Unmanned aerial vehicle detection tracking adaptive low-altitude economic road picketing and logistics method

Abstract

The invention discloses an unmanned aerial vehicle detection and tracking adaptive low-altitude economic road picketing and logistics method, and relates to the technical field of unmanned aerial vehicles. In the aspect of road picketing, road vehicle videos are collected through a combined sensor system carried by an unmanned aerial vehicle, and a driver behavior danger coefficient and a driving danger coefficient are calculated through information processing; vehicle danger levels are evaluated based on the two and corresponding management and control measures are taken, such as conventional monitoring, key tracking, emergency early warning and the like of different levels. In the aspect of logistics application, order receiving and processing, path planning and unmanned aerial vehicle scheduling, flight distribution and goods signing and returning are included, and specific operation and judgment methods of all links of logistics are covered.

Description

Technical Field

[0001] The present invention relates to the technical field of drones, and in particular to a drone detection and tracking method adapted for low-altitude economic road patrol and logistics. Background Art

[0002] With the rapid development of the low-altitude economy, orderly management and efficient operations in this sector have become key challenges. As the global low-altitude economy continues to expand, demand for drone applications in urban road patrol and logistics has surged. To promote the large-scale application of drones in these areas, facilitate the implementation of the "smart transportation" and "unmanned delivery" industries, and accelerate the transition of the low-altitude economy from "pilot" to "regularized operation," it is crucial to develop a drone detection and tracking method suitable for low-altitude economic road patrol and logistics.

[0003] Existing technology, such as the invention patent application CN119068369A, discloses a drone detection and tracking method suitable for low-altitude economic road patrols and logistics. The method includes the following steps: activating the drone and image processing equipment, loading a pre-trained deep learning model to a processor or cloud, and setting the flight path, altitude, and monitoring area. Using the DJI Mobile SDK, the radar vertical coordinates are converted to TapFlyMission target coordinates, and the drone is controlled to automatically plan a route, avoid obstacles, and fly to the target point. By tracking offending entities in real time, the false positive rate is significantly reduced, and offending vehicles and pedestrians are accurately captured. Violation results are uploaded to a backend database in real time, providing law enforcement officers with intuitive access to specific data and effectively reducing the enforcement burden.

[0004] The following defects exist in the existing technology, which are specifically reflected in: 1. In the existing technology, existing road hazard detection mostly relies on a single device, cannot synchronously monitor the driver's status, and the risk assessment is one-sided. The detection accuracy drops significantly at night or in bad weather. It is mostly a "find the problem-punish afterwards" mode, lacks real-time dynamic management and control capabilities, cannot intervene in advance on high-risk vehicles, and has a single early warning method, which lags in response to emergencies.

[0005] 2. In existing technologies, drone logistics screening mostly relies on manual experience or a single weight indicator, and lacks quantitative assessment of the compatibility of cargo volume and cargo compartment. This may cause the drone's load balance to be unbalanced or the cargo compartment to be unable to close, posing a dangerous hazard. It is unable to respond to sudden weather changes or temporary airspace control in real time, causing the drone to be forced to return or be stranded, making it difficult to guarantee delivery timeliness and risk reduction. The drone's emergency mechanism is mostly "one-size-fits-all" and does not classify the severity of the fault. This may result in wasted transportation capacity for minor faults or insufficient response to severe faults. Summary of the Invention

[0006] The purpose of the present invention is to provide a drone detection and tracking method adapted to low-altitude economic road patrol and logistics, which solves the problems existing in the background technology.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions: The present invention provides a method for detecting and tracking drones adapted to low-altitude economic road patrols, including: step one, information acquisition, for using a combined sensor system carried by a drone to collect videos of each vehicle passing through the target road, the combined sensor system including a high-definition visible light camera and an infrared thermal imager.

[0008] Step 2: Information processing: Based on the video of each vehicle on the target road, the behavior risk coefficient of each vehicle driver is calculated, and the driving risk coefficient of each vehicle driver is calculated.

[0009] Step 3: Risk assessment and early warning: Based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, the risk level of each vehicle is assessed, and then control measures are taken based on the risk level of each vehicle.

[0010] Preferably, the behavioral risk coefficient of each vehicle driver is obtained by calculation, and a specific implementation method is: analyzing the video of each vehicle on the target road, detecting and identifying various distracting behaviors and various fatigue driving behaviors of each vehicle driver, summarizing and obtaining the duration and frequency of various distracting behaviors of each vehicle driver, and obtaining the duration and frequency of various fatigue driving behaviors of each vehicle driver.

[0011] The duration and frequency of each type of distracting behavior of each vehicle driver are compared with the duration and frequency of appropriate distracting behaviors stored in the database, and the distracting behavior risk coefficient of each vehicle driver is calculated.

[0012] The duration and frequency of each type of fatigue driving behavior of each vehicle driver are compared with the duration and frequency of appropriate fatigue driving behaviors stored in the database, and the fatigue driving behavior risk coefficient of each vehicle driver is calculated.

[0013] Based on the distraction behavior risk coefficient and fatigue driving behavior risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the behavior risk coefficient of each vehicle driver.

