A target positioning method based on multi-point ranging data of UAV

Abstract

The present invention discloses a target positioning method based on multi-point ranging data from an unmanned aerial vehicle (UAV), with the purpose of improving the positioning accuracy of the UAV positioning method. The technical solution is to construct a target positioning system consisting of a data acquisition module, a coarse position estimation module, and a fine position estimation module. The data acquisition module constructs a data set (Data) based on the data acquisition accuracy specified in the UAV user manual and data collected by the UAV laser payload and the UAV navigation system. The coarse position estimation module reads Data and uses an optimal linear unbiased estimation method to obtain a first estimate #imgabs0# containing the target position and a weighted least squares estimation method to obtain a second estimate #imgabs1# containing the target position. The fine position estimation module constructs a parameter β to be estimated based on the relationship between the elements within #imgabs2# and uses a weighted least squares estimation method to accurately locate the target. The present invention can obtain high-precision target positioning results and a high-precision three-dimensional position of the target.

Description

Technical Field

[0001] The present invention relates to the technical field of target positioning, and in particular to a target positioning method based on multi-point ranging data of an unmanned aerial vehicle. Background Art

[0002] Drones are widely used in military and civilian fields due to their small size, maneuverability, strong concealment, and ability to carry multiple sensors, and their potential to accurately detect target locations.

[0003] The UAV positioning target generally adopts the classic single point positioning method. The specific steps are as follows:

[0004] In the first step, after the operator controls the drone's optoelectronic system to lock and track the ground target, the drone's data acquisition module obtains the drone's position from the drone's navigation system, the drone's attitude from the drone's gyroscope system, and the target's azimuth and pitch angle relative to the drone, as well as the distance between the drone and the target, from the drone's optoelectronic system.

[0005] In the second step, the positioning software on the drone reads the drone's position, drone attitude, azimuth and pitch angle of the target relative to the drone, and the distance from the drone to the target from the data acquisition module. Then, the relative position of the drone and the target and the relationship between the distance and angle between the two are used to establish an expression for estimating the target position and solve the target position.

[0006] Single-point positioning uses angle and distance information to calculate the target position. Its principle is simple and easy to implement. However, this method places high demands on the attitude measurement accuracy of the drone's gyroscope system and the angle measurement accuracy of the optoelectronic system. High-precision gyroscopes are too large to be installed on drone platforms, while miniaturized gyroscopes exhibit significant cumulative errors in drone attitude measurement, which gradually increase with measurement time. Furthermore, the mechanically rotating optoelectronic system on the drone is subject to installation errors, making it impossible to obtain accurate pointing angle information. This results in large single-point positioning errors on existing drone platforms. For targets 1 kilometer away, the positioning error can even reach hundreds of meters, making it impossible to effectively estimate the position of ground targets.

[0007] In order to overcome the problem of low target positioning accuracy caused by inaccurate measurement of UAV attitude and pointing angle, a total least squares method for positioning the target using multi-point ranging information of UAV was proposed (Zheng Kai, Zheng Xianmin, Yin Shaofeng, et al. Real-time target positioning method of UAV based on total least squares [J]. Electro-Optics and Control, 2019 26(10):26-29.).

[0008] The total least squares method is:

[0009] In the first step, the mission operator controls the drone to lock onto the target and fly around it. At several measurement points along the orbital trajectory, the drone's data acquisition module obtains the drone's position from the drone's navigation system and the distance to the target from the optoelectronic system.

[0010] In the second step, the positioning software on the drone reads the drone position and the distance from the drone to the target in the data acquisition module. Based on the relationship between the relative position of the drone and the target and the distance between the two, an expression for estimating the target position is established based on the least squares principle, and the singular value decomposition method is used to solve the expression for the target position to obtain the target position.

