Method for managing position of unmanned aerial vehicle cluster based on array assisted positioning

Abstract

This invention discloses a method for UAV swarm location management based on array-assisted positioning, applied in the field of UAV technology. Addressing the problem of coexistence of positioning and communication signals at the same frequency during UAV swarm location management within GNSS denied areas, resulting in significant mutual interference and the difficulty of single-node positioning tasks in existing technologies, this invention utilizes the assisted positioning system carried by the swarm leader node. By constructing a received signal separation model through array reception, it separates the received time-frequency conflicting positioning and communication signals, eliminating mutual interference and enabling location updates within the UAV swarm. The separated positioning signals are further processed for single-node positioning, ultimately achieving efficient location management within the UAV swarm.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicles (UAVs), and specifically relates to a location update technology for UAV swarms. Background Technology

[0002] With the continuous development of drone technology and the increasing diversification of application scenarios, drones are playing an increasingly important role in fields such as reconnaissance, transportation, and search and rescue. However, a single drone can hardly meet the needs of complex tasks, and drone swarms are needed to work together to complete tasks and improve efficiency. Position management within a drone swarm is an important prerequisite for ensuring the smooth completion of their collaboration.

[0003] Existing methods for UAV swarm location management largely rely on Global Navigation Satellite Systems (GNSS). Guohao Zhang et al., in their paper "Intelligent GNSS / INS integrated navigation system for a commercial UAV flight control system," Aerospace Science and Technology, 2018, 80:368-380, utilized a Kalman filter to combine GNSS with an Inertial Navigation System (INS), eliminating GNSS positioning errors and obtaining precise location information for each node within the UAV swarm. The swarm leader node can periodically communicate with other nodes in the network to collect their location information for swarm management and monitoring. However, in cases where UAV signals are lost within GNSS denial zones, the swarm leader node must re-search and locate the UAV to obtain its relative position. Positioning methods based on Time of Arrival (TOA) and Time Difference of Arrival (TDOA) have been widely studied and applied as important positioning tools. However, these methods require utilizing TOA / TDOA measurements from positioning signals to multiple known receiving nodes, solving for the position coordinates of the node to be positioned through geometric modeling, and demanding precise clock synchronization between the node to be positioned and the receiving nodes. Therefore, these methods are not suitable for single-node scenarios. Furthermore, to conserve frequency domain resources, positioning signals and communication signals coexist on the same frequency, resulting in significant mutual interference. Therefore, potential time-frequency domain resource conflicts between the positioning task and the original communication task of the swarm leader node can severely impact the position update efficiency and positioning accuracy within the swarm. How to efficiently utilize time-frequency domain resources to achieve accurate and real-time position management within a UAV swarm remains a challenge. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a method for managing the location of a UAV swarm based on array-assisted positioning. The swarm leader node periodically receives communication signals from other nodes within the swarm and, within GNSS-denied areas, utilizes the assisted positioning system carried by the swarm leader node to obtain the relative positions of lost UAVs, thereby achieving location updates and management of the UAV swarm.

[0005] The technical solution adopted in this invention is: a drone swarm location management method based on array-assisted positioning. When a drone signal is lost in a GNSS denied area, the swarm leader node re-searches for and locates the drone while receiving communication signals from other nodes. The specific processing includes the following steps:

[0006] S1. Establish the array receiving signal model; specifically: the positioning subsystem and communication subsystem within the group leader node share the same radio frequency hardware channel. When the group leader node activates the auxiliary positioning system to receive positioning signals, a UAV simultaneously sends communication signals to the group leader node; considering that the receiving array is an M×N uniform area array, and the positioning signal and communication signal have the same carrier frequency, the M×N array element received signal, after down-conversion processing, can be represented in the following matrix form:

[0007] x(t) = As(t) + n(t)

[0008] Where A is the array manifold matrix, x(t) is the received signal vector, and n(t) is the M×N noise signal vector;

[0009] S2. Calculate the MUSIC spectrum estimate of the received signal x(t);

[0010] S3. Solve for the unmixing matrix W based on the whitened received signal x(t);

[0011] S4. Use the demixing matrix W to separate the received signal x(t) to obtain the communication signal and the positioning signal;

[0012] S5. Demodulate the separated communication signals to achieve position updates within the UAV cluster; perform pulse compression processing on the separated positioning signals to obtain distance information; combine the MUSIC spectrum estimation from step S2 with the azimuth information of the communication signals to determine the azimuth of the UAV to be located, thereby achieving accurate positioning within the UAV cluster.