[0014] Preferably, the driving risk coefficient of each vehicle driver is obtained by calculation, and a specific implementation method is as follows: from the video of each vehicle on the target road, the speed of each vehicle on the target road at each monitoring time point is extracted, and the speed is compared with the maximum speed threshold of the target road stored in the database; if the speed of a vehicle passing through the target road at a certain monitoring time point is greater than the maximum speed threshold of the target road stored in the database, it is determined that the vehicle passing through the target road has speeding behavior, and then the speed risk coefficient of each vehicle is calculated based on the speed of each vehicle passing through the target road at each monitoring time point and the maximum speed threshold of the target road stored in the database.

[0015] The video of each vehicle on the target road is analyzed, and the various types of violations of each vehicle are detected and analyzed. The duration and frequency of each type of violation of each vehicle are statistically obtained, and the duration and frequency of the violations are compared with the duration and frequency of appropriate violations stored in the database to calculate the violation risk factor of each vehicle.

[0016] Based on the speed risk coefficient and violation risk coefficient of each vehicle, the driving risk coefficient of each vehicle driver is obtained by weighted calculation.

[0017] Preferably, the assessment obtains the danger level of each vehicle, and the specific implementation method is: based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the comprehensive risk coefficient of each vehicle, and the comprehensive risk coefficient of each vehicle is compared with the comprehensive risk coefficient range of vehicles corresponding to each danger level stored in the database. If the comprehensive risk coefficient of a vehicle is within the comprehensive risk coefficient range of vehicles corresponding to a certain danger level stored in the database, then the danger level is used as the danger level of the vehicle.

[0018] Preferably, the control measures are taken based on the danger level of each vehicle, and the specific implementation method is: extract the danger level of each vehicle. If the danger level of a vehicle is level one, the drone automatically adjusts its flight attitude and position, focuses on tracking the vehicle, and transmits relevant information to the ground command center in real time. The ground command center can remind the vehicle driver of the danger through text messages or voice prompts. If the danger level of a vehicle is level two, the drone immediately sends an emergency warning to the ground command center, and at the same time turns on the sound and light warning device to warn the vehicle in the air. At the same time, the ground command center dispatches law enforcement vehicles to intercept and inspect the vehicle.

[0019] A method for drone detection and tracking adapted to low-altitude economic road logistics includes: Step 1, order reception and processing, receiving a user's logistics order, screening the goods in the order, and determining whether the goods are suitable for drone delivery. If so, a delivery task is generated, wherein the delivery task includes the starting point, end point, and timeliness requirements of the goods.

[0020] Step 2: Path planning and drone scheduling: extract real-time weather information, airspace restriction information, and drone endurance, plan the optimal flight path for the delivery mission, dispatch available drones from storage nodes, load the cargo onto the drones, and calibrate the load balance.

[0021] Step 3: Flight delivery: Control the drone to fly according to the planned flight path. During the flight, the drone will perform obstacle avoidance operations in real time through the sensors it carries. If an emergency occurs, the corresponding emergency mechanism will be triggered.

[0022] Step 4: Cargo receipt and return. When the drone arrives at the destination, it adjusts its delivery posture in real time at the designated take-off and landing point and automatically lands. The user signs for the cargo, and the drone chooses to return to recharge or go to the next mission point based on its own power status.

[0023] Preferably, the determination of whether the goods are suitable for drone delivery is specifically implemented as follows: obtaining the weight and volume of the goods in the order, and comparing them with the load capacity and cargo hold volume of the dispatchable drone; if the weight of the goods in the order is higher than the load capacity of the dispatchable drone or the volume of the goods in the order is higher than the cargo hold volume of the dispatchable drone, then the goods are determined to be unsuitable for drone delivery; if the weight of the goods in the order is lower than the load capacity of the dispatchable drone and the volume of the goods in the order is lower than the cargo hold volume of the dispatchable drone, then the goods are determined to be suitable for drone delivery.

[0024] Preferably, the optimal flight path for the delivery task is planned, and the specific implementation method is as follows: obtaining real-time weather information from the meteorological department, including wind speed, wind direction, temperature, and precipitation probability, obtaining airspace restriction information from the airspace management department, and obtaining the drone's endurance from the database, including the maximum flight distance and remaining power, performing preprocessing, and constructing a map containing the starting point, end point, and surrounding environment information. Through a path search algorithm, various possible paths from the starting point to the end point are obtained, and then the safety factor of each possible path is calculated. The possible path corresponding to the maximum safety factor is screened and used as the optimal flight path.

[0025] Preferably, the triggering of the corresponding emergency mechanism is specifically implemented as follows: by equipping the drone with multiple sensors, the operating data of the drone is monitored in real time, and based on the operating data of the drone, the flight safety factor of the drone is calculated, and the flight safety factor of the drone is compared with the flight safety factor range of the drone corresponding to each safety level stored in the database. If the flight safety factor of the drone is within the flight safety factor range of the drone corresponding to a certain safety level stored in the database, then the safety level is used as the safety level of the drone.

