Low-altitude unmanned aerial vehicle operation risk and grade identification method
Through real-time data collection and three-dimensional grid path planning, combined with gray prediction model, the static early warning lag and dynamic early warning unreliability in drone operation risk identification and avoidance is solved, and the precise obstacle avoidance and safety of drone operation is achieved Sex improvement.
Patent Information
- Application Number
- CN202411153596.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-08-21
AI Technical Summary
The existing technology has problems such as static early warning lag, inaccurate path planning and unreliable dynamic early warning in terms of drone operation risks, which affects the operation efficiency and safety of drones and makes it difficult to achieve accurate obstacle avoidance.
By collecting drone operation data in real time, setting static and dynamic alert areas, path planning is performed based on the three-dimensional grid method, monitoring drone location in real time, predicting future locations using gray prediction models, and selecting the optimal evasion action execution.
It realizes accurate analysis and avoidance of drone operation risks, and improves the operational safety and efficiency of drones in complex environments.
Smart Images

Figure CN118968823A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of unmanned aerial vehicle operation monitoring, and in particular to a method for identifying risks and levels of low-altitude unmanned aerial vehicle operation. Background Art
[0002] In recent years, low-altitude UAV (Unmanned Aerial Vehicles) technology has developed rapidly in many fields, including logistics distribution, agricultural monitoring, environmental protection, security monitoring and disaster relief. With the continuous advancement of UAV technology and the expansion of its application scope, the operation safety of low-altitude UAVs has gradually attracted widespread attention. Traditional UAV operation risk identification and avoidance methods mainly rely on ground station monitoring and manual operation. Although this method ensures the operation safety of UAVs to a certain extent, it cannot fully cope with the complex low-altitude flight environment due to the limitations of manual reaction speed and judgment ability. In recent years, UAV operation risk identification and avoidance technology based on automation and intelligence has gradually become a research hotspot. Although some results have been achieved, there are still many problems. The existing technology has problems in the identification and avoidance of UAV operation risks, such as static warning lag, inaccurate path planning and unreliable dynamic warning, which affects the operation efficiency and safety of UAVs and makes it difficult to achieve accurate obstacle avoidance. Summary of the invention
[0003] In view of the problems existing in the above-mentioned existing low-altitude UAV operation risk and level identification methods, the present invention is proposed.
[0004] Therefore, the problem to be solved by the present invention is that the existing technology has problems such as static warning lag, inaccurate path planning and unreliable dynamic warning in the identification and avoidance of risks in the operation of drones, which affects the operating efficiency and safety of drones and makes it difficult to achieve accurate obstacle avoidance.
[0005] To solve the above technical problems, the present invention provides the following technical solutions: a method for identifying the risks and levels of low-altitude UAV operation, which includes collecting UAV operation data in real time, setting static alarm areas, and setting dynamic alarm areas according to the operation data; planning the UAV operation route based on a three-dimensional grid method, and predicting the future UAV position based on the UAV operation data; monitoring the UAV position in real time, and issuing static warnings based on the static alarm areas, and synchronously calculating the relative distances between UAVs and issuing dynamic warnings based on the dynamic alarm areas; selecting the optimal avoidance action for the UAV and executing it.
[0006] As a preferred solution of the method for identifying the risk and level of low-altitude UAV operation of the present invention, wherein: the real-time collection of UAV operation data, setting of static alarm areas, and setting of dynamic alarm areas according to the operation data refers to collecting the operation data of the UAV in real time through sensors, generating a unique identification of the UAV according to the UAV information, and setting the static alarm area of the UAV according to the safe operation interval of the UAV; Define the dynamic alarm zone update interval as , predicting the time interval based on the collected real-time drone operation data The drone position after: Where x(t), y(t) and z(t) are the coordinate positions of the drone at time t, V(t) is the speed of the drone at time t, and are the azimuth and pitch angles of the UAV at time t, , as well as For time interval The coordinate position of the UAV after the Calculate time t and time interval The distance D between the coordinate position of the UAV and the dynamic alarm area is used as the radius of the dynamic alarm area to form a spherical area, and the spherical area is used as the dynamic alarm area.
