicle type identified by the drone.

[0069] For the determination of actual traffic volume, refer to the correction factor f of the existing traffic composition HV .f HV The expression is as follows:

[0070]

[0071] p i is the proportion of each model (%); E i is the small amount conversion factor. HV The determination of the hard shoulder width also includes the correction factor f SW and lane width correction factor f W The correction coefficient is affected. As shown in Table 1:

[0072] Table 1

[0073]

[0074] Since this is a calculation of highway capacity, the correction coefficient for the leftmost lane uses the correction coefficient f in the table. SW =1,f W =1.

[0075] Longitudinal slope correction factor f LG:

[0076] P T is the proportion of trucks and buses; P R The proportion of trucks and tourist cars; E i E is the equivalent value of trucks and buses converted into passenger cars; R This is the equivalent value of trucks and tourist buses converted to passenger cars. Since a drone is set up every 200 meters to monitor the proportion of various vehicles on the road, its change is a dynamic process. The data needs to be uploaded uniformly by the drone group within the mutual monitoring area to a terminal to obtain the longitudinal slope correction coefficient f at that moment. LG .

[0077] The actual traffic capacity of a single lane is C 实 , multiply the maximum traffic volume by various correction factors:

[0078] C 实 =C·f SW ·f W ·f LG

[0079] f LG is the longitudinal slope correction coefficient; f SW f is the correction factor for hard shoulder width; W is the lane width correction factor. For multi-lane capacity calculation N 多 is the following formula:

[0080] N 多 =N i ΣK n

[0081] N i is the traffic capacity of the first lane (vehicles / h), that is, the traffic capacity of the leftmost lane of the expressway; K n is the reduction coefficient corresponding to the lane. (N i =2200)

[0082] For K n , optimized through the nature of the street, the frequency of vehicle entry and exit and transfer lanes, and the impact of slow-moving vehicles on both sides.

[0083] The frequency of vehicles entering, exiting, and changing lanes is detected by a swarm of drones. A drone is set up every 200 meters to hover in the middle of the highway lane, forming a highway drone swarm. Each drone detects the frequency of vehicles entering, exiting, and changing lanes every 200 meters. In the hour t1, all vehicles change lanes a total of n times, and their lane change frequency is During this period, a total of a vehicle left the road monitored by the drone group, and b vehicles entered the area monitored by the drone group, then the frequency of leaving Entry frequency

[0084]

[0085]

[0086]

[0087] f HV Represents the correction factor for traffic composition, C 实1 represents the theoretical capacity of the leftmost road of the expressway, C 实2 represents the theoretical capacity of the middle road of the highway, C 实3 Represents the theoretical capacity of the rightmost road of the highway.

[0088] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0089] (1) The present invention combines a drone with a traffic fault warning device. The drone has the characteristics of strong maneuverability and fast response speed. It can arrive at the scene at the first time, provide fast and accurate fault identification and positioning, and provide traffic management departments with more accurate and timely traffic status reports. It not only provides timely and efficient early warning to vehicles passing through the accident scene, but also effectively avoids and reduces the occurrence of secondary traffic accidents.

[0090] (2) The early warning device of the present invention is fully functional and has a fast response speed. It uses a GPS locator to quickly and accurately send the location coordinates of the current accident scene to the traffic management department, so that the traffic management department can respond to the accident in a timely manner. Multiple drones work together, with high work efficiency and high response speed.

[0091] (3) The camera in the present invention provides real-time images of the accident scene to traffic control and road rescue departments, and transmits real-time photos of the scene to the single-chip microcomputer controller for image analysis. The single-chip microcomputer controller processes and analyzes the images collected by the camera to determine the lane in which the vehicle is located.

[0092] (4) The calculation of the optimal path for urban roads takes into account a wide range, making the calculated data applicable.