[0013] The beneficial effects of this invention are as follows: This invention comprehensively considers the resource conflicts between positioning and communication tasks in UAV swarm location management. It employs a feature matrix joint diagonalization method to separate time-frequency domain conflicting communication signals from positioning signals, eliminating mutual interference between signals. Furthermore, it combines the received signal music spectrum with the separated communication information, enabling a single node to complete the positioning task. This effectively reduces the number of required nodes, avoids waste of time-frequency domain resources, and achieves efficient utilization of communication-sensing positioning resources. The method of this invention has the following advantages:

[0014] 1. The method of the present invention comprehensively considers the potential time-frequency domain resource conflicts between communication positioning tasks and perception positioning tasks in UAV swarm positioning, and uses separation methods to eliminate the mutual interference between simultaneous same-frequency positioning and communication signals at the receiving end, thereby achieving efficient utilization of time-frequency domain resources for communication and perception positioning.

[0015] 2. The method of the present invention can complete the mutual positioning between nodes in the GNSS denied area with only a single node. By receiving signals through a single-node array, the direction of incoming waves can be effectively confirmed and the number of nodes required for UAV swarm perception and positioning tasks can be reduced. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention;

[0017] Figure 2 This is a schematic diagram of a scenario provided for an embodiment of the present invention;

[0018] Figure 3 This is a diagram of a uniform array signal receiving model.

[0019] Figure 4 Here is a flowchart of the joint diagonalization algorithm for the feature matrix;

[0020] Figure 5 The waveform diagram shows the positioning signal and the communication signal.

[0021] Wherein, (a) is the positioning signal waveform, and (b) is the communication signal waveform;

[0022] Figure 6 The waveform diagram is for the separated signal.

[0023] Wherein, (a) is the separation signal 1, and (b) is the separation signal 2;

[0024] Figure 7 This is a diagram showing the location results;

[0025] Where (a) represents distance information and (b) represents distance information. Detailed Implementation

[0026] To facilitate understanding of the technical content of this invention by those skilled in the art, the following description, in conjunction with the accompanying drawings, further illustrates the invention.

[0027] To address the issue of UAV signal loss within GNSS denied zones, this invention utilizes an auxiliary positioning system carried by the swarm leader node to rapidly search for and locate lost UAVs within a certain range. The auxiliary positioning system shares the same radio frequency hardware channel with the communication system. Therefore, the signal received by the swarm leader node may contain a mixture of positioning and communication signals. By constructing a received signal separation model using an array receiver, the received time-frequency conflicting positioning and communication signals are separated, eliminating mutual interference between positioning and communication signals. This enables position updates within the UAV swarm. Further processing of the separated positioning signals yields accurate positioning of the node to be located, ultimately achieving efficient position management within the UAV swarm.

[0028] Combined with appendix Figure 1 The specific implementation steps of the present invention are described as follows:

[0029] Step 1: Establish the array receiving signal model

[0030] Assuming the positioning and communication subsystems within the group leader node share the same radio frequency hardware channel, and the positioning and communication signals operate on the same carrier frequency, when the group leader node activates its auxiliary positioning system to receive positioning signals, simultaneously, a drone is sending communication signals to the group leader node as shown in the attached diagram. Figure 2 As shown. Considering the receiving array is an M×N uniform area array, its received signal model is as follows. Figure 3 As shown, the element spacing is d, θ and Let x be the azimuth and elevation angles of the incoming wave direction, respectively. Taking the lower left element as the first row and first column, then after down-conversion, the baseband echo x received by the element in the m-th row and n-th column is... mn The expression for (t) is:

[0031]

[0032] Where t represents time, s(t) is the positioning signal, and θ s , θ represents the azimuth and elevation angles, respectively, representing the direction of arrival of the positioning signal; c(t) represents the communication signal, and θ represents the elevation angle. c , These are the azimuth and elevation angles, respectively, representing the direction of arrival of the communication signal; n mn (t) represents the Gaussian white noise received by the m-th row and n-th column element, and λ is the carrier frequency wavelength.