[0007] As a preferred solution of the method for identifying the risk and level of low-altitude UAV operation of the present invention, wherein: the UAV operation route planning based on the three-dimensional grid method refers to dividing the airspace into three-dimensional grids, and each grid unit represents a spatial unit; The center point of the space unit is used as the jumping point, the initial space unit of the drone is used as the initial jumping point, and the destination space unit is used as the final jumping point. The JPS algorithm is used for initial path planning, and the path jumping points are randomly selected from the surroundings of the initial jumping point. The moving cost g(v,c) from the initial jumping point to the path jumping point is calculated: Where g(v) is the movement cost of jump point v. If jump point v is the initial jump point, the movement cost g(v) is 0. d(v, c) is the distance from jump point v to jump point c. Use Manhattan distance to calculate the estimated cost h(c, z) of the path from hop c to the final hop z: where x c ,y c and z c is the coordinate position of the jump point c, x z ,y z and z z is the coordinate position of the final jump point z; When selecting a path jump point, calculate the comprehensive cost f of the path jump point: Select the path jump point with the smallest comprehensive cost f as the next jump point, and use the YOLO algorithm to identify obstacles in the space unit of the next jump point, connect the diagonal of the space unit, and identify the position of the obstacle. If the obstacle is not on the diagonal of the space unit, the path of the drone in the space unit is diagonal. If the obstacle is on the diagonal of the space unit, reselect the jump point from the adjacent space units around the space unit. Repeatedly select path jump points until the final jump point z is reached, record the selected path jump points, determine the UAV path according to the spatial unit of the path jump points, and perform path smoothing through Bezier curves. Calculate the position of the UAV at each time point according to the UAV's running speed, and record the time point and position coordinates.
[0008] As a preferred solution of the method for identifying the risk and level of low-altitude UAV operation of the present invention, the method of predicting the future UAV position based on the UAV operation data includes: Collect the historical operation data of the drone, construct a vector of the operation data at each time point, and form the initial vector sequence P: Where p(n) is the running data vector at the nth time point; Perform cumulative transformation on the initial vector sequence P to generate a new sequence P (1): Where p(i) is the running data vector at the i-th time point; Smoothing the sequence P(1) generates the background value sequence Z(1): Where P(1) k is the kth sequence vector in the sequence P(1), P(1) k-1 is the k-1th sequence vector in the sequence P (1); Construct the grey prediction model formula: Where p(t+1) is the running data vector at time point t+1, and a and b are model parameter vectors; Establish the differential equation of grey prediction model: The observation matrix Y is constructed based on the initial vector sequence P, and the background value matrix Q is constructed based on the background value sequence Z (1). The model parameter vector is solved by the least squares method: Where T is the transpose operation; After obtaining the model parameter vector, the grey prediction model formula is inserted to obtain the operation data vector p(t+1) at time point t+1, and the position coordinates of the UAV at time point t+1 are extracted from the operation data vector p(t+1); Define the safety range of each drone operation data according to the drone design specification, and calculate the fuzzy score S of each drone operation data u : o is the operation data of a single type of drone, including the operation data of all types of drones, omax and min is the upper and lower limits of the safety range of a single type of UAV operation data, o t+1 is the single type of UAV operation data value at time t+1, o cen It is the middle value of the upper and lower limits of the safety range of the operating data of a single type of drone; The comprehensive evaluation value at time t+1 is calculated based on the fuzzy score of each drone operation data: Where L(t+1) is the comprehensive evaluation value at time t+1, m is the type of UAV operation data, S u is the fuzzy score of the u-th type of UAV operation data; Starting from the time point when the historical operation data of the drone is collected, the comprehensive evaluation value of each time point is calculated in turn, and the model parameter vector adjustment coefficient r is calculated according to the comprehensive evaluation value of the t+1 time point: Where L(w) is the comprehensive evaluation value at the wth time point; Use the adjustment coefficient r to adjust the model parameter vectors a and b: where a new and b new is the adjusted model parameter vector, a old and b old The model parameter vector before adjustment is substituted into the grey prediction model formula to continuously predict the future operation data vector of the UAV, and the model parameter vector is adjusted iteratively.
[0009] As a preferred solution of the method for identifying the risks and levels of low-altitude UAV operations described in the present invention, the real-time monitoring of the UAV position and the static early warning based on the static alarm area refers to real-time collection of UAV operation data to extract the UAV position coordinates, and automatically issuing a static early warning when an invading UAV is detected entering the static alarm area of our UAV to remind staff to fly carefully and display the unique identification of the invading UAV.