[0011] Compared to single-point positioning, the total least squares positioning method does not require angle measurement information, thus avoiding the problem of low positioning accuracy caused by large angle measurement errors, and can achieve target positioning accuracy superior to that of single-point positioning. However, the total least squares positioning method relies on the UAV's flight trajectory and can only obtain good positioning results under cylindrical spiral and Archimedean spiral trajectories. For targets 1 kilometer away, the positioning accuracy can reach the meter level. However, for typical flight trajectories such as flying around the target's circumference, the positioning error is large at the target's altitude, and the positioning accuracy can even drop to the hundred-meter level. This makes it impossible to effectively obtain the target's three-dimensional spatial position and cannot meet the requirements of high-precision target indication.

[0012] Therefore, how to improve the accuracy of UAV target positioning has always been a difficult and hot issue in this field. Summary of the Invention

[0013] The technical problem to be solved by the present invention is that, in response to the problem that the existing UAV positioning method has low positioning accuracy, a new target positioning method based on UAV multi-point ranging data is proposed to achieve high-precision determination of the target's three-dimensional spatial position and improve the target positioning accuracy.

[0014] To solve the above technical problems, the technical solution of the present invention is to construct a target positioning system consisting of a data acquisition module, a rough position estimation module, and a fine position estimation module. The data acquisition module constructs a data set Data based on the data acquisition accuracy specified in the drone manual and the data collected by the drone laser payload and the drone navigation system; the rough position estimation module reads the Data and uses the optimal linear unbiased estimation method to obtain the first estimated value containing the target position. And use the weighted least squares estimation method to get the second estimated value of the target position The position estimation module is based on The internal element relationship is used to construct the parameter β to be estimated, and the weighted least squares estimation method is used to accurately locate the target. The present invention can obtain a high-precision target positioning result.

[0015] In the mathematical formulas of the present invention, adding "T" to the upper right corner of a matrix or vector indicates transposing the matrix or vector, and adding "-1" to the upper right corner of a matrix indicates inverting the matrix.

[0016] The present invention comprises the following steps:

[0017] The first step is to build a target positioning system, which consists of a data acquisition module, a coarse position estimation module, and a fine position estimation module. The data acquisition module is built into the drone, while the coarse position estimation module and the fine position estimation module are installed on the drone's computing unit.

[0018] The data acquisition module, connected to the coarse position estimation module, consists of a UAV laser payload and a UAV navigation system. The UAV laser payload is required to measure distances of at least 10 kilometers, with ranging accuracy better than 9 meters; the UAV navigation system requires positioning accuracy better than 5 meters. The UAV laser payload collects the distances to the target from N measurement points as the UAV orbits the target; the UAV navigation system collects the UAV's three-dimensional position from N measurement points as the UAV orbits the target, thereby generating the following dataset:

[0019]

[0020] in is the distance set from N measurement points to the target during the UAV’s flight around the target. The distance from the nth measurement point to the target during the UAV's flight around the target, in meters, 1≤n≤N, where N is the number of measurement points in the UAV's flight path around the target; is the three-dimensional position of the UAV at N measurement points during the UAV’s flight around the target, is the three-dimensional position of the UAV at the nth measurement point during the UAV’s flight around the target, is the x-axis position of the nth measurement point in the WGS84 coordinate system, is the y-axis position of the nth measurement point in the WGS84 coordinate system, is the z-axis position of the nth measurement point in the WGS84 coordinate system. The unit of the three-dimensional position of the drone is meter. u is the UAV navigation accuracy, which indicates the measurement accuracy of the UAV position, in square meters; Q R Q is the UAV laser ranging accuracy, which indicates the ranging accuracy of the UAV laser payload, in square meters. u and Q R Both can be obtained through the drone user manual.

[0021] The rough position estimation module is connected to the data acquisition module and the precise position estimation module. The rough position estimation module reads the UAV multi-point ranging data set Data from the UAV data acquisition module, estimates the target position using the optimal linear unbiased estimation method and the weighted least squares method, obtains the initial estimated value of the target position, and sends the initial estimated value to the precise position estimation module.

[0022] The precise position estimation module is connected to the coarse position estimation module, receives the initial estimated value of the target position from the coarse position estimation module, constructs an expression for precisely estimating the target position based on the initial estimated value, and then uses the weighted least squares estimation method to solve the target position to obtain the final target positioning result.

