Multi-uav cooperative search positioning method and system

Abstract

The present application relates to the technical field of aviation search and rescue, and discloses a multi-unmanned aerial vehicle cooperative search and positioning method and system, which comprises the following steps: at least three unmanned aerial vehicles are caused to fly along straight lines at an included angle at the same time, a lifesaving radio receiver is carried to listen to a target signal, starting point and ending point coordinates of signal receiving are independently recorded and returned to a data processing center; the data processing center generates a signal bearing line based on the starting and ending points, and when the number of effective bearing lines is greater than or equal to three, a weighted least square method is used to solve a target estimated position which makes the weighted distance square sum of coordinate points to each bearing line minimum. The present application does not need to synchronize a direction-finding antenna and a high-precision clock, and through a three-level data screening mechanism, reliability is ensured, the search and rescue response time is shortened from several hours to several tens of minutes, the positioning error can be converged to within 50 meters, and the timeliness, accuracy and economy of aviation search and rescue are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of aviation search and rescue technology, and more specifically, to a method and system for collaborative search and positioning of multiple unmanned aerial vehicles (UAVs). Background Technology

[0002] In maritime or wilderness search and rescue missions, quickly locating people in distress is crucial for a successful rescue. In recent years, the industry has proposed various technical solutions focusing on rapid delineation of search and rescue areas, optimization of search and rescue strategies, and multi-platform mission collaboration.

[0003] Regarding the calculation of search and rescue areas, Chinese invention patents CN116541637A (titled "A Method for Calculating Search and Rescue Areas for Parachute Landings") and CN118657261A (titled "A Method for Calculating Search and Rescue Areas for Plane Crashes on Land") provide methods for calculating search and rescue areas for different scenarios. These methods can quickly delineate the search and rescue area at a macro level, providing regional guidance for subsequent refined searches. Regarding search and rescue strategies and coordination, two Chinese invention patents, CN118839089A (titled "A Search Strategy Method for Land-based Air Search and Rescue Missions") and CN117591795A (titled "A Method for Estimating the Success Rate of Multi-Aircraft Participation in Large-scale Maritime Search and Rescue"), propose search and rescue strategies and corresponding success rate assessments, as well as mission coordination methods for dual-aircraft search and rescue missions. These patents can optimize route planning and resource allocation under multi-platform participation conditions. Meanwhile, CN119673000A (titled "A DogLeg-based Escort Search and Rescue Calculation Method and System") proposes route and time coordination methods between search and rescue aircraft and auxiliary aircraft for escort scenarios to ensure the effectiveness of escort navigation during search and rescue operations. These existing technologies primarily focus on macro-level aspects such as search and rescue zone delineation, route coordination, and search and rescue force allocation. However, a complete technological loop has not yet been formed regarding how to quickly and accurately locate the target within the search and rescue zone.

[0004] Currently, commonly used search and rescue methods include visual search, infrared imaging search, airborne rescue radio direction finding, and satellite-based positioning rescue devices. Among these, the rescue radio is a crucial device for distressed personnel to transmit distress signals; its emitted VHF band 121.5MHz or UHF band 243MHz signals are standard distress signals used in aerial search and rescue. Positioning technology based on rescue radios has significant application value in aerial search and rescue due to its advantages such as requiring no additional operation from distressed personnel, wide coverage, and independence from third-party infrastructure.