[0010] As a preferred solution of the method for identifying the risks and levels of low-altitude UAV operations described in the present invention, the synchronous calculation of the relative distance between UAVs and the dynamic warning based on the dynamic alarm area means that when an invading UAV is detected to have invaded the static alarm area of our UAV, the position coordinates of the invading UAV and the position coordinates of our UAV are obtained, and the distance between the invading UAV and our UAV is calculated in real time. When the distance between the invading UAV and our UAV is less than the radius of the dynamic alarm area, a dynamic warning is automatically issued to remind the staff of the risk of collision, and an avoidance command is generated synchronously.
[0011] As a preferred solution of the method for identifying the operation risk and level of low-altitude UAVs of the present invention, wherein: after the optimal evasive action of the UAV is selected and the evasive command is generated, all possible evasive actions are simulated starting from the initial node according to the current operation data of our UAV as the initial node, each evasive action is simulated and the future position coordinates of the invading UAV and our UAV are predicted using the grey prediction model formula to detect the evasive effect; If the invading UAV and our UAV have no overlap in their future position coordinates at the same time after the avoidance action simulation, it will be recorded as a successful avoidance, and the time point and UAV position coordinates after the successful avoidance will be obtained. The distance between the UAV position coordinates at that time point and the UAV position coordinates at that time point in the planned UAV path will be calculated, and the avoidance action with the smallest distance will be selected as the optimal avoidance action.
[0012] As a preferred solution of the method for identifying the risks and levels of low-altitude UAV operations described in the present invention, the UAV operation data includes the UAV position coordinates, operation speed, acceleration, direction angle and pitch angle, and operation time series.
[0013] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of the above-mentioned method for identifying the risks and levels of low-altitude unmanned aerial vehicle operations when executing the computer program.
[0014] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for identifying the risks and levels of low-altitude unmanned aerial vehicle operations.
[0015] The beneficial effects of the present invention are as follows: the present invention sets static alarm areas and dynamic alarm areas by collecting drone operation data, plans drone paths, predicts the future position coordinates of the drone in real time, realizes monitoring of the future operation status of the drone, and warns of drone operation risks in combination with dynamic alarm areas, realizes accurate drone operation risk analysis, and calculates the optimal avoidance action execution, thereby improving the operation safety of drones in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0017] Figure 1 This is a flowchart of the method for identifying the risks and levels of low-altitude UAV operations.
[0018] Figure 2 This is a structural diagram of the method for identifying the risks and levels of low-altitude UAV operations. DETAILED DESCRIPTION
[0019] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.
[0020] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.
[0021] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments. Example
[0022] Reference Figure 1 and Figure 2 , which is the first embodiment of the present invention, and provides a method for identifying the risks and levels of low-altitude UAV operations. The method for identifying the risks and levels of low-altitude UAV operations includes: S1. Collect drone operation data in real time, set static alarm areas, and set dynamic alarm areas based on the operation data; Specifically, real-time collection of drone operation data, setting of static alarm areas, and setting of dynamic alarm areas based on the operation data refers to real-time collection of drone operation data through sensors, generation of a drone unique identifier based on drone information, and setting of drone static alarm areas based on drone safe operation intervals; Define the dynamic alarm zone update interval as , predicting the time interval based on the collected real-time drone operation data The drone position after: Where x(t), y(t) and z(t) are the coordinate positions of the drone at time t, V(t) is the speed of the drone at time t, and are the azimuth and pitch angles of the UAV at time t, , as well as For time interval The coordinate position of the drone after Calculate time t and time interval The distance D between the coordinate position of the UAV and the dynamic alarm area is used as the radius of the dynamic alarm area to form a spherical area, and the spherical area is used as the dynamic alarm area.