[0023] In the second step, the data acquisition module constructs the dataset Data based on the data acquisition accuracy in the UAV manual and the data collected by the UAV laser payload and the UAV navigation system. The method is as follows:

[0024] 2.1 According to the UAV navigation accuracy Q in the UAV user manual u and laser ranging accuracy Q R , put them into the data set Data in turn, and get the first two items in Data.

[0025] 2.2 Collection and The method is:

[0026] 2.2.1 Initialize n = 1;

[0027] 2.2.2 During the UAV's flight around the target, the UAV's laser payload obtains the distance from the UAV to the target at the nth measurement point (1≤n≤N, N is the number of measurement points of the UAV's flight path around the target, N>3) The UAV navigation system obtains the position of the UAV at the nth measurement point Will Put In Put middle.

[0028] 2.2.3 If n≤N, go to 2.2.2; if n>N, it means it is completed and The collection will and Put it into the data set Data.

[0029] In the third step, the position rough estimation module reads the data from the data acquisition module and uses the optimal linear unbiased estimation method to obtain the first estimated value of the target position. On this basis, the weighted least squares estimation method is further used to obtain the second estimated value of the target position. and will Send to the position estimation module.

[0030] Here’s how:

[0031] 3.1: Arrange the data to generate the first vector b and the first coefficient matrix A.

[0032] b=[b1,…,b n ,…,b N ] T ,in

[0033] in

[0034] 3.2: According to the Gauss-Markov theorem, using the relationship between the target position and the distance between the UAV and the target, based on the optimal linear unbiased estimation method (see the literature "SM Kay, Fundamentals of Statistical Signal Processing [M]. Prentice Hall PTR, 1993." SM Kay's monograph: Fundamentals of Statistical Signal Processing) the first vector b and the first coefficient matrix A are derived to obtain formula (4), and the estimated value of the first target position is calculated.

[0035]

[0036] Where W is the covariance matrix, W is generated using Data,

[0037]

[0038] in Is a diagonal matrix, the N elements on the main diagonal are vectors N distance values ​​in ; I N Represents the identity matrix of size N×N; represents a three-dimensional vector whose elements are all 1; Is a diagonal matrix, the elements on the main diagonal are vectors The coordinate values ​​in are composed of; Represents the Kronecker product mathematical operation.

[0039] 3.3: After getting Based on the weighted least squares estimation method, formula (6) is obtained, and the estimated value of the second target position is calculated using formula (6)

[0040]

[0041] Where Ξ is the weight matrix, Ξ is the use of Data and generated,

[0042]

[0043] in Represents a vector The vector consisting of the first three elements in . It is a 4-dimensional vector, the first to third elements are the target 3D positions, and the fourth element is the mean square sum of the target 3D positions.

[0044] 3.4: The rough position estimation module will Send to the position estimation module.

[0045] The fourth step is that the position estimation module receives the based on The internal element relationship is used to construct the parameter β to be estimated, and then the weighted least squares estimation method is used to accurately locate the target. The method is as follows:

[0046] 4.1: Utilization Generate the second vector q and the second coefficient matrix H,

[0047]

[0048]

[0049] in express The vector consisting of the first to third elements in , express The fourth element in α=[x α ,y α ,z α ] T is the auxiliary variable introduced, where Indicates that The largest express The first element in; Indicates that The largest Indicates quantity The second element in Indicates that The largest express The third element in .

[0050] 4.2: Using the weighted least squares estimation method to obtain the estimation formula (10), the estimated value of the third target position is calculated according to formula (10)

[0051]

[0052] Where M is the weight matrix, and its calculation expression is

[0053] M=T(A T Ξ -1A ) -1 T T (11)

[0054] in

[0055] 4.3: Using Vectors Through the matrix operation of formula (12), we can get the target three-dimensional position estimation value for

[0056]

[0057] where sgn(·) is a sign function that satisfies

[0058] is the estimated value of the target horizontal coordinate in the WGS84 coordinate system, is the estimated value of the target vertical coordinate, Estimated value of target height coordinate.