[0026] If the drone's safety level is level one, the drone will continue to fly along the optimal flight path. If the drone's safety level is level two, it will send an emergency fault message to the ground control center, requesting the operator's guidance, while maintaining flight balance and reducing flight speed. If the drone's safety level is level three, the emergency landing procedure will be immediately initiated, and a safe landing point will be selected based on the current position and surrounding environment. At the same time, an alarm will be issued to the surrounding area through sound and light signals and wireless communication equipment. After receiving the serious fault information, the ground control center will immediately activate the emergency rescue plan and organize personnel to go to the landing point for rescue.

[0027] Preferably, the delivery posture is adjusted in real time at the designated take-off and landing point, and the specific implementation method is: extract the pitch angle, roll angle, and yaw angle of the drone, combine them with the current wind speed, calculate the delivery safety factor of the drone, and compare it with the delivery safety factor threshold stored in the database. If the delivery safety factor of the drone is higher than the delivery safety factor threshold stored in the database, it is determined that the drone can be delivered. If the delivery safety factor of the drone is lower than the delivery safety factor threshold stored in the database, it is determined that the drone cannot be delivered, and the pitch angle, roll angle, and yaw angle of the drone are automatically changed until the delivery safety factor of the drone is higher than the safety risk factor threshold stored in the database.

[0028] The beneficial effects of the present invention are: 1. In the present invention, multi-dimensional detection of driver behavior and driving behavior is achieved through combined sensors, and a dual evaluation system of "behavior risk coefficient + driving risk coefficient" is constructed. At the same time, the detection accuracy at night or in bad weather is guaranteed, and vehicles are divided into three levels of danger based on the comprehensive risk coefficient, matching differentiated management and control measures to achieve a transition from "post-interception" to "real-time intervention".

[0029] 2. In the present invention, standardized screening rules are established through clear comparison of weight and volume dual thresholds to ensure that drone loads are dangerous and avoid flight accidents or delivery failures caused by overloading or space mismatch. Real-time meteorological data, airspace restrictions and drone endurance are integrated, and the optimal path is selected through a risk factor model. The drone operation data is monitored in real time through sensors, the flight risk factor is calculated and a graded response is given. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0031] Figure 1Schematic diagram of the process flow for implementing the road patrol method of the present invention.

[0032] Figure 2 The figure is a schematic flow chart of the steps for implementing the logistics method of the present invention. DETAILED DESCRIPTION

[0033] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0034] Reference Figure 1 As shown, the present invention provides a method for detecting and tracking drones adapted to patrol low-altitude economic roads, including: step one, information acquisition, for collecting videos of vehicles passing through the target road using a combined sensor system carried by the drone, wherein the combined sensor system includes a high-definition visible light camera and an infrared thermal imager.

[0035] Step 2: Information processing: Based on the video of each vehicle on the target road, the behavior risk coefficient of each vehicle driver is calculated, and the driving risk coefficient of each vehicle driver is calculated.

[0036] In a specific embodiment, the calculation obtains the behavioral risk coefficient of each vehicle driver, and the specific implementation method is: analyzing the video of each vehicle on the target road, detecting and identifying various distracting behaviors and various fatigue driving behaviors of each vehicle driver, summarizing and obtaining the duration and frequency of various distracting behaviors of each vehicle driver, and obtaining the duration and frequency of various fatigue driving behaviors of each vehicle driver.

[0037] It should be noted that the distracting behaviors include but are not limited to smoking and eating, and the fatigue driving behaviors include but are not limited to yawning and rubbing eyes; the detection and identification of various distracting behaviors and fatigue driving behaviors of each vehicle driver is an existing technology, which specifically includes but is not limited to line of sight deviation detection, hand movement detection, head posture detection, fatigue feature detection, yawn detection, and head droop detection.

[0038] The duration and frequency of each type of distracting behavior of each vehicle driver are compared with the duration and frequency of appropriate distracting behaviors stored in the database, and the distracting behavior risk coefficient of each vehicle driver is calculated.

[0039] It should be noted that the calculation formula for obtaining the distraction risk coefficient of each vehicle driver is: ,in Indicates the The risk factor of distracted behavior of each vehicle driver, Indicates the Vehicle driver No. The duration of distraction-like behavior, Indicates the Vehicle driver No. Frequency of distracting behaviors, 、 The duration and frequency of appropriate distracting behaviors stored in the database are respectively represented by the following method: tens of thousands of hours of safe driving videos are collected through the vehicle-mounted DMS (Driver Monitoring System), and the duration and frequency of various types of driver distracting behaviors in accident-free / hazard-free scenarios are marked. The average processing is performed to obtain the duration and frequency of various types of distracting behaviors after the average processing, and these are used as the duration and frequency of appropriate distracting behaviors stored in the database. Indicates the number of each vehicle, , is a positive integer greater than 2, Numbers representing various types of distracting behaviors, , is a positive integer greater than 2, 、 The formulas represent the appropriate weighting factors for distraction duration and frequency, respectively, stored in the database and set by professionals. Firstly, this formula sets appropriate thresholds based on actual safe driving data, ensuring that risk factor calculations are closer to real-world driving scenarios and improving assessment accuracy. Secondly, it integrates duration, frequency, and weighting factors for multi-dimensional quantification, scientifically reflecting the comprehensive risk of distraction and providing a precise basis for safety management.

[0040] The duration and frequency of each type of fatigue driving behavior of each vehicle driver are compared with the duration and frequency of appropriate fatigue driving behaviors stored in the database, and the fatigue driving behavior risk coefficient of each vehicle driver is calculated.