[0023] By collecting drone operation data in real time, the system can ensure that it always has the latest status information of the drone, thereby improving the accuracy of early warning and path planning. Real-time data collection can promptly reflect the speed, position and attitude changes of the drone, enabling the system to respond quickly to emergencies. The setting of the static alarm area is based on the safe operation interval of the drone, which can effectively prevent the drone from issuing an alarm before entering the dangerous area and take measures to avoid it in advance. The fixedness of the area ensures basic safety protection. The setting of the dynamic alarm area is based on real-time data to predict the future position of the drone, and can adjust the warning range in real time with the movement of the drone. The dynamic alarm area is more flexible and accurate than the static alarm area, and can more effectively prevent potential collision risks. Reasonable setting of the time interval can ensure the prediction accuracy of the dynamic alarm area and avoid warning errors caused by too long or too short time intervals. Through the unique identification of the drone, independent monitoring and management of multiple drones can be achieved to avoid data confusion and improve the management efficiency and accuracy of the system. By calculating the time t and the time interval The distance D of the UAV's position after the detection is used as the radius of the dynamic alarm area to form a spherical early warning area. This area can effectively cover the possible future flight range of the UAV and provide a wider range of early warning protection.
[0024] It should also be noted that the drone operation data includes the drone position coordinates, operation speed, acceleration, direction angle and pitch angle, and operation time series.
[0025] S2, UAV route planning based on 3D grid method, and predicting the future UAV position based on UAV operation data; Specifically, the UAV operation route planning based on the three-dimensional grid method refers to dividing the airspace into three-dimensional grids, where each grid cell represents a spatial unit; The center point of the space unit is used as the jumping point, the initial space unit of the drone is used as the initial jumping point, and the destination space unit is used as the final jumping point. The JPS algorithm is used for initial path planning, and the path jumping points are randomly selected from the surroundings of the initial jumping point. The moving cost g(v,c) from the initial jumping point to the path jumping point is calculated: Where g(v) is the movement cost of jump point v. If jump point v is the initial jump point, the movement cost g(v) is 0. d(v, c) is the distance from jump point v to jump point c. Use Manhattan distance to calculate the estimated cost h(c, z) of the path from hop c to the final hop z: where x c ,y c and zc is the coordinate position of the jump point c, x z ,y z and z z is the coordinate position of the final jump point z; When selecting a path jump point, calculate the comprehensive cost f of the path jump point: Select the path jump point with the smallest comprehensive cost f as the next jump point, and use the YOLO algorithm to identify obstacles in the space unit of the next jump point, connect the diagonal of the space unit, and identify the position of the obstacle. If the obstacle is not on the diagonal of the space unit, the path of the drone in the space unit is diagonal. If the obstacle is on the diagonal of the space unit, reselect the jump point from the adjacent space units around the space unit. Repeatedly select path jump points until the final jump point z is reached, record the selected path jump points, determine the UAV path according to the spatial unit of the path jump points, and perform path smoothing through Bezier curves. Calculate the position of the UAV at each time point according to the UAV's running speed, and record the time point and position coordinates.
[0026] Divide the airspace into three-dimensional grids, each grid unit represents a spatial unit, which can effectively manage and simplify complex flight space. Through the gridding method, the characteristics of each spatial area can be accurately defined, which is convenient for path planning and obstacle detection. The JPS algorithm optimizes the traditional A* algorithm through jump point search, reduces the inspection of redundant nodes, and improves the efficiency and speed of path search. Using the JPS algorithm, the optimal path from the initial jump point to the final jump point of the drone can be quickly found. By calculating the movement cost and estimated cost of the jump point, the actual movement cost and estimated cost of the path can be comprehensively considered to ensure that the selected path is both economical and efficient. Manhattan distance is used For cost estimation, it can quickly estimate the distance from the jump point to the final jump point. Through the YOLO algorithm, it can identify obstacles in the space unit where the path jump point is located in real time, accurately detect potential flight risks, and adjust the flight path according to the location of the obstacle. If the obstacle is not on the diagonal of the space unit, the drone can fly along the diagonal. If the obstacle is on the diagonal, reselect the jump point and smooth the path through the Bezier curve. The flight path of the drone can be made more natural and smooth, reducing sharp turns and frequent speed changes, and optimizing flight performance. The Bezier curve can generate a smooth curve based on the control point to meet the actual flight needs of the drone.