[0059] Compared with the prior art, the present invention can achieve the following beneficial effects:

[0060] (1) The present invention adopts a two-step positioning method of coarse positioning + fine positioning. Through the coarse-to-fine positioning strategy, it overcomes the problem of poor positioning effect of the direct target position solution method caused by the high-order nonlinear relationship between the distance measurement value and the target position, and realizes high-precision estimation of the target position.

[0061] (2) The third step of the present invention is to use the rough position estimation module to update the weighted matrix in the weighted least squares estimation method through the preliminary estimated target position value, so that the rough position estimation module can make more effective use of the valuable information in the measurement data, so that the rough position estimation module can obtain a more accurate target position estimate than using only the optimal linear unbiased estimation method.

[0062] (3) The fourth step position estimation module of the present invention reduces the sensitivity of the coefficient matrix in the weighted least squares estimation method to the UAV flight trajectory by introducing auxiliary variables, and can estimate the three-dimensional position of the target with high precision under the trajectory of the UAV flying around the target circle. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is the overall flow chart of the present invention;

[0064] Figure 2 This is the logical structure diagram of the target positioning system constructed in the first step of the present invention;

[0065] Figure 3 This is a scene diagram of a drone positioning target in a simulation example of the present invention. Figure 3 (a) is a top view of the positioning scene, Figure 3 (b) is a three-dimensional image of the positioning scene;

[0066] Figure 4 In order to use the total least squares method of the present invention and the background technology to Figure 3 Comparison of positioning results for the simulation example shown. DETAILED DESCRIPTION

[0067] The present invention will be further described in detail below with reference to the accompanying drawings and specific simulation examples.

[0068] Figure 1 This is the overall flow chart of the present invention, which includes four steps.

[0069] The first step is to build a target positioning system. The target positioning system consists of a drone data acquisition module, a rough position estimation module, and a precise position estimation module. The data acquisition module is built into the drone, and the rough position estimation module and the precise position estimation module are installed on the drone computing unit. The logical structure of the target positioning system is shown in the figure below. Figure 2 shown.

[0070] The data acquisition module, connected to the coarse position estimation module, consists of a UAV laser payload and a UAV navigation system. The UAV laser payload is required to measure distances of at least 10 kilometers, with ranging accuracy better than 9 meters; the UAV navigation system requires positioning accuracy better than 5 meters. The UAV laser payload collects the distances to the target from N measurement points as the UAV orbits the target; the UAV navigation system collects the UAV's three-dimensional position from N measurement points as the UAV orbits the target, thereby generating the following dataset:

[0071]

[0072] in is the distance set from N measurement points to the target during the UAV’s flight around the target. The distance from the nth measurement point to the target during the UAV's flight around the target, in meters, 1≤n≤N, where N is the number of measurement points in the UAV's flight path around the target; is the three-dimensional position of the UAV at N measurement points during the UAV’s flight around the target, is the three-dimensional position of the UAV at the nth measurement point during the UAV’s flight around the target, is the x-axis position of the nth measurement point in the WGS84 coordinate system, is the y-axis position of the nth measurement point in the WGS84 coordinate system, is the z-axis position of the nth measurement point in the WGS84 coordinate system. The unit of the three-dimensional position of the drone is meter. u is the UAV navigation accuracy, which indicates the measurement accuracy of the UAV position, in square meters; Q R Q is the UAV laser ranging accuracy, which indicates the ranging accuracy of the UAV laser payload, in square meters. u and Q R Both can be obtained through the drone user manual.

[0073] The rough position estimation module is connected to the UAV data acquisition module and the precise position estimation module. After reading the UAV multi-point ranging data set Data from the UAV data acquisition module, the rough position estimation module uses the optimal linear unbiased estimation method and the weighted least squares method to estimate the target position, obtains the initial estimated value of the target position, and sends the initial estimated value to the precise position estimation module.

[0074] The precise position estimation module is connected to the coarse position estimation module, receives the initial estimated value of the target position from the coarse position estimation module, constructs an expression for accurately estimating the target position based on the initial estimated value, and then uses the weighted least squares estimation method to perform high-precision positioning of the target to obtain the final target position estimate.