[0093] (5) The present invention is based on the Dijkstra algorithm and the Expressway Traffic Distribution Algorithm, introduces the vehicle entry and exit frequency as a coefficient for correction, analyzes and processes accidents on urban roads and highways, and improves the accuracy and speed of fault identification; and automatically identifies traffic accidents, road obstacles, and traffic congestion faults, and quickly processes and analyzes them, providing detailed traffic status reports to regulatory authorities, providing accurate decision-making basis for traffic management and emergency response, and has the advantages of wide monitoring range, fast response speed, accurate early warning, and simple operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] Figure 1 This is a flow chart of the traffic fault warning method based on multi-UAV collaboration of the present invention;

[0095] Figure 2 This is a schematic structural diagram of a traffic fault warning device based on multi-UAV collaboration according to the present invention;

[0096] Figure 3 This is a schematic diagram of the top view of the traffic fault warning device based on multi-UAV collaboration of the present invention. DETAILED DESCRIPTION

[0097] like Figure 2 and Figure 3 As shown, the traffic fault warning device of the present invention includes a drone body 1, a GPS locator 2, a camera 3, a voice system 4, an infrared rangefinder 5, a single-chip microcomputer controller 6, a main warning light 7, an auxiliary warning light 8, a direction indicator light 9 and a solar panel 10; the drone body is equipped with a battery pack; the solar panel 10 supplies power to the battery pack; the camera 3 and the infrared rangefinder 5 are located below the drone body; the voice system 4 is located on the drone body; the GPS locator 2 and the single-chip microcomputer controller 6 are located inside the drone body; the main warning light 7 is located above the drone body; the auxiliary warning light 8 is located below the four arms of the drone; the direction indicator light 9 is located above the main warning light 7; and the solar panel 10 is located above the direction indicator light 9.

[0098] The solar panel 10 is located above the direction indicator light 9 and converts the photoelectric effect of solar radiation energy into electrical energy to provide energy for the entire warning system and ensure the endurance of the warning system.

[0099] The GPS tracker 2 is mounted magnetically underneath the drone fuselage 1. The GPS tracker consists of two components. The first is the positioning system, which calculates the tracker's geographic coordinates by receiving signals from satellites in the sky. The second is the communication system, which transmits these coordinates to a server or the driver's mobile phone. Once the GPS tracker 2 confirms its location, its built-in communication module transmits the data to traffic management via SMS or GPRS internet access. This data also includes the calculated optimal route, enabling traffic management to quickly and promptly reach the accident scene, improving accident resolution rates.

[0100] The single chip controller 6 collects traffic data and ground road condition data, provides data for calculating the passage of the road and the optimal path, and provides a basis for the direction indicator 9 to make judgments.

[0101] The infrared rangefinder 5 will only be turned on when the GPS positioning system determines that the vehicle is located on a highway. If a moving target appears 500 meters behind, the indicator light of the early warning system will be turned on.

[0102] When an accident occurs, the GPS locator 2 receives the signal and the geographical coordinates of the operating fault, and then transmits the information of the positioning system operating in the communication system directly to the traffic management department through the GPS locator 2. The monitoring device carried by the drone monitors the scene of the traffic accident.

[0103] When a traffic accident occurs, the drone automatically activates. If the breakdown is caused solely by a vehicle malfunction, the driver can simply press a button to start the drone. After the drone activates, the GPS locator 2 inside the drone determines the vehicle's location, providing traffic control with the precise location of the accident. If the vehicle is on an urban road, the warning device's voice system 4 is activated. Meanwhile, the infrared rangefinder is deactivated. Urban roads are densely populated and traffic is slow, and the warning lights are always on. The camera is activated to collect ground information and determine the road section. If the vehicle is on the far left side of the road, the drone's right turn signal illuminates; if it is on the far right side, the drone's left turn signal illuminates; if it is in the center of the road, both left and right turn indicators illuminate. Simultaneously, other drones collect information about surrounding road conditions. Using an algorithm, they calculate the optimal route and provide it to rescue organizations.

[0104] If the vehicle is on a highway, due to the high speeds on highways, the warning system's voice system is disabled and the infrared rangefinder is activated. When the vehicle behind the drone is 500 meters away, the warning lights activate, and the voice system is disabled. This, combined with the warning lights being activated only after the vehicle behind is 500 meters away, increases the system's endurance. Cameras collect ground information to determine road sections and traffic flow. By calculating the road's capacity, multiple drones provide guidance in different sections, implementing phased evacuation to ensure smooth traffic flow.