[0027] Furthermore, the prediction of the future position of the drone based on the drone operation data includes: Collect the historical operation data of the drone, construct a vector of the operation data at each time point, and form the initial vector sequence P: Where p(n) is the running data vector at the nth time point; Perform cumulative transformation on the initial vector sequence P to generate a new sequence P (1): Where p(i) is the running data vector at the i-th time point; Smoothing the sequence P(1) generates the background value sequence Z(1): Where P(1) k is the kth sequence vector in the sequence P(1), P(1) k-1 is the k-1th sequence vector in the sequence P (1); Construct the grey prediction model formula: Where p(t+1) is the running data vector at time point t+1, and a and b are model parameter vectors; Establish the differential equation of grey prediction model: The observation matrix Y is constructed based on the initial vector sequence P, and the background value matrix Q is constructed based on the background value sequence Z (1). The model parameter vector is solved by the least squares method: Where T is the transpose operation; After obtaining the model parameter vector, the grey prediction model formula is inserted to obtain the operation data vector p(t+1) at time point t+1, and the position coordinates of the UAV at time point t+1 are extracted from the operation data vector p(t+1); Define the safety range of each drone operation data according to the drone design specification, and calculate the fuzzy score S of each drone operation data u : o is the operation data of a single type of drone, including the operation data of all types of drones, o max and min is the upper and lower limits of the safety range of a single type of UAV operation data, o t+1 is the single type of UAV operation data value at time t+1, o cen It is the middle value of the upper and lower limits of the safety range of the operating data of a single type of drone; The comprehensive evaluation value at time t+1 is calculated based on the fuzzy score of each drone operation data: Where L(t+1) is the comprehensive evaluation value at time t+1, m is the type of UAV operation data, S u is the fuzzy score of the u-th type of UAV operation data; The comprehensive evaluation value of each time point is calculated in sequence from the time point when the historical operation data of the drone is collected. The model parameter vector adjustment coefficient r is calculated according to the comprehensive evaluation value of the t+1 time point: Where L(w) is the comprehensive evaluation value at the wth time point; Use the adjustment coefficient r to adjust the model parameter vectors a and b: where a new and b newis the adjusted model parameter vector, a old and b old The model parameter vector before adjustment is substituted into the grey prediction model formula to continuously predict the future operation data vector of the UAV, and the model parameter vector is adjusted iteratively.
[0028] By collecting and organizing the historical operation data of the UAV and constructing the initial vector sequence P, the operation status of the UAV at different time points can be systematically recorded and analyzed, providing basic data for subsequent data processing and model construction. The cumulative transformation can smooth the data, reduce the random fluctuations of the data, highlight the trend of the data, and facilitate the subsequent model construction and prediction. This step can enhance the smoothness and predictability of the data and improve the accuracy and reliability of the grey prediction model. The smoothing process can further reduce the fluctuations of the data, improve the stability and continuity of the data, and facilitate the subsequent model construction and parameter estimation. It can improve the quality of the background value sequence and provide more stable data support for the accurate modeling of the grey prediction model. By establishing the differential equation of the grey prediction model, the data after the cumulative transformation can be associated with time to form a mathematical model that can be used to predict the future state. It can provide a mathematical basis for the prediction of the UAV operation data and ensure the scientificity and operability of the model. The measured value matrix and the background value matrix are calculated, and the least squares method is used to solve the model parameters, which can accurately determine the parameters of the gray prediction model and improve the model's fit and prediction accuracy. By fuzzy scoring the UAV operation data, the UAV's operating status can be quantitatively evaluated, which is convenient for comprehensive evaluation and risk warning. It can provide quantitative indicators for the evaluation and management of UAV operation safety, improve the scientificity and accuracy of safety assessment, and by calculating the comprehensive evaluation value, the UAV's operating status can be comprehensively evaluated. The model parameters can be adjusted according to the evaluation results to improve the model's adaptability and prediction accuracy. It can ensure that the gray prediction model remains efficient and accurate in a changing environment, and provide technical support for the real-time evaluation and adjustment of the UAV's operating status. Through continuous prediction and iterative adjustment of model parameters, the UAV's operation prediction can be updated in real time, and the accuracy and real-time nature of the prediction can be improved. It can provide technical support for real-time monitoring and dynamic adjustment of UAVs, and improve the operating efficiency and safety of UAVs.
[0029] S3, real-time monitoring of the drone position, and static warning according to the static alarm area, synchronous calculation of the relative distance between drones and dynamic warning according to the dynamic alarm area; Specifically, real-time monitoring of the drone's position and static warning based on the static alarm area means real-time collection of drone operation data to extract the drone's position coordinates. When an invading drone is detected entering the static alarm area of our drone, a static warning will be automatically issued to remind staff to fly carefully and display the unique identification of the invading drone.