[0005] However, existing positioning methods based on rescue radios have the following shortcomings: First, single-point direction finding accuracy is low. Traditional airborne rescue radio direction finding equipment can typically only provide a rough azimuth angle in a single flight. Affected by changes in UAV flight attitude, multipath effects, and the electromagnetic environment, the random error of a single direction finding is relatively large, making it difficult to directly provide high-precision target position information. Second, there is a lack of effective methods for multi-azimuth line position calculation. In actual search and rescue scenarios, due to the complex signal propagation environment, multiple azimuth lines obtained from multiple flights or multiple direction findings often do not strictly intersect at a single point. Traditional manual mapping intersection methods are slow, inaccurate, and cannot quickly provide the optimal estimated position; while simple geometric intersection methods cannot effectively utilize the redundant information from multiple measurements, resulting in insufficient reliability of the positioning results. Therefore, there is an urgent need to provide a multi-UAV collaborative search and positioning method and system to solve the above problems. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a multi-UAV collaborative search and positioning method and system. It employs multiple UAVs flying in parallel in a straight line, resulting in a fast response. By combining weighted least squares and three-level data filtering, it achieves high positioning accuracy and strong reliability. UAVs replace manned aircraft, which is low-cost and safe, and realizes fully automated collaboration throughout the entire process.

[0007] To achieve the above objectives, the technical solution of the present invention is as follows:

[0008] A multi-UAV cooperative search and localization method includes:

[0009] At least three drones are ordered to fly simultaneously or nearly simultaneously along their respective predetermined straight-line routes. Each drone is equipped with a rescue radio receiver to listen for radio signals emitted by the target rescue station, and any two straight-line routes form an angle with each other.

[0010] During their flight along their straight flight path, each UAV independently records the coordinates of the starting point of the first received signal and the coordinates of the ending point of the last received signal when it receives the radio signal.

[0011] Each drone will transmit the recorded starting and ending point coordinates back to the data processing center;

[0012] The data processing center generates a signal azimuth line for each UAV based on the received sets of start and end points;

[0013] The data processing center counts the total number of signal azimuth lines, K; if K < 3, it continues to receive new data transmitted back by the UAV; if K ≥ 3, it uses the weighted least squares method to solve for the coordinate point, minimizing the sum of the squared weighted distances from the coordinate point to each signal azimuth line, and the resulting coordinates are the estimated position of the rescue platform target.

[0014] As a preferred embodiment of the present invention, the included angle between any two straight routes is set to 30° to 150°.

[0015] The 30° to 150° range ensures that the azimuth line forms a good intersection geometry, minimizing positioning error.

[0016] As a preferred embodiment of the present invention, the UAV monitors the strength or signal-to-noise ratio of the received signal in real time during its flight along the assigned straight flight path. When the signal strength first exceeds a preset detection threshold, the UAV records its real-time coordinates as the signal start point coordinates. When the signal strength drops below the detection threshold and the duration reaches a preset stabilization time threshold, the UAV records its real-time coordinates as the signal end point coordinates. The preset stabilization time threshold is 3-5 seconds.

[0017] The first time the detection threshold is exceeded, it corresponds to the boundary of the rescue beacon signal coverage area where the drone enters, and the starting point can be accurately marked. Setting a 3-5 second stabilization time threshold is to filter out signal strength fluctuations caused by multipath fading and instantaneous interference, avoid misjudging the signal termination time, and ensure that the termination point coordinates can truly reflect the location where the drone flies out of the signal coverage area.

[0018] As a preferred embodiment of the present invention, before each UAV transmits the recorded starting point coordinates and ending point coordinates back to the data processing center, each UAV calculates the straight-line distance between its recorded starting point and ending point. Only when the straight-line distance is greater than a preset minimum effective distance threshold is the starting point and ending point transmitted back to the data processing center.

[0019] If the distance between the starting and ending points is too short, it indicates that the drone only grazed the edge of the signal coverage area. The directional error of the line connecting the two points will increase exponentially, resulting in extremely low reliability of the generated azimuth line. By pre-screening using a minimum effective distance threshold, invalid data can be eliminated on the airborne end in advance, reducing communication bandwidth usage and preventing low-quality data from interfering with backend processing.

[0020] As a preferred embodiment of the present invention, after the data processing center generates all signal azimuth lines based on the received start points and end points, it first calculates the mutual angles between all generated signal azimuth lines. If the average angle between a certain azimuth line and the other azimuth lines is less than a preset first angle threshold, the azimuth line is determined to be a redundant azimuth line and is removed. Then, the total number K of the remaining valid signal azimuth lines is counted.