[0105] like Figure 1 As shown, when an accident occurs on an urban road, the traffic fault warning method based on multi-UAV collaboration of the present invention includes the following steps:

[0106] (1) Start the drone and place a warning sign 50-100 meters behind the faulty vehicle and hover the drone;

[0107] (2) The infrared rangefinder is turned off and the voice system is turned on to warn vehicles coming from behind;

[0108] (3) The camera collects road condition information. When the vehicle is on the left side of the road, the right direction running light is turned on; when the vehicle is in the middle of the road, the left and right direction running lights are turned on at the same time; when the vehicle is on the right side of the road, the left direction running light is turned on;

[0109] (4) When the faulty vehicle is operating normally, the drone automatically retracts the warning sign and returns to the faulty vehicle.

[0110] When an accident occurs on an urban road, the traffic fault warning method based on multi-UAV collaboration of the present invention includes the following steps:

[0111] (1) Activate multiple drones. After the first drone is activated, place a warning sign 50-100 meters behind the faulty vehicle.

[0112] (2) Turn off the infrared rangefinder and turn on the voice system to warn vehicles coming from behind;

[0113] (3) The camera collects road condition information. If the vehicle is on the left side of the road, the right-hand driving light is on; if the vehicle is in the middle of the road, the left and right-hand driving lights are on at the same time; if the vehicle is on the right side of the road, the left-hand driving light is on.

[0114] (4) The first drone collects road conditions and vehicle information on the accident section, and the remaining drones are responsible for obtaining data on the entire traffic area;

[0115] (5) After the drone completes data collection, it transmits the data to the microcontroller for T ij Signalized intersection delay on the path, Q ij Actual speed of vehicles on the road section, L ijThe calculation formula for the three data of real-time road conditions is as follows:

[0116]

[0117]

[0118] L ij =α1U1+α2(U2+U3+U4) (2)

[0119] In the above formula, d is the green light time (seconds); t is the signal cycle time (seconds); λ is the green-to-signal ratio, that is, the ratio of the effective green light time to the signal cycle time; c is the traffic capacity of the exit road (vehicles / hour); q is the traffic volume of the entrance road (vehicles / hour); t1 is the driving time of a vehicle; x is the displacement distance of the front (or rear) of the vehicle within time t1; v0 is the maximum speed allowed on the road section when there are no vehicles on the road; U i (i = 1, 2, 3, 4) represent four road conditions, with a value of 1 if present and 0 if absent. The influencing factor α1 + α2 = 1. α1 is the special weather influencing factor, which is divided into four levels: 0.25, 0.50, 0.75, and 1.00, corresponding to the severity of the weather: mild, moderate, severe, and severe. α2 is calculated using α1 + α2 = 1. The three parameters w1, w2, and w3 are 0.5, 0.3, and 0.2, respectively.

[0120] (6) Substitute the calculated results of the three data into Z ij =w1T ij +w2Q ij +w3L ij Calculate the path difficulty of each path;

[0121] (7) Z ij Substitute Dijkstra to calculate the optimal path, and send the optimal path to the accident scene to the rescue organization so that the rescue organization can reach the accident scene.

[0122] When an accident occurs on a highway, the traffic fault warning method based on multi-UAV collaboration of the present invention includes the following steps:

[0123] (1) All drones take off and hover in the middle of the road every 200 meters behind the accident scene to warn the following vehicles. At the same time, the infrared rangefinder of the first drone is turned on. When the rear vehicle is 500 meters away from the first drone, the warning light turns on. Each drone detects the frequency of vehicles entering and exiting and changing lanes within 200 meters of itself.