[0030] By collecting the operation data of the drone in real time through sensors and extracting the current position coordinates of the drone, the flight status of the drone can be accurately tracked. Real-time monitoring of the drone's position can ensure the accuracy and safety of the flight path, and timely detect and handle abnormal situations. According to the flight environment and mission requirements of the drone, the static alarm area is pre-set. When the drone enters the area, the system can trigger an early warning in time to remind the operator or the automatic system to intervene. The setting of the static alarm area can effectively prevent the drone from colliding with fixed obstacles or other drones, thereby improving flight safety. When an invading drone is detected entering the static alarm area, the system will automatically trigger the static early warning mechanism to remind the staff to pay attention to flight safety and display the unique identification of the invading drone. The static early warning can issue an alarm as soon as the drone enters the dangerous area to prevent potential accidents. By displaying the unique identification of the invading drone, the potential threat source can be clearly identified, which is convenient for the operator to monitor and handle. The unique identification can help the operator quickly identify and locate the invading drone and take effective avoidance measures.
[0031] Furthermore, the relative distance between drones is calculated synchronously and dynamic warning is given according to the dynamic alarm area. When an invading drone is detected to have invaded the static alarm area of our drone, the position coordinates of the invading drone and the position coordinates of our drone are obtained, and the distance between the invading drone and our drone is calculated in real time. When the distance between the invading drone and our drone is less than the radius of the dynamic alarm area, a dynamic warning is automatically given to remind the staff of the risk of collision, and an avoidance command is generated synchronously.
[0032] Through sensors and data links, the position coordinates of the invading drone and our drone are obtained in real time, and the flight status and relative position of the drone can be accurately tracked, which provides accurate data support for subsequent distance calculation and dynamic warning. By calculating the relative distance between the invading drone and our drone in real time, the risk of collision between the two can be judged in time. When the distance is less than the set dynamic alarm area radius, the system can quickly issue an early warning. When the invading drone enters the static alarm area of our drone and the relative distance between the two is less than the dynamic alarm area radius, the system will automatically issue a dynamic early warning. This early warning mechanism can reflect changes in the flight environment in real time and promptly remind operators to pay attention to potential collision risks. While issuing a dynamic early warning, the system will automatically generate an avoidance command to guide the drone to change its flight path and avoid potential collision risks. The automatic generation and execution of avoidance commands can reduce delays in human operations and improve the response speed and safety of drones.
[0033] S4, select the optimal evasive action for the drone and execute it; Specifically, after selecting the optimal evasive action of the drone and executing the generated evasive command, the current operating data of our drone is used as the initial node, and all possible evasive actions are simulated from the initial node. Each evasive action is simulated and the future position coordinates of the invading drone and our drone are predicted using the grey prediction model formula to detect the evasive effect; If the invading UAV and our UAV have no overlap in their future position coordinates at the same time after the avoidance action simulation, it will be recorded as a successful avoidance, and the time point and UAV position coordinates after the successful avoidance will be obtained. The distance between the UAV position coordinates at that time point and the UAV position coordinates at that time point in the planned UAV path will be calculated, and the avoidance action with the smallest distance will be selected as the optimal avoidance action.
[0034] After detecting a potential collision risk, the system automatically generates an avoidance command, which includes a series of possible avoidance actions, such as changing speed, heading, and altitude. These commands are generated based on the current flight status of the drone and the surrounding environment data. Starting from the initial node, all possible avoidance actions are simulated, and the effect of each avoidance action is tested through simulation. During the simulation, the gray prediction model is used to predict the future position coordinates of the invading drone and our drone, and the effectiveness of the avoidance action is evaluated. The gray prediction model is used to predict the future position coordinates of the invading drone and our drone after the avoidance action. The model can make effective predictions in the case of incomplete data and provide highly accurate future position data. During the simulation, it is detected whether the position coordinates of the invading drone and our drone at the same time point in the future overlap. If there is no overlap, it is recorded as a successful avoidance, and the time point and drone position coordinates after the successful avoidance are obtained. The distance between the drone position coordinates at the time point after the successful avoidance and the position coordinates at the time point in the planned path is calculated, and the avoidance action with the smallest distance is selected as the optimal avoidance action to be executed. The optimal avoidance action can not only effectively avoid the risk of collision, but also make the drone as close to the original planned path as possible to reduce flight deviation. Example
[0035] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.
[0036] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.
[0037] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.