[0021] Redundant azimuth lines that are parallel or nearly parallel do not provide additional geometric constraint information; instead, they increase the computational cost of the solution and may lead to a rank deficiency in the solution matrix. Removing redundant azimuth lines with excessively small average included angles preserves valid observations with independent geometric constraints, improving the stability and convergence speed of the solution.

[0022] As a preferred embodiment of the present invention, when the number of remaining valid azimuth lines after redundancy removal is ≥4, the RANSAC random sampling consensus algorithm is used to perform consistency verification on all remaining valid azimuth lines, and abnormal azimuth lines determined to be external points are removed, while only internal point azimuth lines are retained to participate in subsequent position calculation.

[0023] The RANSAC algorithm can robustly estimate model parameters from datasets containing a large number of outliers. In search and rescue scenarios, outliers are mainly composed of erroneous azimuth lines caused by signal obstruction, multipath effects, and electromagnetic interference. When the number of valid azimuth lines is ≥4, a statistical basis for distinguishing inliers from outliers is established. By using RANSAC to remove outliers, the reliability of positioning results in harsh electromagnetic environments can be significantly improved.

[0024] As a preferred embodiment of the present invention, the weighted least squares method for solving the coordinate points includes:

[0025] Each signal azimuth line is uniformly converted into a standardized straight line general expression, and the coefficients are normalized so that the absolute value of the result obtained after substituting any coordinate represents the vertical distance from the arbitrary coordinate to the corresponding signal azimuth line.

[0026] Assign weights to each azimuth line based on the signal reception duration and / or signal quality corresponding to each azimuth line;

[0027] Construct a weighted sum of squared distance objective function, which takes the coordinates of the target location as variables and represents the sum of the weighted squared distances from the point to be found to all azimuth lines;

[0028] By taking the partial derivatives of the objective function with respect to the two coordinate directions and setting the partial derivatives to zero, we obtain the necessary condition for the objective function to reach its minimum value.

[0029] The above conditions are organized into a system of two linear equations. By directly solving this system of equations analytically, the estimated target position coordinates that minimize the weighted sum of squared distances can be obtained.

[0030] Coefficient normalization eliminates the distance measurement bias caused by scale differences in different azimuth equations, ensuring the comparability of the distance contributions of each azimuth. The weighted sum of squared distance objective function is a convex quadratic function with respect to coordinates, which has a unique global minimum. By taking the partial derivative and setting it to zero, it can be transformed into a system of linear equations, enabling fast analytical solutions and avoiding the computational overhead and local optima problems of iterative algorithms.

[0031] As a preferred embodiment of the present invention, the weight of each signal azimuth line is set according to the signal reception duration recorded by the corresponding UAV, the average signal strength, or the product of the two.

[0032] The longer the signal reception duration, the more complete the path the UAV takes through the signal coverage area, and the smaller the azimuth error between the two points. Higher average signal strength results in lower signal interference and higher accuracy in recording the coordinates of the starting and ending points. Dynamically allocating weights based on these factors allows highly reliable observation data to dominate the calculation, effectively suppressing the negative impact of low-quality data.

[0033] As a preferred embodiment of the present invention, the UAV is any one of a fixed-wing UAV, a multi-rotor UAV, or a vertical take-off and landing UAV; the radio signal emitted by the target rescue platform is a 121.5MHz signal in the VHF band or a 243MHz signal in the UHF band.

[0034] Different types of drones are suitable for different search and rescue scenarios: fixed-wing drones have long endurance and high speed, making them suitable for rapid searches over large areas; multi-rotor drones are highly maneuverable, suitable for complex terrain and precise searches in small areas; vertical takeoff and landing drones combine the advantages of both. 121.5MHz and 243MHz are the standard frequency bands for global general aviation emergency location beacons as stipulated by the International Civil Aviation Organization (ICAO), compatible with the ELT equipment carried by all existing crashed aircraft, ensuring the system's universality and interoperability.