[0075] In the second step, the drone data acquisition module constructs the dataset Data based on the data acquisition accuracy in the drone manual and the data collected by the drone laser payload and the drone navigation system. The method is as follows:

[0076] 2.1 According to the UAV navigation accuracy Q in the UAV user manual u and laser ranging accuracy Q R , put them into the data set Data in turn, and get the first two items in Data.

[0077] 2.2 The operator controls the drone to fly around the target and collect data and The flight trajectory is as follows Figure 3 As shown, Figure 3 This is a positioning scene diagram of a simulation example of the present invention, simulating the positioning process of the active KVB series UAV (i.e., short-range reconnaissance UAV). Figure 3 (a) is a top view of the positioning scene, Figure 3 (b) is a 3D image of the positioning scenario. The drone's flight trajectory is obtained by the operator maneuvering the drone to fly in a circle centered on the target O. Meanwhile, the drone's flight position fluctuates during flight due to airflow. and The collection method is:

[0078] 2.2.1 Initialize n = 1;

[0079] 2.2.2 The laser payload of the UAV obtains the distance from the UAV to the target at the nth measurement point (1≤n≤N, N is the number of measurement points of the UAV's flight path around the target. In this simulation example, N=30) The UAV navigation system obtains the position of the UAV at the nth measurement point Will Put In Put middle.

[0080] 2.2.3 If n≤N, go to 2.2.2; if n>N, it means it is completed and The collection will and Put it into the data set Data.

[0081] In the third step, the position rough estimation module reads the data from the drone data acquisition module and uses the optimal linear unbiased estimation method to obtain the first estimated value of the target position. On this basis, the weighted least squares estimation method is further used to obtain the second estimated value of the target position. and will Send it to the position estimation module. The method is as follows:

[0082] 3.1: Arrange the data to generate the first vector b and the first coefficient matrix A.

[0083] b=[b1,…,b n ,…,b N ] T ,in

[0084] in

[0085] 3.2: According to the Gauss-Markov theorem, using the relationship between the target position and the distance between the drone and the target, the first vector b and the first coefficient matrix A are derived based on the optimal linear unbiased estimation method to obtain formula (4) to calculate the estimated value of the first target position.

[0086]

[0087] Where W is the covariance matrix, W is generated using Data,

[0088]

[0089] in Is a diagonal matrix, the N elements on the main diagonal are vectors N distance values ​​in ; I N Represents the identity matrix of size N×N; represents a three-dimensional vector whose elements are all 1; Is a diagonal matrix, the elements on the main diagonal are vectors The coordinate values ​​in are composed of; Represents the Kronecker product mathematical operation.

[0090] 3.3: After getting the estimated value Based on the weighted least squares estimation method, formula (6) is obtained, and the estimated value of the second target position is calculated using formula (6)

[0091]

[0092] Where Ξ is the weight matrix, Ξ is the use of Data and generated,

[0093]

[0094] in Represents a vector The vector consisting of the first three elements in .

[0095] 3.4: The rough position estimation module will Send to the position estimation module.

[0096] The fourth step is that the position estimation module receives the based on The internal element relationship is used to construct the parameter β to be estimated, and then the weighted least squares estimation method is used to accurately locate the target. The method is as follows:

[0097] 4.1: Utilization Generate the second vector q and the second coefficient matrix H,

[0098]

[0099]

[0100] in Represents a vector The vector consisting of the first to third elements in , Represents a vector The fourth element in α=[x α ,y α ,z α ] T is the auxiliary variable introduced, where Indicates that The largest in Represents a vector The first element in; Indicates that The largest in Represents a vector The second element in Indicates that The largest in Represents a vector The third element in .