[0038] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.
[0039] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.
Claims
1. A method for identifying the risks and levels of low-altitude UAV operations, characterized by: include, Collect drone operation data in real time, set static alarm areas, and set dynamic alarm areas based on the operation data; The UAV operation route is planned based on the 3D grid method, and the future UAV position is predicted based on the UAV operation data; Monitor the position of drones in real time and issue static warnings based on static warning areas. Simultaneously calculate the relative distance between drones and issue dynamic warnings based on dynamic warning areas. Select the best evasive action for the drone and execute it.
2. The method for identifying the risk and level of low-altitude UAV operation according to claim 1, characterized in that: The real-time collection of drone operation data, setting of static alarm areas, and setting of dynamic alarm areas according to the operation data refers to collecting drone operation data in real time through sensors, generating a drone unique identifier according to drone information, and setting a drone static alarm area according to the drone safe operation interval; Define the dynamic alarm zone update interval as , predicting the time interval based on the collected real-time drone operation data The drone position after: Where x(t), y(t) and z(t) are the coordinate positions of the drone at time t, V(t) is the speed of the drone at time t, and are the azimuth and pitch angles of the UAV at time t, , as well as For time interval The coordinate position of the UAV after the Calculate time t and time interval The distance D between the coordinate position of the UAV and the dynamic alarm area is used as the radius of the dynamic alarm area to form a spherical area, and the spherical area is used as the dynamic alarm area.
3. The method for identifying the risk and level of low-altitude UAV operation as claimed in claim 2, characterized in that: The UAV operation route planning based on the three-dimensional grid method refers to dividing the airspace into three-dimensional grids, each grid unit represents a spatial unit; The center point of the space unit is used as the jumping point, the initial space unit of the drone is used as the initial jumping point, and the destination space unit is used as the final jumping point. The JPS algorithm is used for initial path planning, and the path jumping points are randomly selected from the surroundings of the initial jumping point. The moving cost g(v,c) from the initial jumping point to the path jumping point is calculated: Where g(v) is the movement cost of jump point v. If jump point v is the initial jump point, the movement cost g(v) is 0. d(v, c) is the distance from jump point v to jump point c. Use Manhattan distance to calculate the estimated cost h(c, z) of the path from hop c to the final hop z: where x c ,y c and z c is the coordinate position of the jump point c, x z ,y z and z z is the coordinate position of the final jump point z; When selecting a path jump point, calculate the comprehensive cost f of the path jump point: Select the path jump point with the smallest comprehensive cost f as the next jump point, and use the YOLO algorithm to identify obstacles in the space unit of the next jump point, connect the diagonal of the space unit, and identify the position of the obstacle. If the obstacle is not on the diagonal of the space unit, the path of the drone in the space unit is diagonal. If the obstacle is on the diagonal of the space unit, reselect the jump point from the adjacent space units around the space unit. Repeatedly select path jump points until the final jump point z is reached, record the selected path jump points, determine the UAV path according to the spatial unit of the path jump points, and perform path smoothing through Bezier curves. Calculate the position of the UAV at each time point according to the UAV's running speed, and record the time point and position coordinates.