[0035] A multi-UAV cooperative search and positioning system includes:

[0036] At least three search and rescue drones, each equipped with an onboard rescue radio receiver module, a flight status perception module, and a data acquisition and preprocessing module; wherein, the data acquisition and preprocessing module is used to independently extract the coordinates of the start and end points of the received radio signals during the drone's flight along a straight flight path, and form a position data packet;

[0037] The data processing center communicates with all UAVs and includes an azimuth line generation module, a quantity judgment module, and a weighted least squares solution module. The azimuth line generation module receives position data packets transmitted from each UAV and generates signal azimuth lines based on each set of start and end points. The quantity judgment module determines whether the number of signal azimuth lines meets a preset requirement. The weighted least squares solution module calculates the estimated position of the target rescue platform using the weighted least squares method when the preset requirement is met.

[0038] The data processing center also includes an azimuth line filtering unit, used to perform redundant azimuth line removal and abnormal azimuth line removal based on a random sampling consensus algorithm; the data processing center also includes a weight calculation unit, used to calculate the weight for each signal azimuth line based on the signal reception duration, average signal strength, or a combination thereof; the data processing center also includes a positioning result distribution module, used to send the calculated target location information to each participating UAV via a data link.

[0039] The beneficial technical effects of this invention are:

[0040] By having at least three drones fly simultaneously or nearly simultaneously along straight routes that form an angle between each other, a distributed parallel acquisition system is formed in space. A single coordinated flight can obtain a sufficient number of signal azimuth lines, completely changing the traditional operational mode of multiple serial flights by a single drone. This method reduces search and rescue response time from several hours to tens of minutes, significantly improving the timeliness of aerial search and rescue.

[0041] By employing weighted least squares to fuse multiple sets of azimuth data, random errors from single-shot direction finding can be effectively suppressed, fully utilizing redundant measurement information from multiple drones. As the number of participating drones increases, the positioning error can converge to within 50 meters, significantly outperforming traditional manual mapping or simple geometric intersection methods. Furthermore, by setting the included angles of each drone's flight path between 30° and 150°, the spatial intersection geometry is optimized, further ensuring positioning accuracy.

[0042] The system performs self-validation of the start and end point data of a single UAV by setting a preset minimum effective distance threshold; redundant azimuth lines are eliminated by calculating the angle between azimuth lines; and when there are sufficient azimuth lines, the RANSAC random sampling consensus algorithm is further used to eliminate outliers. These mechanisms effectively address complex situations such as poor signal reception, packet loss, or abnormal interference from some UAVs, ensuring the reliability of positioning results in harsh environments.

[0043] Replacing traditional manned aircraft with fixed-wing, multi-rotor, or vertical takeoff and landing (VTOL) drones in search and rescue missions significantly reduces the risk and manpower costs of search and rescue operations. The system can be widely deployed on various aerial search and rescue platforms, is compatible with VHF / UHF standard rescue radio signals, and possesses excellent versatility and engineering practicality.

[0044] Each drone independently performs signal detection, automatic recording of start and end points, and data preprocessing. The data processing center centrally completes azimuth line generation, quality screening, weighted calculation, and result distribution. It also transmits the target location information back to each participating drone in real time via data link, guiding them to converge and confirm the target location. This achieves fully automated collaboration from data collection and processing to guiding rescue efforts. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the overall process of the present invention.

[0046] Figure 2 This is a geometric diagram of three drones flying together to search and locate a target. Detailed Implementation

[0047] In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, the specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0048] Combination Figure 1 The present invention provides the following embodiments:

[0049] Example 1:

[0050] A collaborative search and localization method for multiple unmanned aerial vehicles (UAVs) includes the following steps:

[0051] Step 1: Have at least three drones fly simultaneously or nearly simultaneously along their respective predetermined straight routes that form an angle with each other. Each drone is equipped with a rescue radio receiver to listen for radio signals emitted by the target rescue station.