[0124] (2) Within t1 hour, all vehicles change lanes a total of n times, and their lane-changing frequency During this period, a total of a vehicle left the road monitored by the drone group, and b vehicles entered the area monitored by the drone group, then the frequency of leaving Entry frequency

[0125] (3) The obtained data is uploaded to the single chip computer and the following formula is used according to the Expressway Traffic Distribution Algorithm model:

[0126] The actual traffic capacity of the road requires parameter correction of the theoretical traffic capacity, so the traffic composition correction coefficient f can be calculated according to the following formula: HV , slope correction coefficient f LG :

[0127]

[0128]

[0129] The theoretical multi-lane capacity calculation formula is as follows:

[0130] N 多 =N i ·∑K n

[0131] However, the actual traffic capacity of the road needs to take into account the frequency of lane changes of vehicles.

[0132] C 实 =C·f SW ·f W ·f LG

[0133] Since the leftmost lane of the expressway has the highest design speed (120km / h), SW =1

[0134] f W =1

[0135] This leads to the following three formulas:

[0136]

[0137]

[0138]

[0139] The maximum traffic throughput obtained by the Expressway Traffic Distribution Algorithm model is calculated, that is, the basic traffic capacity of the road section where the traffic accident occurred.HV Represents the correction factor for traffic composition, C 实1 represents the actual traffic capacity of the leftmost road of the expressway, C 实2 represents the actual traffic capacity of the middle road of the highway, C 实3 Represents the actual capacity of the rightmost road of the highway.

[0140] (4) Divert the rear traffic flow based on the calculation results to avoid secondary accidents caused by vehicles coming from behind due to insufficient reaction or lack of attention, thereby reducing or even avoiding highway congestion.

[0141] Where p i is the proportion of each model (%); E i is the small amount conversion factor; P T is the proportion of trucks and buses; P R The proportion of trucks and tourist cars; E i E is the equivalent value of trucks and buses converted into passenger cars; R is the equivalent value of trucks and tourist buses converted into passenger cars; f LG is the longitudinal slope correction coefficient; f SW f is the correction factor for hard shoulder width; W is the lane width correction factor; N i is the traffic capacity of the first lane (vehicles / h), that is, the traffic capacity of the leftmost lane of the expressway; K n is the reduction coefficient corresponding to the lane (N i =2200); t1 is a period of time; n is the total number of lane changes by all vehicles; a is the number of vehicles leaving the road monitored by the drone swarm; b is the number of vehicles entering the area monitored by the drone swarm.

Claims

1. A traffic fault warning method based on multi-UAV collaboration, implemented by a traffic fault warning device based on UAV collaboration, characterized by: The warning device comprises a fuselage (1), a body, an arm, a GPS locator (2), a camera (3), a voice system (4), an infrared rangefinder (5), a single-chip controller (6), a main warning light (7), a secondary warning light (8), a direction indicator light (9) and a battery panel (10); The camera (3) and the infrared rangefinder (5) are located below the machine body, and the voice system (4) is located on the machine body; the GPS locator (2) and the single-chip microcomputer controller (6) are located inside the machine body, the main warning light (7) is located above the machine body, the auxiliary warning light (8) is located below the machine arm, and the direction indicator light (9) is located above the main warning light (7); The method comprises the following steps: (1) When an accident occurs on a city road, multiple drones are activated. After the first drone is launched, a warning sign is placed behind the faulty vehicle. (2) Turn off the infrared rangefinder and turn on the voice system to warn vehicles coming from behind; (3) The camera collects road condition information. If the vehicle is on the left side of the road, the right running light is on; if the vehicle is in the middle of the road, the left and right running lights are on at the same time; if the vehicle is on the right side of the road, the left running light is on; (4) The first drone collects road conditions and vehicle information on the accident section, and the remaining drones obtain data for the entire traffic area; (5) After the drone completes data collection, it transmits the data to the microcontroller to calculate the signal intersection delay T of the path. ij , Actual speed of vehicles on the road section Q ij 、Real-time road conditions ij Calculation of three data; (6) Substitute the calculated results of the three data into the path difficulty Z ij =w1T ij +w2Q ij +w3L ij Calculate the path difficulty of each path; (7) Z ij Substitute the Dijkstra algorithm to calculate the optimal path, and send the optimal path to the accident site to the rescue organization.