[0033] The received signal from the M×N array elements, after down-conversion processing, can be represented in the following matrix form.

[0034] x(t)=As(t)+n(t) (2)

[0035] Where, x(t)=[x 11 (t),x 12 (t),…,x MN (t)] T Let be the received signal vector, where (·) T It is the transpose symbol; This is an array manifold matrix, which can also be referred to as a signal mixing matrix in the method of this invention, wherein... Defined as:

[0036]

[0037] θ=θ s , Or θ = θ c , s(t) = [s(t), c(t)] T The signal vector containing the positioning signal s(t) and the communication signal c(t) can be called the source signal; n(t) is an M×N noise signal vector, defined as n(t) = [n 11 (t),n 12 (t),…,n MN (t)] T Its mean is 0, and its covariance matrix is ​​σ. 2 I M×N I M×N It is an M×N dimensional identity matrix.

[0038] Step 2: Calculate the MUSIC (Multiple Signal Classification) spectrum estimate of the received signal x(t).

[0039] Step 2-1: Calculate the correlation matrix R of the received signal x(t)

[0040] R = E[x(t)x H (t)] (4)

[0041] in(·) H E[] is the conjugate transpose symbol, and E[] represents the mathematical expectation.

[0042] Step 2-2: Perform eigenvalue decomposition on the correlation matrix R, and arrange the eigenvalues ​​in a monotonically increasing order to obtain the eigenvectors g1, g2, ..., g corresponding to the first M×N-2 eigenvalues. M×N-2 That is, to obtain the vectors of the noise subspace and construct the matrix G.

[0043]

[0044] in Represents the complex field.

[0045] Step 2-3: Construct the scan function

[0046]

[0047] in, As the array guide vector, spatial angles are assigned according to θ and Divide the data and scan sequentially for calculation. The Music spectrum estimate of the received signal x(t) can then be obtained.

[0048] Step 3: Solve for the unmixing matrix W using joint diagonalization of the characteristic matrices:

[0049] Step 3-1: Whiten the received signal x(t)

[0050] z(t)=Tx(t) (7)

[0051] Where z(t) is the whitened signal vector, and T is the whitening matrix, defined as follows:

[0052]

[0053] Λ is a diagonal matrix formed by the first two largest eigenvalues ​​of the correlation matrix R, and G max R is a matrix consisting of the eigenvectors corresponding to each eigenvalue of the correlation matrix R.

[0054] Step 3-2: Construct the fourth-order cumulant matrix of the whitened signal z(t) Confirm the cumulative characteristic matrix C z (M i ), i = 1, 2:

[0055] Calculate the fourth-order cumulant of the whitened signal z(t).

[0056]

[0057] Where E(·) represents the expected value, (·) * The conjugate symbol is used. All fourth-order cumulants of the whitened signal z(t) are constructed as 2. 2 ×2 2 Matrix C z For C z Eigenvalue decomposition yields eigenvalues ​​and eigenvectors. The eigenvalues ​​are sorted in monotonically decreasing order, and the first two eigenvalues ​​λ1, λ2 and their corresponding eigenvectors u1, u2 are taken to obtain the cumulant eigenmatrix C, which requires approximate joint diagonalization. z (M i )=λ i M i ,i=1,2, where Mi Let vec(M) be the characteristic matrix. i )=u i vec(·) represents the vectorization operation of a matrix.

[0058] Step 3-3: Find the unmixing matrix W; for the cumulant characteristic matrix C z (M i Perform joint diagonalization on i = 1 and 2:

[0059] Step 3-3-1: Let the initial matrix W = I2, which is a 2×2 dimensional identity matrix, and set the algorithm's stopping iteration threshold ξ.

[0060] Step 3-3-2: Calculation in

[0061] h i =[cm i,pp -cm i,qq ,cm i,pp +cm i,qq ,j(cm i,qp -cm i,pq )] T (10)

[0062] cm i,pp cm i,qq cm i,qp cm i,pq C respectively z (M i The values ​​of the (p,p), (q,q), (q,p), and (p,q)th elements of ), where p = 1 and q = 2 in the method of this invention.