4. The method for identifying the risk and level of low-altitude UAV operation as claimed in claim 3, characterized in that: The prediction of the future position of the drone based on the drone operation data includes: Collect the historical operation data of the drone, construct a vector of the operation data at each time point, and form the initial vector sequence P: Where p(n) is the running data vector at the nth time point; Perform cumulative transformation on the initial vector sequence P to generate a new sequence P (1): Where p(i) is the running data vector at the i-th time point; Smoothing the sequence P(1) generates the background value sequence Z(1): Where P(1) k is the kth sequence vector in the sequence P(1), P(1) k-1 is the k-1th sequence vector in the sequence P (1); Construct the grey prediction model formula: Where p(t+1) is the running data vector at time point t+1, and a and b are model parameter vectors; Establish the differential equation of grey prediction model: The observation matrix Y is constructed based on the initial vector sequence P, and the background value matrix Q is constructed based on the background value sequence Z (1). The model parameter vector is solved by the least squares method: Where T is the transpose operation; After obtaining the model parameter vector, the grey prediction model formula is inserted to obtain the operation data vector p(t+1) at time point t+1, and the position coordinates of the UAV at time point t+1 are extracted from the operation data vector p(t+1); Define the safety range of each drone operation data according to the drone design specification, and calculate the fuzzy score S of each drone operation data u : o is the operation data of a single type of drone, including the operation data of all types of drones, o max and min is the upper and lower limits of the safety range of a single type of UAV operation data, o t+1 is the single type of UAV operation data value at time t+1, o cen It is the middle value of the upper and lower limits of the safety range of the operating data of a single type of drone; The comprehensive evaluation value at time t+1 is calculated based on the fuzzy score of each drone operation data: Where L(t+1) is the comprehensive evaluation value at time t+1, m is the type of UAV operation data, S u is the fuzzy score of the u-th type of UAV operation data; Starting from the time point when the historical operation data of the drone is collected, the comprehensive evaluation value of each time point is calculated in turn, and the model parameter vector adjustment coefficient r is calculated according to the comprehensive evaluation value of the t+1 time point: Where L(w) is the comprehensive evaluation value at the wth time point; Use the adjustment coefficient r to adjust the model parameter vectors a and b: where a new and b new is the adjusted model parameter vector, a old and b old The model parameter vector before adjustment is substituted into the grey prediction model formula to continuously predict the future operation data vector of the UAV, and the model parameter vector is adjusted iteratively.
5. The method for identifying the risk and level of low-altitude UAV operation as claimed in claim 4, characterized in that: The real-time monitoring of the drone position and static warning based on the static alarm area refers to real-time collection of drone operation data to extract the drone position coordinates. When an invading drone is detected entering the static alarm area of our drone, a static warning is automatically issued to remind staff to fly carefully and display the unique identification of the invading drone.
6. The method for identifying the risk and level of low-altitude UAV operation according to claim 5, characterized in that: The synchronous calculation of the relative distance between drones and the dynamic warning based on the dynamic alarm area means that when an invading drone is detected to have invaded the static alarm area of our drone, the position coordinates of the invading drone and the position coordinates of our drone are obtained, and the distance between the invading drone and our drone is calculated in real time. When the distance between the invading drone and our drone is less than the radius of the dynamic alarm area, a dynamic warning is automatically issued to remind the staff of the risk of collision, and an avoidance command is generated synchronously.
7. The method for identifying the risk and level of low-altitude UAV operation according to claim 6, characterized in that: After the optimal evasion action of the drone is selected and the evasion command is generated, the current operation data of our drone is used as the initial node, all possible evasion actions are simulated from the initial node, each evasion action is simulated, and the future position coordinates of the invading drone and our drone are predicted using the grey prediction model formula to detect the evasion effect; If the invading UAV and our UAV have no overlap in their future position coordinates at the same time after the avoidance action simulation, it will be recorded as a successful avoidance, and the time point and UAV position coordinates after the successful avoidance will be obtained. The distance between the UAV position coordinates at that time point and the UAV position coordinates at that time point in the planned UAV path will be calculated, and the avoidance action with the smallest distance will be selected as the optimal avoidance action.
8. The method for identifying the risk and level of low-altitude UAV operation according to claim 7, characterized in that: The UAV operation data includes the UAV position coordinates, operation speed, acceleration, direction angle and pitch angle, and operation time series.
9. A computer device comprising: Memory and processor; The memory stores a computer program, which is characterized in that when the processor executes the computer program, the steps of the method for identifying the risks and levels of low-altitude unmanned aerial vehicle operations described in any one of claims 1 to 8 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying the risks and levels of low-altitude UAV operations described in any one of claims 1 to 8 are implemented.
Citation Information
Patent Citations
Air traffic collision avoidance method based on state prediction
CN106548661A
Multi-target route planning and united searching method of drone in urban low-altitude environment
CN106970648A
Real-time acquisition method for early warning range applicable to low-altitude drone
CN110223539A
Dynamic risk situation awareness method for low-altitude airspace unmanned aerial vehicle operation
CN118191844A
Low-altitude airspace management method and system based on communication and sensing integration
CN118280168A
Cited By
Risk identification and emergency method for unmanned aerial vehicle flying in airspace adjacent to expressway
CN120335478A
Risk identification and emergency response methods when drones fly in airspace near highways
CN120335478B
Low-altitude unmanned aerial vehicle traffic operation risk early warning method and system
CN120356365A
A low-altitude UAV traffic operation risk warning method and system
CN120356365B