[0101] 4.2: Using the weighted least squares estimation method to obtain the estimation formula (10), the estimated value of the third target position is calculated according to formula (10)

[0102]

[0103] Where M is the weight matrix, and its calculation expression is

[0104]

[0105] in

[0106] 4.3: Using Vectors Through the matrix operation of formula (24), we can get the target three-dimensional position estimation value for

[0107]

[0108] where sgn(·) is a sign function that satisfies

[0109] Contains the estimated horizontal coordinate of the target in the WGS84 coordinate system Target ordinate estimate And the target height coordinate estimate

[0110] Figure 4 The positioning results of the overall minimum method in the present invention and the background technology are shown under different numbers of drone measurement points (i.e., N is different, N is 4, 10, 20, and 30 respectively). The results show the positioning errors of the positioning method in three directions in the WGS84 coordinate system, as well as the total error of the positioning errors in the three directions. The total error is calculated as the square root of the sum of the squares of the positioning errors in the three directions. The following takes the number of drone measurement points N as 4 as an example to explain the data set Data input to the target positioning system. In a certain flight, the values ​​of each element in the data set Data are as follows:

[0111] Q u =25·I 12

[0112] Q R =81·I4

[0113]

[0114]

[0115] The drone position of the first measurement point is The drone position at the second measurement point is The drone position at the third measurement point is The drone position at the fourth measurement point is I 12 and I4 represent identity matrices of size 12×12 and 4×4 respectively. Figure 4 It can be found that when the number of measurement points N is 4, the present invention significantly improves the target positioning accuracy compared to the total least squares method described in the background art, especially reducing the positioning error in the z-axis direction from 337.33m to 8.19m, while the positioning accuracy in the x and y directions is also improved. Figure 4 The positioning results presented for measurement points N of 10, 20, and 30 demonstrate that the present invention significantly improves target positioning accuracy compared to existing total least squares methods, with a particularly significant improvement in the z-axis direction, demonstrating superior positioning accuracy. These results demonstrate that the present invention overcomes the shortcomings of existing positioning methods and achieves high-precision target positioning.