[0063] Step 3-3-3: Perform eigenvalue decomposition on matrix B to obtain the normalized eigenvector [bx,by,bz] corresponding to the largest eigenvalue of matrix B. T .make

[0064]

[0065] Step 3-3-4: Use e1 and e2 as the Givens matrix G e The elements in G e Defined as

[0066]

[0067] Step 3-3-5: Update the unmixing matrix and the cumulant feature matrix:

[0068] W=WG e (13)

[0069]

[0070] Step 3-3-6: Determine whether e² ≥ ξ holds true, where ξ = 1 / e -6 The stopping condition for the algorithm is a given threshold. If the threshold is met, the algorithm returns to step 3-3-2; otherwise, the desired W is the unmixing matrix. The algorithm flow is shown in the appendix. Figure 4 As shown.

[0071] Step 4: Use the obtained unmixing matrix W to achieve signal separation:

[0072] The obtained demixing matrix W is left-multiplied by the received signal to achieve signal separation.

[0073]

[0074] The obtained separation signal is estimated.

[0075] Step 5: Demodulate the separated communication signals to achieve position updates within the UAV swarm; perform pulse compression processing on the separated positioning signals to obtain more accurate distance information for the UAV to be located. Combine the distance information with the received signal music spectrum and the communication signal azimuth information to determine the orientation of the UAV to be located, achieving accurate positioning within the UAV swarm.

[0076] Simulation verification and analysis

[0077] Simulation parameters:

[0078] A 6×6 uniform array is considered for reception, with an element spacing of 0.015m. The carrier frequency for both positioning and communication signals is 10GHz, with a bandwidth of 50MHz and a sampling frequency of 100MHz. The positioning signal is a linear frequency modulated (LFM) signal with a pulse repetition period of 6μs, a pulse width of 2μs, an elevation angle of 20° and an azimuth angle of -20° relative to the group leader node, and the UAV to be positioned is 100m from the group leader node. The communication signal is modulated using differential binary phase-shift keying (DBPSK), with an elevation angle of 5° and an azimuth angle of 30° relative to the group leader node. The signal-to-noise ratio (SNR) for the positioning signal is set to -5dB, and the SNR for the communication signal is set to 15dB.

[0079] Simulation analysis:

[0080] Appendix Figure 5 The waveforms of the positioning signal and communication signal are attached. Figure 6 For the separated signal waveform diagram, compare with the attached diagram. Figure 5 As can be clearly seen from the waveform, Separation Signal 1 is the received positioning signal, and Separation Signal 2 is the received communication signal. That is, after processing by this method, the time-frequency conflicting positioning and communication signals can be successfully separated, effectively eliminating mutual interference and realizing node position updates. (Appendix) Figure 7This demonstrates the distance information obtained after pulse compression processing of separated signal 1, as well as the music spectrum of the received signal. The music spectrum contains the azimuth information of both the communication and positioning signals. By combining the separated communication information with the music spectrum to confirm the azimuth, the positioning signal's azimuth information is obtained through filtering. (See appendix...) Figure 7 It can be seen that the distance and orientation information of the obtained positioning signal are consistent with the settings, that is, the method can achieve single-node positioning.

[0081] In summary, the UAV swarm location management method based on array-assisted positioning proposed in this invention can effectively solve the problem of mutual interference between communication signals and positioning signals under time-frequency domain resource conflict environment, and realize the efficient utilization of communication sensing and positioning resources.

[0082] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the scope of the claims of the invention.

Claims

1. A method for managing the location of UAV swarms based on array-assisted positioning, characterized in that, When a UAV loses its signal within a GNSS denied area, the swarm leader node re-searches for and locates the UAV while receiving communication signals from other nodes. The specific processing steps include: S1. Establish the array receiving signal model; specifically: the positioning subsystem and communication subsystem within the group leader node share the same radio frequency hardware channel. When the group leader node activates the auxiliary positioning system to receive positioning signals, a UAV simultaneously sends communication signals to the group leader node; considering that the receiving array is an M×N uniform area array, and the positioning signal and communication signal have the same carrier frequency, the M×N array element received signal, after down-conversion processing, can be represented in the following matrix form: x(t) = As(t) + n(t) Where A is the array manifold matrix, x(t) is the received signal vector, s(t) is the signal vector containing positioning and communication signals, which can be called the source signal, and n(t) is the M×N noise signal vector; S2. Calculate the MUSIC spectrum estimate of the received signal x(t); S3. Solve for the unmixing matrix W based on the whitened received signal x(t); S4. Use the demixing matrix W to separate the received signal x(t) to obtain the communication signal and the positioning signal; S5. Demodulate the separated communication signals to achieve position updates within the UAV cluster; perform pulse compression processing on the separated positioning signals to obtain distance information; combine the distance information with the MUSIC spectrum estimation in step S2 and the orientation information of the communication signals to determine the orientation of the UAV to be located, thereby achieving accurate positioning within the UAV cluster.