[0052] Based on the information of the area to be searched, the ground control station performs payload and route planning for multiple UAVs participating in the mission, generates corresponding mission planning files, and loads them onto the corresponding UAV platforms.

[0053] Each drone is assigned a predetermined straight-line flight path, with the angle between any two paths set between 30° and 150° to improve positioning accuracy. All drones fly simultaneously at a constant altitude and speed along their assigned straight-line paths within the designated search and rescue airspace. Each drone is equipped with a rescue platform receiver, continuously monitoring radio distress signals emitted by the target rescue platform during flight. Generally, the predetermined straight-line flight path is at least 5 kilometers long to ensure the drones have sufficient range to completely traverse the signal coverage area; the flight altitude is set between 300 meters and 3000 meters to achieve effective line-of-sight reception of radio signals.

[0054] Step 2: Drone signal monitoring and data recording

[0055] Each drone independently performs the following sub-steps:

[0056] Sub-step 2a: Signal detection and starting point recording. As the UAV flies along its assigned straight flight path, the strength or signal-to-noise ratio of the received signal is calculated in real time. When the signal strength first exceeds a preset detection threshold, it is determined as the start of signal reception, and the UAV's position coordinates obtained by the onboard satellite navigation module at that moment are immediately recorded as the starting point. ,in Number the drone. .

[0057] Sub-step 2b: Continuous signal tracking and termination point recording. After detecting a valid signal, the UAV continuously tracks changes in signal strength. When the signal strength attenuates below the detection threshold and remains below the threshold for a preset stabilization time threshold (e.g., 3 to 5 seconds), it is determined as the signal reception termination moment, and the UAV's position coordinates at this moment are recorded as the termination point. .

[0058] Sub-step 2c: Validity self-verification. Each drone calculates its recorded starting point. and the end point The straight-line distance between .like Greater than the preset minimum effective distance threshold In this example, the minimum effective distance threshold If the distance is set to 500 meters, the record is considered valid, forming a data packet determined by the drone. Otherwise, discard the data set and the drone will continue to fly along the flight path to listen for data.

[0059] Step 3: Centralized Data Aggregation

[0060] Each drone will generate its own valid data packets. The data is transmitted in real time to a designated data processing center via a data link. The data processing center can be deployed at a ground station or a cluster drone; in this example, it is set up at the cluster drone to improve data processing efficiency.

[0061] Step 4: Centralized azimuth line generation and quality screening

[0062] After receiving valid data packets from multiple drones, the data processing center processes the data.

[0063] Sub-step 4a: Azimuth line generation. For each set of valid data [ A straight line is determined by two points, and this line is extended infinitely in both directions to obtain a signal azimuth line. .

[0064] Sub-step 4b: Redundant and Abnormal Azimuth Line Screening. Calculate the angles between all pairs of azimuth lines. If the average angle between a azimuth line and the other azimuth lines is less than a preset first angle threshold (e.g., 15°), the azimuth line is determined to be a redundant azimuth line and is removed. When the number of valid azimuth lines reaches 4 or more, the Random Sampling Consensus (RANSAC) algorithm is used to randomly select 2 azimuth lines each time to find their intersection, calculate the distance of the other azimuth lines to the intersection point, and screen out the set of interior azimuth lines that meet the consistency condition, while removing the exterior azimuth lines.

[0065] Step 5: Weighted least squares intersection solution

[0066] The data processing center uses the weighted least squares method to obtain an estimated target coordinate point. This ensures that the point is connected to all signal azimuth lines selected and retained in step 4. to The weighted sum of squared distances reaches the global minimum. Specifically, this includes:

[0067] Sub-step 5a: Standardize the equation of the line. For the first... A azimuth line, consisting of two points and Calculate the coefficients of the equation of the straight line. Let:

[0068] ;

[0069] ;

[0070] ;

[0071] Normalize the coefficients by setting the normalization factor. The standardized equation is:

[0072] .

[0073] in , , Substitute any point absolute value That is, the distance from that point to the line. The vertical distance.

