Unmanned aerial vehicle group passive positioning path optimization and node selection method based on CRLB

Through the CRLB-based drone cluster path optimization and node selection method, the node selection and grouping problems of drone clusters during multi-target passive positioning are solved, the resource usage efficiency and positioning accuracy are improved, and high-precision target positioning tracking is achieved .

CN119085644APending Publication Date: 2024-12-06SICHUAN UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410988978.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-23
Publication Date
2024-12-06

AI Technical Summary

Technical Problem

When existing drone groups are in passive positioning for multi-targets, the problems of node selection and packetization have not been effectively solved, resulting in low resource usage efficiency and poor positioning accuracy.

Method used

Adopting the CRLB-based drone cluster path optimization and node selection method, the efficiency of resource utilization of drone clusters is improved and Improvement of target positioning accuracy.

Benefits of technology

Through this method, the multi-target passive positioning accuracy and resource utilization efficiency are improved, high-precision positioning tracking is achieved, node selection and grouping scheme of the drone group are optimized, and target passive positioning performance of the drone group is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119085644A_ABST
    Figure CN119085644A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle group passive positioning path optimization and node selection method based on CRLB. The method comprises the following steps: establishing a target passive positioning signal model; calculating an estimated position of a target point based on a Chan-TDOA positioning algorithm and deducing CRLB; unmanned aerial vehicle group nodes are selected based on CRLB, and an optimal node selection method and a grouping scheme are obtained, so that the resource utilization efficiency of the unmanned aerial vehicle group and the target passive positioning precision are improved; an unmanned aerial vehicle group path is optimized based on CRLB, real-time effective configuration of unmanned aerial vehicle group node positions is realized, unmanned aerial vehicle group path optimization in continuous time is formed, and target positioning precision is improved; fusing the positioning target position information and the target position prediction information to obtain a final position of the target; according to the invention, unmanned aerial vehicle group path optimization and node selection are carried out based on CRLB, an unmanned aerial vehicle group path optimization scheme in continuous time is formed, and multi-target passive positioning precision and resource utilization efficiency are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle target positioning, and in particular to a CRLB-based unmanned aerial vehicle swarm passive positioning path optimization and node selection method. Background Art

[0002] With the rapid development of electronic information, the current positioning technology is divided into active positioning and passive positioning. Although active positioning has high positioning accuracy, it has the risk of being easily exposed due to the need to radiate electromagnetic waves. At the same time, it also has high requirements for energy consumption and equipment. The above problems can be solved by passive positioning technology. Passive positioning refers to a technology that does not actively radiate electromagnetic waves to the target being reconnaissance, but only relies on receiving electromagnetic waves radiated by the target to achieve reconnaissance and positioning of the target. Therefore, passive positioning has good concealment, strong survivability, and does not require signal transmission equipment.

[0003] From the perspective of UAV swarm path optimization, the configuration of UAV swarm has an important impact on the accuracy of target passive positioning. By configuring the configuration of UAV swarm in real time, the path optimization of UAV swarm in continuous time can be realized, which can effectively improve the target positioning performance of UAV swarm. S.Bayili et al. proposed an A* algorithm based on heuristic function and cost function for the path search problem in complex environment. By evaluating the cost and heuristic estimate of each node to be traversed, the path search was carried out, and the shortest path was quickly searched when performing positioning tasks. M.Darrah et al. proposed a genetic algorithm based on improved fitness function for multi-UAV collaborative path planning. By optimizing the fitness function, the optimal path was searched from the candidate solution space, and the global path was searched stably to avoid UAV collision. K.Price proposed a differential evolution algorithm with fast convergence speed and high solution accuracy for the problem of fast search for the shortest overall journey of UAV, which reduced the energy and time consumption of UAV. X.Wang et al. proposed a bat global search algorithm that can cope with dynamic changes in the environment for the problem of minimizing the time to complete the UAV task, and improved the time efficiency of the UAV task completion.

[0004] The above research results optimize the path of drone swarm from aspects of improving positioning accuracy, shortest path search, obstacle avoidance path search, etc. However, they have not yet considered the node selection and grouping problem when the drone swarm performs multiple tasks, especially when the drone swarm needs to passively locate multiple targets, the drone swarm nodes are not reasonably selected and grouped, resulting in low resource utilization efficiency and poor multi-target passive positioning accuracy. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the present invention provides a CRLB-based UAV swarm passive positioning path optimization and node selection method, which realizes the formation of a continuous time UAV swarm path optimization scheme by performing UAV swarm path optimization and node selection based on CRLB, thereby improving the multi-target passive positioning accuracy and improving resource utilization efficiency.