[0041] It should be noted that the calculation formula for obtaining the fatigue driving behavior risk coefficient of each vehicle driver is: ,in Indicates the The risk factor of fatigue driving behavior of each vehicle driver, Indicates the Vehicle driver No. Duration of fatigue-like driving behavior, Indicates the Vehicle driver No. The frequency of fatigue-like driving behaviors, 、 They respectively represent the duration and frequency of suitable fatigue driving behaviors stored in the database, and the setting method is as follows: extracting the duration and frequency of various fatigue driving behaviors of all vehicle drivers on the target road within a preset time period, screening out the minimum value of the duration and frequency of various fatigue driving behaviors, and using it as the duration and frequency of suitable fatigue driving behaviors stored in the database; the vehicle is driven by one person during driving, and the vehicle number is the driver number; Numbers representing various types of distracting behaviors, , is a positive integer greater than 2, 、 The weighting factors for the appropriate duration and frequency of distracted driving, respectively, are stored in the database and are set by professionals. The advantages of this formula are: first, it uses the minimum duration and frequency of fatigue driving behavior within a preset time period as a safety benchmark, strengthening the bottom-line approach to risk management through strict threshold setting; second, it quantitatively calculates the duration, frequency, and professionally set weighting factors to scientifically reflect the overall risk level of fatigue driving behavior, providing a basis for graded warnings and precise intervention.

[0042] Based on the distraction behavior risk coefficient and fatigue driving behavior risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the behavior risk coefficient of each vehicle driver.

[0043] It should be noted that the behavioral risk coefficient of each vehicle driver is calculated using the following formula: ,in Indicates the The behavioral risk factor of each vehicle driver is 、 The weight factors represent the appropriate distraction behavior weight factor and fatigue driving behavior weight factor stored in the database, respectively. The setting method is as follows: the number of accidents caused by distraction behavior on the target road is screened from the database, compared with the total number of accidents on the target road in the database, and the percentage of accidents caused by distraction behavior is calculated. This percentage is used as the appropriate distraction behavior weight factor stored in the database. The number of accidents caused by fatigue driving behavior on the target road is screened from the database, compared with the total number of accidents on the target road in the database, and the percentage of accidents caused by fatigue driving is calculated. This percentage is used as the appropriate fatigue driving behavior weight factor stored in the database. The advantages of this formula are: first, the weight factor is dynamically set based on accident data, objectively reflecting the actual impact of distraction and fatigue driving behaviors on the safety of the target road; second, the risk weight is quantified by the accident percentage, making the behavior risk factor calculation more closely aligned with the causes of road accidents, providing a scientific basis for accurately identifying high-risk driving behaviors.

[0044] In a specific embodiment, the driving risk coefficient of each vehicle driver is calculated, and the specific implementation method is as follows: from the video of each vehicle on the target road, the speed of each vehicle on the target road at each monitoring time point is extracted, and the speed is compared with the maximum speed threshold of the target road stored in the database. If the speed of a vehicle passing through the target road at a certain monitoring time point is greater than the maximum speed threshold of the target road stored in the database, it is determined that the vehicle passing through the target road is speeding. Then, based on the speed of each vehicle passing through the target road at each monitoring time point and the maximum speed threshold of the target road stored in the database, the speed risk coefficient of each vehicle is calculated.

[0045] It should be noted that the speed risk coefficient of each vehicle is obtained by the calculation formula: ,in Indicates the The speed hazard factor of each vehicle, Indicates the The speed of the vehicle, represents the maximum speed threshold of the target road stored in the database, and e represents a natural constant.

[0046] The video of each vehicle on the target road is analyzed, and the various types of violations of each vehicle are detected and analyzed. The duration and frequency of each type of violation of each vehicle are statistically obtained, and the duration and frequency of the violations are compared with the duration and frequency of appropriate violations stored in the database to calculate the violation risk factor of each vehicle.

[0047] It should be noted that the various types of violations include but are not limited to crossing a solid line and changing lanes on a solid line. The risk factor of each vehicle's violation is calculated using the following formula: ,in Indicates the The risk factor of the violation of each vehicle, 、 Respectively represent Vehicle driver No. The duration and frequency of such violations, 、 They represent the duration and frequency of appropriate violations stored in the database. Considering the occurrence of accidents, the duration and frequency of appropriate violations are set to 0.5 and 1 respectively. Numbers representing various types of distracting behaviors, , is a positive integer greater than 2. The advantages of this formula are: first, it uses extremely low appropriate duration (0.5) and frequency (1) as safety benchmarks to strictly define the risk bottom line of violations and strengthen the zero-tolerance assessment of violations such as crossing the solid line and changing lanes on the solid line; second, through the quantitative calculation of duration and frequency, it scientifically reflects the comprehensive risk level of violations, providing an intuitive basis for real-time monitoring and accurate early warning.

[0048] Based on the speed risk coefficient and violation risk coefficient of each vehicle, the driving risk coefficient of each vehicle driver is obtained by weighted calculation.