[0074] Sub-step 5b: Weight assignment. For each azimuth line... Assign a weight The weight is determined based on one or more of the following factors in combination.

[0075] Signal duration weight Weight The duration of signal reception recorded by the corresponding drone Proportional, that is ∝ . The longer the azimuth line, the more complete the drone's traverse of the signal coverage area, and the more reliable the azimuth line's direction estimation.

[0076] Signal quality weight Weights and the average signal-to-noise ratio of the corresponding UAV during signal reception. or average received signal strength indicator Proportional.

[0077] Composite weight: = × That is, the product of the duration weight and the quality weight.

[0078] Sub-step 5c: Construction and solution of the objective function. Construct the weighted sum of squared distance objective function:

[0079] ;

[0080] in It's about variables. and The global minimum point of the function is the least squares intersection point we are looking for. Find the values ​​of the function with respect to the least squares intersection point. and Take the partial derivatives and set them to zero:

[0081] =0;

[0082] =0.

[0083] The system of 2×2 linear equations is obtained by rearranging the equations.

[0084]

[0085] The calculation result can be obtained by solving this equation.

[0086] Step 6: Distribution and guidance of location results

[0087] The data processing center will calculate the estimated location of the target. The data is distributed to all participating drones via data link, guiding them to the estimated location for target confirmation. Simultaneously, the information is transmitted back to ground command to guide subsequent rescue operations.

[0088] Example 2:

[0089] A multi-UAV cooperative search and positioning system includes:

[0090] At least three search and rescue drones, each equipped with: (1) an airborne rescue radio receiver module, used to listen to and receive radio distress signals from the target rescue station in the VHF / UHF band, and output demodulated audio and signal strength indication data; (2) a flight status perception module, including at least a global satellite navigation system receiver and an inertial measurement unit, used to acquire the drone's position coordinates, speed and heading information in real time; (3) an airborne data acquisition and preprocessing module, connected to the receiver module and the flight status perception module respectively, used to perform signal detection threshold judgment and validity self-verification, automatically extract the coordinates of the signal start point and end point, form and send local azimuth data packets through the airborne communication module.

[0091] It also includes a data processing center, which is deployed on one of the UAVs (leader aircraft) or a ground control station to maintain communication with all participating UAVs. The data processing center includes the following modules: (1) a multi-aircraft data aggregation module, used to receive valid azimuth data packets transmitted back by each UAV; (2) an azimuth line generation and filtering module, used to generate signal azimuth lines and perform redundant and abnormal azimuth line removal; (3) a weight calculation module, used to calculate the weight for each azimuth line based on the signal duration and / or signal quality parameters; (4) a weighted least squares solution module, used to read the azimuth line equation and weight, construct and solve the 2×2 linear equation system, and output the least squares estimate of the target position; (5) a positioning result distribution module, used to package the estimated target position and confidence area information and send it to each search and rescue unit through a data link.

[0092] Example 3:

[0093] This embodiment simulates a maritime search and rescue scenario involving three aircraft working together to locate and rescue a person in distress. The person in distress carries a standard survival radio, which, once activated, continuously transmits a distress alarm signal at a frequency of 121.5MHz.

[0094] Step 1: Multi-UAV Collaborative Task Planning

[0095] The ground control station planned and loaded coordinated search and rescue routes for three search and rescue drones, designated UAV-1, UAV-2, and UAV-3. The parameters of the planned straight-line routes for the three drones are shown in Table 1. The angle between any two of the three routes is greater than 30°. All drones flew at an altitude of 1000 meters and a speed of 30 meters per second.

[0096] Table 1. Parameters of the planned straight-line flight path for the three UAVs

[0097] Step 2: Drone signal monitoring and data recording

[0098] Three drones took off simultaneously and flew along their respective routes.

[0099] UAV-1 record: at position P 11 A valid signal was first detected at (4.50, 2.90) at location P. 12 The signal disappears at (14.50, 4.90). Calculation shows D1 = 10.20km > 0.5km, so the data is valid. Data packet: [(4.50, 2.90), (14.50, 4.90)]. Signal duration Δt1 = 340 seconds.