[0006] In order to achieve the above-mentioned invention object, the present invention adopts the following technical solutions:

[0007] The first aspect of the present application provides a CRLB-based passive positioning path optimization and node selection method for a drone swarm, comprising the following steps:

[0008] S101, establishing a target passive positioning signal model;

[0009] S102, calculating the estimated position of the target point based on the Chan-TDOA positioning algorithm and deriving the CRLB;

[0010] S103, selecting drone swarm nodes based on CRLB, obtaining the best node selection method and grouping scheme to improve the resource utilization efficiency of the drone swarm and the target passive positioning accuracy;

[0011] S104, optimize the drone group path based on CRLB, realize the real-time and effective configuration of the drone group node position, form the drone group path optimization in continuous time, and improve the target positioning accuracy;

[0012] S105: Fusing the positioning target position information and the target position prediction information to obtain the final position of the target.

[0013] Furthermore, step S101 includes the following steps:

[0014] S11, set the number of targets to m, and the position coordinates of the jth target to [x j ,y j ](j=1,2,3,......m); Assume that the number of drones carrying radar is n, and the position coordinates of the i-th drone are [x i ,y i ](i=0,1,2…n-1), [x0,y0] represents the coordinates of the main drone;

[0015] S12. Calculate the distance from the i-th UAV to the j-th target. The calculation expression is as follows:

[0016]

[0017] Among them, r i is the distance from the i-th UAV to the j-th target, x j is the horizontal coordinate of the jth target position, yj is the ordinate of the jth target position, x i is the horizontal coordinate of the position of the i-th UAV, y i is the ordinate of the position of the i-th UAV;

[0018] Calculate the distance difference between the i-th UAV and the main UAV. The calculation expression is as follows:

[0019] r i -r0=c(τ i -τ0)

[0020] i=1,2,3,...,n-1

[0021] Among them, τ i is the time it takes for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation, r i is the distance from the i-th UAV to the j-th target, r0 is the distance from the main UAV to the j-th target;

[0022] S13. Construct the covariance matrix of TDOA measurement values, which is expressed as follows:

[0023]

[0024] Where Q is the covariance matrix of TDOA measurements, σ i,1 is the variance of the distance difference measurement between the ith UAV and the benchmark UAV.

[0025] Further, step S102 includes the following steps:

[0026] S21, constructing a TDOA nonlinear equation system based on the initialization parameters of step S101 and the relationship between the parameters;

[0027] S22, performing the first weighted least square method of the Chan algorithm based on the TDOA nonlinear equations to obtain a maximum likelihood estimate;

[0028] S23, executing the second weighted least square method of Chan algorithm based on maximum likelihood estimation to obtain an initial estimate of the target position;

[0029] S24. Derivation of CRLB based on Chan-TDOA positioning algorithm.

[0030] Further, step S21 includes the following steps:

[0031] The expression of the distance difference between the i-th UAV and the main UAV and the target is as follows:

[0032] r i -r0=c(τi -τ0)

[0033] i=1,2,3,...,n-1

[0034] Among them, τ i is the time it takes for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation, r i is the distance from the i-th UAV to the j-th target, r0 is the distance from the main UAV to the j-th target;

[0035] Based on the distance difference between the i-th UAV and the main UAV from the target, we get:

[0036]

[0037] in, x i,0 =x i -x0,y i,0 =y i -y0,x i is the horizontal coordinate of the position of the i-th UAV, y i is the ordinate of the position of the i-th UAV, x0 is the abscissa of the position of the main UAV, and y0 is the ordinate of the position of the main UAV;

[0038] By processing the TDOA nonlinear equations converted into the matrix equation form, the expression of the TDOA nonlinear equations converted into the matrix equation form is as follows:

[0039] h=GZ

[0040] in, Z=[x j ,y j , r0] T .

[0041] Further, step S22 includes the following steps:

[0042] Based on the TDOA nonlinear equations, the error vector of TDOA noise is: ψ = h-GZ 0 ;

[0043] The maximum likelihood estimate obtained by weighted least squares is expressed as follows:

[0044] Z = arg min {(h-GZ) T Ψ -1 (h-GZ)}

[0045] ≈(G T Q -1 G)G T Q-1 h

[0046] Among them, Z is the maximum likelihood estimate, Z=[x j ,y j , r0] T , Q is the measurement covariance matrix, ψ=h-GZ 0 .

[0047] Further, step S23 includes the following steps:

[0048] Construct the residual vector ψ′=h′-G′Z′, and the expression of the second weighted least squares method to estimate the position is as follows:

[0049] Z′=(G′TΨ′ -1 G′)G′TΨ′ -1 h′

[0050] Among them, ψ′=h′-G′Z′, Z′ is the second weighted least squares estimated position;

[0051] The second weighted least squares method of the Chan algorithm is performed based on the maximum likelihood estimation to obtain the initial estimate of the target position, which is expressed as follows:

[0052]

[0053] Among them, x is the initial estimated value of the target horizontal coordinate position, y is the initial estimated value of the target vertical coordinate position, Z' is the second weighted least squares estimated position, x0 is the horizontal coordinate of the main UAV position, and y0 is the vertical coordinate of the main UAV position.