2. The traffic fault warning method based on multi-UAV collaboration according to claim 1 is characterized by: The calculation formula of step (5) is as follows: L ij =α1U1+α2(U2+U3+U4) (2) Where d is the green light time; t is the signal cycle duration; c is the traffic volume at the exit per hour; q is the traffic volume at the entrance per hour; t1 is the driving time of a vehicle; x is the displacement of the front or rear of the vehicle within time t1; v0 is the maximum speed allowed on the road section when there are no vehicles on the road; U1 is the road surface abnormality caused by special weather, U2 is the road surface construction, U3 is the road section with damaged road surface that has not yet been repaired, U4 is the peak traffic section, α1 is the weather impact factor, and α2 is the road surface condition factor.

3. The traffic fault warning method based on multi-UAV collaboration according to claim 2 is characterized by: In step (5), α1+α2=1.

4. The traffic fault warning method based on multi-UAV collaboration according to claim 1 is characterized by: In step (1), after the first drone is started, a warning sign is placed 50 to 100 meters behind the faulty vehicle.

5. The traffic fault warning method based on multi-UAV collaboration according to claim 1 is characterized by: In step (1), the UAV obtains the road network plan and establishes a labeled mathematical model to represent the planned path with a road network structure diagram: G=(V ij ,E ij ,With ij ) where node u ij The set of V ij , E ij is the weight set path set, Z ij The difficulty of the path.

6. The traffic fault warning method based on multi-UAV collaboration according to claim 1 is characterized by: The following steps are involved: (1) When a traffic accident occurs on a highway, a swarm of drones takes off and hovers at intervals behind the accident scene. The infrared rangefinder of the first drone is turned on. When the vehicle behind is several meters away from the first drone, the warning light turns on. Each drone detects the frequency of vehicles entering and exiting and changing lanes in its own lane. (2) In the hour t1, suppose all vehicles change lanes n times, and the lane-changing frequency During this period, a total of a vehicle left the road monitored by the drone group, and b vehicles entered the area monitored by the drone group, then the frequency of leaving Entry frequency (3) Upload the data from step (2) to the microcontroller and calculate the traffic composition correction coefficient f according to the following formula HV and slope correction factor f LG : The theoretical multi-lane capacity calculation formula is as follows: N 多 =N i ·∑K n The leftmost lane of the expressway is designed for a speed of 120 km / h. f SW ＝1 f W ＝1 This leads to the following three formulas: The maximum traffic volume is calculated, where f HV Represents the correction factor for traffic composition, C 实1 represents the actual traffic capacity of the leftmost road of the expressway, C 实2 represents the actual traffic capacity of the middle road of the highway, C 实3 Represents the actual traffic capacity of the rightmost road of the highway; Where p i is the proportion of each model; E i is the small amount conversion factor; P T is the proportion of trucks and buses; P R The proportion of trucks and tourist cars; E i E is the equivalent value of trucks and buses converted into passenger cars; R is the equivalent value of trucks and tourist buses converted into passenger cars; f LG is the longitudinal slope correction coefficient; f SW f is the correction factor for hard shoulder width; W is the lane width correction factor; N i K is the capacity of the leftmost lane of the highway; n is the reduction coefficient corresponding to the lane; t1 is the time; n is the total number of lane changes by all vehicles; a is the number of vehicles leaving the road monitored by the drone swarm; b is the number of vehicles entering the area monitored by the drone swarm; Maximum traffic volume Among them, l0 is the minimum headway, l 车 is the average length of the vehicle; l 安 is the safe distance between vehicles; v0 is the speed of the vehicle closest to the fault warning sign in the lane where the faulty vehicle is located; L is the distance between the faulty vehicle and the vehicle closest to the fault warning sign in the lane where the faulty vehicle is located; w1 is the driver's intuitive reaction time, w2 is the driver's reaction time, and w3 is the driver's actual operation time; j max Indicates the car's momentum performance; (4) Divert the vehicles behind based on the calculation results.

Images