[0100] UAV-2 record: at position P 21 The signal was first detected at (7.00, 11.00) at location P. 22The signal disappears at (13.00, 7.00). Calculation shows D2 = 7.21km > 0.5km, indicating valid data. Data packet: [(7.00, 11.00), (13.00, 7.00)]. Signal duration Δt2 = 240 seconds.

[0101] UAV-3 recording: at position P 31 The signal was first detected at (12.00, 4.00) at location P. 32 The signal disappears at (6.00, 10.00). Calculation shows D3 = 8.49 km > 0.5 km, indicating valid data. Data packet: [(12.00, 4.00), (6.00, 10.00)]. Signal duration Δt3 = 283 seconds.

[0102] Step 3: Centralized Data Aggregation

[0103] The three drones transmit their respective data packets back to the data processing center on UAV-1 (the lead drone) in real time via an inter-drone data link.

[0104] Step 4: Centralized azimuth line generation and quality screening

[0105] The data processing center generates three azimuth lines based on the three sets of data:

[0106] L1 from UAV-1 packets: by P 11 (4.50, 2.90) and P 12 (14.50, 4.90) is determined. The slope k1 = (4.90 - 2.90) / (14.50 - 4.50) = 0.20. The equation is: y - 2.90 = 0.20(x - 4.50), which means: 0.20x - y + 2.00 = 0.

[0107] L2 from UAV-2 packets: by P 21 (7.00, 11.00) and P 22 (13.00, 7.00) is determined. The slope k2 = (7.00 - 11.00) / (13.00 - 7.00) = -0.667. The equation is: y - 11.00 = -0.667(x - 7.00), which is: 0.667x + y - 15.67 = 0.

[0108] L3 from UAV-3 packets: by P 31 (12.00, 4.00) and P 32 (6.00, 10.00) is determined. The slope k3 = (10.00 - 4.00) / (6.00 - 12.00) = -1.00. The equation is: y - 4.00 = -1.00(x - 12.00), which means: 1.00x + y - 16.00 = 0.

[0109] Step 5: Weighted least squares intersection solution

[0110] First, the equations of the three azimuth lines are standardized and weighted. In this embodiment, the signal duration ratio is used as the weighting benchmark, as shown in Table 2: Table 2 Weight Allocation of Three Drones

[0111] Substitute the parameters into the system of linear equations ,get: = Therefore, the calculation yields: = .

[0112] Step 6: Distribution and guidance of location results

[0113] The UAV-1 distributed the calculated target's estimated location (8.999, 11.00) to the other two types of UAVs via data link to guide target confirmation. Simultaneously, the information was transmitted back to ground command to guide subsequent rescue operations.

[0114] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multi-UAV cooperative search and positioning method, characterized in that, include: At least three drones are ordered to fly simultaneously or nearly simultaneously along their respective predetermined straight-line routes. Each drone is equipped with a rescue radio receiver to listen for radio signals emitted by the target rescue station, and any two straight-line routes form an angle between each other. During their flight along their straight flight path, each UAV independently records the coordinates of the starting point of the first received signal and the coordinates of the ending point of the last received signal when it receives the radio signal. Each drone will transmit the recorded starting and ending point coordinates back to the data processing center; The data processing center generates a signal azimuth line for each UAV based on the received sets of start and end points; The data processing center counts the total number K of signal azimuth lines. If K < 3, continue to receive new data transmitted back by the drone; if K ≥ 3, use the weighted least squares method to solve for the coordinate point, so that the sum of the squared weighted distances from the coordinate point to each signal azimuth line is minimized, and the obtained coordinates are the estimated position of the rescue platform target.

2. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, The angle between any two straight routes is set to 30° to 150°.

3. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, During the flight of the UAV along the assigned straight flight path, the strength or signal-to-noise ratio of the received signal is monitored in real time. When the signal strength first exceeds the preset detection threshold, the real-time coordinates of the UAV are recorded as the coordinates of the starting point of the signal. When the signal strength drops below the detection threshold and the duration reaches the preset stabilization time threshold, the real-time coordinates of the UAV are recorded as the coordinates of the ending point of the signal. The preset stabilization time threshold is 3-5 seconds.

4. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, Before each drone transmits the recorded starting and ending point coordinates back to the data processing center, each drone calculates the straight-line distance between its recorded starting and ending points. Only when this straight-line distance is greater than a preset minimum effective distance threshold will the starting and ending points be transmitted back to the data processing center.

5. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, After generating all signal azimuth lines based on the received start and end points, the data processing center first calculates the mutual angles between all generated signal azimuth lines. If the average angle between a certain azimuth line and the other azimuth lines is less than a preset first angle threshold, the azimuth line is determined to be a redundant azimuth line and is removed. Then, the total number K of the remaining valid signal azimuth lines is counted.

6. The multi-UAV cooperative search and positioning method according to claim 5, characterized in that, When the number of valid azimuth lines remaining after redundancy removal is ≥4, the RANSAC random sampling consensus algorithm is used to perform consistency verification on all remaining valid azimuth lines, and abnormal azimuth lines that are determined to be outside points are removed, and only the azimuth lines of inside points are retained to participate in subsequent position calculation.

7. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, The weighted least squares method for solving coordinate points includes: Each signal azimuth line is uniformly converted into a standardized straight line general expression, and the coefficients are normalized so that the absolute value of the result obtained after substituting any coordinate represents the vertical distance from the arbitrary coordinate to the corresponding signal azimuth line. Assign weights to each azimuth line based on the signal reception duration and / or signal quality corresponding to each azimuth line; Construct a weighted sum of squared distance objective function, which takes the coordinates of the target location as variables and represents the sum of the weighted squared distances from the point to be found to all azimuth lines; By taking the partial derivatives of the objective function with respect to the two coordinate directions and setting the partial derivatives to zero, we obtain the necessary condition for the objective function to reach its minimum value. The above conditions are organized into a system of two linear equations. By directly solving this system of equations analytically, the estimated target position coordinates that minimize the weighted sum of squared distances can be obtained.

8. The multi-UAV cooperative search and positioning method according to claim 7, characterized in that, The weight of each signal azimuth line is set based on the signal reception duration recorded by the corresponding UAV, the average signal strength, or the product of the two.

9. The multi-UAV cooperative search and positioning method according to claim 1, characterized in that, The drone is any one of a fixed-wing drone, a multi-rotor drone, or a vertical take-off and landing drone; the radio signal emitted by the target rescue platform is a 121.5MHz signal in the VHF band or a 243MHz signal in the UHF band.

10. A multi-UAV cooperative search and positioning system, characterized in that, include: At least three search and rescue drones, each equipped with an onboard rescue radio receiver module, a flight status perception module, and a data acquisition and preprocessing module; wherein, the data acquisition and preprocessing module is used to independently extract the coordinates of the start and end points of the received radio signals during the drone's flight along a straight flight path, and form a position data packet; The data processing center communicates with all UAVs and includes an azimuth line generation module, a quantity judgment module, and a weighted least squares solution module. The azimuth line generation module receives position data packets transmitted from each UAV and generates signal azimuth lines based on each set of start and end points. The quantity judgment module determines whether the number of signal azimuth lines meets a preset requirement. The weighted least squares solution module calculates the estimated position of the target rescue platform using the weighted least squares method when the preset requirement is met. The data processing center also includes an azimuth line filtering unit, used to perform redundant azimuth line removal and abnormal azimuth line removal based on a random sampling consensus algorithm; the data processing center also includes a weight calculation unit, used to calculate the weight for each signal azimuth line based on the signal reception duration, average signal strength, or a combination thereof; the data processing center also includes a positioning result distribution module, used to send the calculated target location information to each participating UAV via a data link.

Images