[0049] It should be noted that the driving risk coefficient of each vehicle driver is calculated using the following formula: ,in Indicates the The driving risk factor of each vehicle driver, 、 They represent the appropriate speed hazard weight factor and the appropriate violation hazard weight factor stored in the database, respectively. The setting method is as follows: the number of accidents caused by speed hazards on the target road is screened from the database, and this number is compared with the total number of accidents on the target road in the database to obtain the proportion of accidents caused by speed hazards, which is used as the appropriate speed hazard weight factor stored in the database. The number of accidents caused by violations on the target road is screened from the database, and this number is compared with the total number of accidents on the target road in the database to obtain the proportion of accidents caused by violations, which is used as the appropriate violation weight factor stored in the database. The advantages of this formula are: first, the weight factor is dynamically determined based on the accident data of the target road, objectively quantifying the actual contribution of speed hazards and violations to accident risks; second, by setting weights based on the accident proportion, the driving hazard coefficient calculation closely matches the causes of road accidents, providing scientific support for the accurate assessment of driver risk levels and targeted safety management.

[0050] Step 3: Risk assessment and early warning: Based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, the risk level of each vehicle is assessed, and then control measures are taken based on the risk level of each vehicle.

[0051] In a specific embodiment, the assessment obtains the danger level of each vehicle, and the specific implementation method is: based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the comprehensive risk coefficient of each vehicle, and the comprehensive risk coefficient of each vehicle is compared with the comprehensive risk coefficient range of vehicles corresponding to each danger level stored in the database. If the comprehensive risk coefficient of a vehicle is within the comprehensive risk coefficient range of vehicles corresponding to a certain danger level stored in the database, then the danger level is used as the danger level of the vehicle.

[0052] It should be noted that the above calculations obtain the comprehensive risk factor of each vehicle, and the calculation formula is: ,in Indicates the The comprehensive risk factor of each vehicle driver, 、 They respectively represent the appropriate behavioral risk weight factor and the appropriate driving risk weight factor stored in the database. The setting method is the same as the setting method of the appropriate speed risk weight factor and the appropriate violation risk weight factor, and will not be elaborated here. The database stores the comprehensive risk coefficient range of the vehicle corresponding to each risk level. The specific setting method is to analyze historical accident data, calculate the accident rate in different coefficient intervals, and use the critical point where the accident rate increases significantly as the level division boundary. The advantages of this formula are: First, the weight factor is dynamically set according to the target road accident data, and the actual impact of behavioral risk and driving risk on accident risk is objectively quantified; second, the risk level boundaries are divided based on historical accident rates, and the risk intervals are scientifically defined to provide data support for accurate graded assessment and differentiated management of vehicle risk levels. In a specific embodiment, the control measures are taken based on the danger level of each vehicle. The specific implementation method is: extract the danger level of each vehicle. If the danger level of a vehicle is level one, the drone automatically adjusts its flight attitude and position, focuses on tracking the vehicle, and transmits relevant information to the ground command center in real time. The ground command center can remind the vehicle driver of the danger through text messages or voice prompts. If the danger level of a vehicle is level two, the drone immediately sends an emergency warning to the ground command center, and at the same time turns on the sound and light warning device to warn the vehicle in the air. At the same time, the ground command center dispatches law enforcement vehicles to intercept and inspect the vehicle.

[0053] It should be noted that the higher the hazard level, the more dangerous it is.

[0054] In the present invention, multi-dimensional detection of driver behavior and driving behavior is achieved through combined sensors, and a dual evaluation system of "behavior risk coefficient + driving risk coefficient" is constructed. At the same time, the detection accuracy at night or in bad weather is guaranteed. Based on the comprehensive risk coefficient, vehicles are divided into three levels of risk, and differentiated management and control measures are matched to achieve the transition from "post-interception" to "real-time intervention".

[0055] Reference Figure 2 As shown, the present invention provides a method for detecting and tracking drones to adapt to low-altitude economic logistics, including: a method for detecting and tracking drones to adapt to low-altitude economic road logistics, including: step one, order reception and processing, receiving the user's logistics order, screening the goods in the order, and judging whether the goods are suitable for drone delivery. If suitable, a delivery task is generated, and the delivery task includes the starting point, end point and time requirements of the goods.

[0056] In a specific embodiment, the determination of whether the goods are suitable for drone delivery is specifically implemented as follows: obtaining the weight and volume of the goods in the order, and comparing them with the load capacity and cargo hold volume of the dispatchable drone; if the weight of the goods in the order is higher than the load capacity of the dispatchable drone or the volume of the goods in the order is higher than the cargo hold volume of the dispatchable drone, then the goods are determined to be unsuitable for drone delivery; if the weight of the goods in the order is lower than the load capacity of the dispatchable drone and the volume of the goods in the order is lower than the cargo hold volume of the dispatchable drone, then the goods are determined to be suitable for drone delivery.

[0057] It should be noted that the load capacity and cargo hold volume of the dispatchable drone are stored in the database and obtained through the drone's operating instructions.

[0058] Step 2: Path planning and drone scheduling: extract real-time weather information, airspace restriction information, and drone endurance, plan the optimal flight path for the delivery mission, dispatch available drones from storage nodes, load the cargo onto the drones, and calibrate the load balance.