Claims

1. A target positioning method based on multi-point ranging data of unmanned aerial vehicle, characterized in that The following steps are involved: The first step is to build a target positioning system, which consists of a data acquisition module, a rough position estimation module, and a fine position estimation module. The data acquisition module is built into the drone, while the rough position estimation module and the fine position estimation module are installed on the drone's computing unit. The data acquisition module is connected to the rough position estimation module and consists of the UAV laser payload and the UAV navigation system; The UAV laser payload collects the distance from the target to N measurement points during the UAV's flight around the target; the UAV navigation system collects the UAV's three-dimensional position at N measurement points during the UAV's flight around the target; thus, the dataset Data is obtained: in is the distance set from N measurement points to the target during the UAV’s flight around the target. is the distance from the nth measurement point to the target during the UAV's flight around the target, in meters, 1≤n≤N, N is the number of measurement points in the UAV's flight track around the target; u is the three-dimensional position of the UAV at N measurement points during the UAV's flight around the target, is the three-dimensional position of the UAV at the nth measurement point during the UAV’s flight around the target, is the x-axis position of the nth measurement point in the WGS84 coordinate system, is the y-axis position of the nth measurement point in the WGS84 coordinate system, is the z-axis position of the nth measurement point in the WGS84 coordinate system. The unit of the three-dimensional position of the drone is meter. u is the UAV navigation accuracy, which indicates the measurement accuracy of the UAV position, in square meters; Q R Q is the UAV laser ranging accuracy, which indicates the ranging accuracy of the UAV laser payload, in square meters. u and Q R All were obtained through the drone user manual; The rough position estimation module is connected to the data acquisition module and the fine position estimation module. The rough position estimation module reads the UAV multi-point ranging data set Data from the UAV data acquisition module, estimates the target position using the optimal linear unbiased estimation method and the weighted least squares method, obtains the initial estimate of the target position, and sends the initial estimate to the fine position estimation module. The precise position estimation module is connected to the coarse position estimation module, receives the initial estimated value of the target position from the coarse position estimation module, constructs an expression for accurately estimating the target position based on the initial estimated value, and then uses the weighted least squares estimation method to solve the target position to obtain the final target positioning result; In the second step, the data acquisition module constructs the dataset Data based on the data acquisition accuracy in the UAV manual and the data collected by the UAV laser payload and the UAV navigation system. The method is as follows: 2.1 According to the UAV navigation accuracy Q in the UAV user manual u and laser ranging accuracy Q R , put them into the data set Data in turn, and get the first two items in Data; 2.2 UAV laser payload collection UAV navigation system acquisition Will and Put it into the data set Data; In the third step, the position rough estimation module reads the data from the data acquisition module and uses the optimal linear unbiased estimation method to obtain the first estimated value of the target position. And further use the weighted least squares estimation method to obtain the second estimated value of the target position and will Send to the position estimation module; the method is as follows: 3.1: Arrange the data to generate the first vector b and the first coefficient matrix A. In the formula, adding "T" to the upper right corner of the matrix or vector indicates transposing the matrix or vector, and adding "-1" to the upper right corner of the matrix indicates inverting the matrix: b=[b1,···,b n ,···,b N ] T ,in in 3.2: According to the Gauss-Markov theorem, using the relationship between the target position and the distance between the drone and the target, the first vector b and the first coefficient matrix A are derived based on the optimal linear unbiased estimation method to obtain formula (4) to calculate the estimated value of the first target position. Where W is the covariance matrix, W is generated using Data, in Is a diagonal matrix, the N elements on the main diagonal are vectors N distance values ​​in ; I N Represents the identity matrix of size N×N; represents a three-dimensional vector whose elements are all 1; Is a diagonal matrix, the elements on the main diagonal are vectors The coordinate values ​​in are composed of; Represents the Kronecker product mathematical operation; 3.3: After getting Based on the weighted least squares estimation method, formula (6) is obtained, and the estimated value of the second target position is calculated using formula (6) Where Ξ is the weight matrix, Ξ uses Data and generate, in Represents a vector The vector consisting of the first to third elements in ; It is a 4-dimensional vector, the first to third elements are the three-dimensional positions of the target, and the fourth element is the mean square sum of the three-dimensional positions of the target; 3.4: The rough position estimation module will Send to the position estimation module; The fourth step is that the position estimation module receives the based on Internal element relationships construct parameters to be estimated Then the weighted least squares estimation method is used to accurately locate the target; the method is as follows: 4.1: Utilization Generate the second vector q and the second coefficient matrix H, in express The vector consisting of the first to third elements in , express The fourth element in α=[x α ,y α ,z α ] T is the auxiliary variable introduced, where Indicates that The largest express The first element in; Indicates that The largest express The second element in Indicates that The largest express The third element in 4.2: Using the weighted least squares estimation method to obtain the estimation formula (10), the estimated parameters of the third target position are calculated according to formula (10): Where M is the weight matrix, and its calculation expression is M=T(A T Ξ -1 A) -1 T T (11) in 4.3: Utilization Through the matrix operation of formula (12), we can get the target three-dimensional position estimation value where sgn(·) is a sign function that satisfies is the estimated value of the target horizontal coordinate in the WGS84 coordinate system, is the estimated value of the target vertical coordinate, Estimated value of target height coordinate.

2. A target positioning method based on multi-point ranging data of an unmanned aerial vehicle according to claim 1, characterized in that The ranging distance of the UAV laser payload is required to be no less than 10 kilometers, and the ranging accuracy is better than 9 meters; the UAV navigation system requires a positioning accuracy better than 5 meters.

3. A target positioning method based on multi-point ranging data of an unmanned aerial vehicle according to claim 1, characterized in that The number N of measurement points of the track of the UAV flying around the target is greater than 3.

4. A target positioning method based on multi-point ranging data of an unmanned aerial vehicle according to claim 1, characterized in that Acquisition as described in step 2.2 and The method is: 2.2.1 Initialize n = 1; 2.2.2 When the UAV flies around the target, the UAV laser payload obtains the distance from the UAV to the target at the nth measurement point. The UAV navigation system obtains the position of the UAV at the nth measurement point Will Put In Put middle; 2.2.3 If n≤N, go to 2.2.2; if n>N, it means it is completed and The collection will and Put it into the data set Data.

Images