2. The method for managing the location of a UAV swarm based on array-assisted positioning according to claim 1, characterized in that, Step S2 specifically includes the following sub-steps: S21. Calculate the correlation matrix R of x(t): R=E[x(t)x H (t)] in,(·) H E[] represents the conjugate transpose symbol and the mathematical expectation. S22. Perform eigenvalue decomposition on the correlation matrix R, and arrange the eigenvalues ​​in a monotonically increasing order to obtain the eigenvectors g1, g2, ..., g2 corresponding to the first M×N-2 eigenvalues. M×N-2 According to g1, g2, ..., g M×N-2 Construct matrix G: in, Represents the field of complex numbers; S23. Construct a scanning function based on matrix G. Where, θ and These are the azimuth and elevation angles, respectively, in the direction of the incoming wave. For array guiding vector; S24. Arrange spatial angles according to θ and Divide the data and scan sequentially for calculation. The MUSIC spectrum estimate of x(t) can then be obtained.

3. The method for managing the location of a UAV swarm based on array-assisted positioning according to claim 2, characterized in that, Step S3 specifically includes the following sub-steps: S31. Whiten x(t): z(t) = Tx(t) Where z(t) is the whitened signal vector, and T is the whitening matrix, defined as: Λ is a diagonal matrix formed by the first two largest eigenvalues ​​of the correlation matrix R, and G max R is a matrix consisting of the eigenvectors corresponding to each eigenvalue of the correlation matrix R; S32. Construct the fourth-order cumulant matrix of the whitened signal z(t). Confirm the cumulative characteristic matrix C z (M i ), i = 1, 2; the specific process is as follows: Calculate the fourth-order cumulant of the whitened signal z(t). Where E(·) represents the expected value, (·) * The conjugate symbol; Construct all fourth-order cumulants of the whitened signal z(t) as 2 2 ×2 2 Matrix C z , for C z Eigenvalue decomposition yields eigenvalues ​​and eigenvectors, and C... z The eigenvalues ​​are sorted in a monotonically decreasing order. The first two eigenvalues ​​λ1 and λ2, along with their corresponding eigenvectors u1 and u2, yield the cumulant eigenma matrix C, which requires approximate joint diagonalization. z (M i )=λ i M i ,i=1,2, where M i Let vec(M) be the characteristic matrix. i )=u i vec(·) represents the vectorization operation of a matrix; S33, By analyzing the cumulative characteristic matrix C z (M i Perform joint diagonalization on i=1,2 to obtain the unmixed matrix W.

4. The method for managing the location of a UAV swarm based on array-assisted positioning according to claim 3, characterized in that, The specific implementation process of step S33 is as follows: S331. Let the initial matrix W = I2 be a 2×2 identity matrix, and set the algorithm's stopping iteration threshold ξ. S332, Calculation in h i =[cm i,pp -cm i,qq ,cm i,pp +cm i,qq ,j(cm i,qp -cm i,pq )] T cm i,pp cm i,qq cm i,qp cm i,pq C respectively z (M i Let p be the (p,p), (q,q), (q,p), and (p,q)th element values ​​of , where p = 1 and q = 2; S333. Perform eigenvalue decomposition on matrix B to obtain the normalized eigenvector [bx,by,bz] corresponding to the largest eigenvalue of matrix B. T ; make S334. Use e1 and e2 as Givens matrix G. e The elements in G e Defined as: S335, Update the unmixing matrix and the cumulant feature matrix: W=WG e S336. Determine whether e2≥ξ holds true. If it does, return to step S332; otherwise, the required W is the unmixing matrix.

Images