[0059] It should be noted that loading cargo onto the drone and calibrating the load balance involves first detecting the cargo weight using a cargo hold pressure sensor, then determining the center of gravity coordinates through laser scanning or geometric calculations. This is then combined with the drone's empty center of gravity data to verify whether it is within a preset danger zone. If the center of gravity shifts, the counterweight position is adjusted using an adjustable counterweight on the fuselage or a dynamic balance bracket. Meanwhile, an IMU sensor is used to monitor the pitch and roll angles in real time to ensure that the drone's attitude deviation after loading is within the danger zone. Finally, a vibration sensor is used to verify the cargo's securement and overall balance to ensure flight stability.

[0060] In a specific embodiment, the optimal flight path for the delivery task is planned by the following specific implementation method: obtaining real-time weather information from the meteorological department, including wind speed, wind direction, temperature, and precipitation probability; obtaining airspace restriction information from the airspace management department; obtaining the drone's endurance from the database, including the maximum flight distance and remaining power; performing preprocessing; constructing a map containing the starting point, end point, and surrounding environment information; obtaining each possible path from the starting point to the end point through a path search algorithm; and then calculating the safety factor of each possible path. The possible path corresponding to the maximum safety factor is screened and used as the optimal flight path.

[0061] It should be noted that the construction of a map including the starting point, the end point and the surrounding environment information is based on a path search algorithm, which is a prior art and will not be described in detail here. The calculation formula for obtaining the risk coefficient of each possible path is: ,in Indicates the The risk factor of a possible path, 、 、 Respectively represent Wind speed, precipitation probability, and number of no-fly zones for each possible path, 、 、 They represent the appropriate wind speed, appropriate precipitation probability, and appropriate number of no-fly zones stored in the database, respectively. The setting method is: obtain the wind speed, precipitation probability, and number of no-fly zones of all possible paths, perform average processing, and use the averaged wind speed, precipitation probability, and number of no-fly zones as the appropriate wind speed, precipitation probability, and number of no-fly zones stored in the database. 、 、 These factors represent the appropriate wind speed weighting factor, the appropriate precipitation probability weighting factor, and the appropriate no-fly zone number weighting factor stored in the database, respectively, and are set by professionals. The advantages of this formula are: first, it uses the mean of all possible path data as an appropriate benchmark, objectively reflecting the safety threshold of the overall environment and providing a universal reference for risk factor calculations; second, it combines professionally set weighting factors to quantify the combined impact of wind speed, precipitation probability, and the number of no-fly zones, scientifically assessing the path risk level and providing a basis for path planning and safety decision-making.

[0062] Step 3: Flight delivery: Control the drone to fly according to the planned flight path. During the flight, the drone will perform obstacle avoidance operations in real time through the sensors it carries. If an emergency occurs, the corresponding emergency mechanism will be triggered.

[0063] In a specific embodiment, the triggering of the corresponding emergency mechanism is specifically implemented as follows: by equipping the drone with multiple sensors, the operating data of the drone is monitored in real time, and based on the operating data of the drone, the flight safety factor of the drone is calculated, and the flight safety factor of the drone is compared with the flight safety factor range of the drone corresponding to each safety level stored in the database. If the flight safety factor of the drone is within the flight safety factor range of the drone corresponding to a certain safety level stored in the database, then the safety level is used as the safety level of the drone.

[0064] It should be noted that the calculation formula for obtaining the flight safety factor of the drone is: ,in Indicates the The flight safety factor of the drone at each monitoring time point, 、 Respectively represent The temperature and vibration amplitude of the drone at each monitoring time point, 、 The safety temperature and vibration amplitude of the drone are stored in the database and are set by professionals. The temperature of the drone refers to the temperature of the drone's drive motor. The flight safety factor ranges of the drone corresponding to each safety level stored in the database are also set by professionals. The lower the safety level, the safer the drone. The advantages of this formula are: first, by using the safety temperature and vibration amplitude set by professionals as a benchmark, it scientifically defines the safety threshold of the drone's motor operation, providing a reliable reference for calculating the flight safety factor; second, through the quantitative comparison of real-time temperature and vibration amplitude, it intuitively reflects the safety level of the drone's operating status. Combined with the professionally classified safety levels, it provides a convenient basis for real-time safety monitoring and risk management of drones.

[0065] If the drone's safety level is level one, the drone will continue to fly along the optimal flight path. If the drone's safety level is level two, it will send an emergency fault message to the ground control center, requesting the operator's guidance, while maintaining flight balance and reducing flight speed. If the drone's safety level is level three, the emergency landing procedure will be immediately initiated, and a safe landing point will be selected based on the current position and surrounding environment. At the same time, an alarm will be issued to the surrounding area through sound and light signals and wireless communication equipment. After receiving the serious fault information, the ground control center will immediately activate the emergency rescue plan and organize personnel to go to the landing point for rescue.

[0066] Step 4: Cargo receipt and return. When the drone arrives at the destination, it adjusts its delivery posture in real time at the designated take-off and landing point and automatically lands. The user signs for the cargo, and the drone chooses to return to recharge or go to the next mission point based on its own power status.

[0067] In a specific embodiment, the delivery posture is adjusted in real time at the designated take-off and landing point. The specific implementation method is: extract the pitch angle, roll angle, and yaw angle of the drone, combine them with the current wind speed, calculate the delivery safety factor of the drone, and compare it with the delivery safety factor threshold stored in the database. If the delivery safety factor of the drone is higher than the delivery safety factor threshold stored in the database, it is determined that the drone can be delivered. If the delivery safety factor of the drone is lower than the delivery safety factor threshold stored in the database, it is determined that the drone cannot be delivered, and the pitch angle, roll angle, and yaw angle of the drone are automatically changed until the delivery safety factor of the drone is higher than the safety risk factor threshold stored in the database.

[0068] It should be noted that the drone is delivered at the designated take-off and landing point according to the preset delivery plan in the database. The calculation formula for obtaining the drone's delivery risk factor is: ,in Indicates the The delivery risk factor of the drone in the preset delivery plan, 、 、 Respectively represent The pitch angle, roll angle and yaw angle of the drone in a preset delivery plan, 、 、 The formula represents the pitch angle threshold, roll angle threshold, and yaw angle threshold of the drone stored in the database, which are set by professionals. The delivery risk factor threshold stored in the database is also set by professionals. The advantages of this formula are: first, by using the pitch angle, roll angle, and yaw angle thresholds set by professionals as a safety benchmark, it scientifically defines the safety boundaries of the drone's delivery posture, providing a reliable reference for risk factor calculation; second, through the quantitative comparison of real-time posture parameters and thresholds, it intuitively reflects the risk level of the delivery plan. Combined with the professionally set risk factor threshold, it provides an accurate basis for safety assessment and optimization decision-making of drone delivery plans.

[0069] In the present invention, standardized screening rules are established through clear comparison of weight and volume dual thresholds to ensure that drone loads are dangerous and avoid flight accidents or delivery failures caused by overloading or space mismatch. Real-time meteorological data, airspace restrictions and drone endurance are integrated, and the optimal path is screened through a risk factor model. UAV operation data is monitored in real time through sensors, and the flight risk factor is calculated and a graded response is given.

[0070] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they should all fall within the scope of protection of the present invention.

Claims

1. A method for detecting and tracking drones and adapting to low-altitude economic road patrol, characterized in that: include: Step 1: Information acquisition, which uses the combined sensor system onboard the drone to collect videos of vehicles passing through the target road. The combined sensor system includes a high-definition visible light camera and an infrared thermal imager. Step 2: Information processing: Based on the video of each vehicle on the target road, the behavior risk coefficient of each vehicle driver is calculated, and the driving risk coefficient of each vehicle driver is calculated; Step 3: Risk assessment and early warning: Based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, the risk level of each vehicle is assessed, and then control measures are taken based on the risk level of each vehicle.

2. The method for detecting and tracking drones and adapting to low-altitude economic road patrol according to claim 1 is characterized in that: The calculation obtains the behavior risk coefficient of each vehicle driver, and the specific implementation method is as follows: Analyze videos of each vehicle on the target road, detect and identify various types of distracted and fatigued driving behaviors of each driver, summarize and calculate the duration and frequency of each type of distracted behavior, and the duration and frequency of each type of fatigued driving behavior; Compare the duration and frequency of each type of distraction behavior of each driver with the duration and frequency of appropriate distraction behaviors stored in the database, and calculate the distraction behavior risk coefficient of each driver; Compare the duration and frequency of each type of fatigue driving behavior of each vehicle driver with the duration and frequency of appropriate fatigue driving behavior stored in the database, and calculate the fatigue driving behavior risk factor of each vehicle driver; Based on the distraction behavior risk coefficient and fatigue driving behavior risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the behavior risk coefficient of each vehicle driver.

3. The method for detecting and tracking drones and adapting to low-altitude economic road patrol according to claim 1 is characterized in that: The driving risk coefficient of each vehicle driver is obtained by the calculation, and the specific implementation method is as follows: Extract the speed of each vehicle on the target road at each monitoring time point from the video of each vehicle on the target road, and compare it with the maximum speed threshold of the target road stored in the database. If the speed of a vehicle passing through the target road at a certain monitoring time point is greater than the maximum speed threshold of the target road stored in the database, it is determined that the vehicle passing through the target road is speeding. Then, based on the speed of each vehicle passing through the target road at each monitoring time point and the maximum speed threshold of the target road stored in the database, the speed hazard factor of each vehicle is calculated; Analyze the video of each vehicle on the target road, detect and analyze various types of violations by each vehicle, summarize and obtain the duration and frequency of each type of violation by each vehicle, compare it with the duration and frequency of appropriate violations stored in the database, and calculate the violation risk factor of each vehicle; Based on the speed risk coefficient and violation risk coefficient of each vehicle, the driving risk coefficient of each vehicle driver is obtained by weighted calculation.

4. A method for detecting and tracking drones and adapting to low-altitude economic road patrol according to claim 3, characterized in that: The evaluation obtains the danger level of each vehicle, and the specific implementation method is as follows: Based on the behavioral risk coefficient and driving risk coefficient of each vehicle driver, a weighted calculation is performed to obtain the comprehensive risk coefficient of each vehicle. The comprehensive risk coefficient of each vehicle is compared with the comprehensive risk coefficient range of vehicles corresponding to each risk level stored in the database. If the comprehensive risk coefficient of a vehicle is within the comprehensive risk coefficient range of vehicles corresponding to a certain risk level stored in the database, then the risk level is used as the risk level of the vehicle.

5. The method for detecting and tracking drones and adapting to low-altitude economic road patrol according to claim 4 is characterized in that: The control measures are taken based on the danger level of each vehicle, and the specific implementation method is as follows: The danger level of each vehicle is extracted. If the danger level of a vehicle is level one, the drone will automatically adjust its flight attitude and position, focus on tracking the vehicle, and transmit relevant information to the ground command center in real time. The ground command center can remind the vehicle driver of the danger through text messages or voice prompts. If the danger level of a vehicle is level two, the drone will immediately send an emergency warning to the ground command center, turn on the sound and light warning device, and warn the vehicle in the air. At the same time, the ground command center dispatches law enforcement vehicles to intercept and inspect the vehicle.

6. A drone detection and tracking method adapted to low-altitude economic road logistics, characterized by: include: Step 1: Order reception and processing: Receive the user's logistics order, screen the goods in the order, and determine whether the goods are suitable for drone delivery. If suitable, generate a delivery task, which includes the starting point, destination, and timeliness requirements of the goods; Step 2: Path planning and drone dispatching: extracting real-time weather information, airspace restrictions, and drone endurance to plan the optimal flight path for the delivery mission, dispatching available drones from storage nodes, loading the cargo onto the drones, and calibrating the load balance; Step 3: Flight delivery: Control the drone to fly according to the planned flight path. During the flight, the drone will perform obstacle avoidance operations in real time through the sensors on board. If an emergency occurs, the corresponding emergency mechanism will be triggered. Step 4: Cargo receipt and return. When the drone arrives at the destination, it adjusts its delivery posture in real time at the designated take-off and landing point and automatically lands. The user signs for the cargo, and the drone chooses to return to recharge or go to the next mission point based on its own power status.

7. A method for detecting and tracking drones and adapting to low-altitude economic road logistics according to claim 6, characterized in that: The specific implementation method for determining whether the goods are suitable for drone delivery is as follows: Obtain the weight and volume of the goods in the order and compare them with the load capacity and cargo hold volume of the dispatchable drone. If the weight of the goods in the order is higher than the load capacity of the dispatchable drone or the volume of the goods in the order is higher than the cargo hold volume of the dispatchable drone, the goods are determined to be unsuitable for drone delivery. If the weight of the goods in the order is lower than the load capacity of the dispatchable drone and the volume of the goods in the order is lower than the cargo hold volume of the dispatchable drone, the goods are determined to be suitable for drone delivery.

8. The method for detecting and tracking drones and adapting to low-altitude economic road logistics according to claim 6 is characterized in that: The optimal flight path for the delivery mission is planned by: Real-time weather information, including wind speed, wind direction, temperature, and precipitation probability, is obtained from the meteorological department. Airspace restriction information is obtained from the airspace management department. The drone's endurance, including maximum flight distance and remaining battery power, is obtained from the database. Preprocessing is performed to construct a map containing the starting point, end point, and surrounding environment information. Through the path search algorithm, all possible paths from the starting point to the end point are obtained, and then the safety factor of each possible path is calculated. The possible path corresponding to the maximum safety factor is screened and used as the optimal flight path.

9. A method for detecting and tracking drones and adapting to low-altitude economic road logistics according to claim 8, characterized in that: The triggering of the corresponding emergency mechanism is specifically implemented as follows: The drone is equipped with multiple sensors to monitor its operating data in real time. Based on the operating data, the drone's flight safety factor is calculated. The flight safety factor of the drone is compared with the flight safety factor ranges of drones corresponding to various safety levels stored in the database. If the flight safety factor of the drone is within the flight safety factor range of drones corresponding to a certain safety level stored in the database, the safety level is used as the safety level of the drone. If the drone's safety level is level one, the drone will continue to fly along the optimal flight path. If the drone's safety level is level two, it will send an emergency fault message to the ground control center, requesting the operator's guidance, while maintaining flight balance and reducing flight speed. If the drone's safety level is level three, the emergency landing procedure will be immediately initiated, and a safe landing point will be selected based on the current position and surrounding environment. At the same time, an alarm will be issued to the surrounding area through sound and light signals and wireless communication equipment. After receiving the serious fault information, the ground control center will immediately activate the emergency rescue plan and organize personnel to go to the landing point for rescue.

10. The method for detecting and tracking drones and adapting to low-altitude economic road logistics according to claim 6 is characterized in that: The specific implementation method of adjusting the delivery posture in real time at the designated take-off and landing point is as follows: The pitch angle, roll angle and yaw angle of the drone are extracted and combined with the current wind speed to calculate the delivery safety factor of the drone, which is then compared with the delivery safety factor threshold stored in the database. If the delivery safety factor of the drone is higher than the delivery safety factor threshold stored in the database, the drone is determined to be capable of delivery. If the delivery safety factor of the drone is lower than the delivery safety factor threshold stored in the database, the drone is determined not to be capable of delivery. The drone's pitch angle, roll angle and yaw angle are then automatically changed until the delivery safety factor of the drone is higher than the safety risk factor threshold stored in the database.

Images