[0054] Further, step S24 includes the following steps:

[0055] The measurement equation is established, and its expression is as follows:

[0056] s=[cτ 1,0 , cτ 2,0 , ..., cτ i,0 ]

[0057] i=1,2,3,...,n-1

[0058] Among them, s′=s+m, s is the true value, s′ is the distance difference measurement value, τ i is the time for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation;

[0059] The expression of the probability density function of the system measurement value is as follows:

[0060]

[0061] Where f is the probability density function of the system measurement value, s is the true value, s′ is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 , δ 2 2,0 , …, δ 2 n1,0};

[0062] Based on the probability density function of the system measurement value, its likelihood function is obtained, and its expression is as follows:

[0063]

[0064] Where lnf is the probability density likelihood function of the system measurement value, s is the true value, s′ is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 , δ 2 2,0 , …, δ 2 n-1,0};

[0065] Find its second-order partial derivative

[0066]

[0067]

[0068]

[0069] Find its mathematical expectation:

[0070]

[0071]

[0072]

[0073] The Fisher information matrix is:

[0074]

[0075] The expression of the CRLB matrix is ​​as follows:

[0076] CRLB(s)=FIM- 1 (s)

[0077] Among them, s is the true value.

[0078] Further, step S103 includes the following steps:

[0079] S31. According to the existence of M target signal sources with different positions on the two-dimensional plane, the number of drones carrying radars participating in passive positioning is N, and the number of groups of drone groups after node selection is M;

[0080] S32. Clarify the constraints for selecting drone nodes;

[0081] S33. Select the drone swarm nodes based on CRLB to obtain the best node selection method and grouping scheme.

[0082] Further, step S32 includes the following steps:

[0083] The number of drones locating the same target is the same, and the quantitative association between the number of drones involved in positioning and the number of targets is constructed. The expression is as follows:

[0084] N=λM

[0085] Among them, N is the number of drones carrying radars participating in passive positioning, M is the number of groups in the drone swarm after node selection, and λ is the multiple of the number of drones participating in positioning and the number of targets.

[0086] In order to meet the needs of Chan-TDOA two-dimensional plane positioning solution, each target requires three or more UAVs to locate it, so λ≥3.

[0087] Further, step S33 includes the following steps:

[0088] The optimization goal can be expressed as the average value of the positioning CRLB trace of each group of drones for target j after the drone group node selection at this moment:

[0089]

[0090] Therefore, the mathematical model of the drone node selection optimization problem can be described as:

[0091]

[0092] stN=λM

[0093] λ ≥ 3

[0094] Among them, N is the number of radar-carrying drones participating in passive positioning, and M is the number of groups in the drone swarm after node selection;

[0095] The average values ​​of the CRLB traces of multi-target passive positioning under different node selection schemes are calculated, and the node selection (grouping) scheme that can minimize the average value of the CRLB traces of multi-target passive positioning is selected as the optimal scheme at the current moment.

[0096] Further, step S104 includes the following steps:

[0097] S41. According to the movement of the drone group and the target, let the flight time be t, the path optimization time interval be Δt, the drone flight speed be v, the drone heading angle be X, and the drone turning radius at time t be The minimum turning radius is L min , the minimum approach distance between the UAV and the target is d min ;

[0098] S42. Calculate the turning radius The turning radius of the drone is

[0099] S43. The CRLB-based UAV swarm path optimization algorithm realizes real-time and effective configuration of the UAV swarm node locations.

[0100] Further, step S42 includes the following steps:

[0101] Set up the UAV path optimization constraints;

[0102] Assume the turning radius of the UAV is L i Greater than or equal to the minimum turning radius, The turning arc of the drone is The turning arc is approximately a circular arc, so According to the geometric relationship, a mathematical equation is established, and its expression is as follows:

[0103]

[0104] in, is the turning radius of the i-th UAV, is the turning radius.

[0105] Further, step S43 includes the following steps:

[0106] The distance between the drone and the target r i Greater than or equal to the minimum approach distance:

[0107]

[0108] The optimization goal can be expressed as the positioning CRLB trace of the drone swarm at the next moment under different drone swarm node position configuration schemes:

[0109] Tr(CRLB t )

[0110] Therefore, the mathematical model of the UAV swarm path optimization problem can be described as:

[0111] min arg Tr(CRLB t )

[0112]

[0113] r i t ≥d min

[0114] in, is the turning radius of the i-th UAV, L min is the minimum turning radius, d min The minimum approach distance.

[0115] The beneficial effects of the present application are as follows: by performing UAV swarm path optimization and node selection based on CRLB, a UAV swarm path optimization solution is formed in continuous time, the multi-target passive positioning accuracy is improved, and the resource utilization efficiency is improved.

[0116] When a swarm of drones approaches a target for passive positioning, the optimal node selection method and grouping scheme can be obtained by optimizing the above-mentioned optimization mathematical model at each moment, thereby improving the resource utilization efficiency of the swarm of drones and the passive positioning accuracy of the target.

[0117] It is achieved by traversing different node selection schemes that meet the constraints, calculating the average CRLB trace of multi-target passive positioning under different node selection schemes, and selecting the node selection (grouping) scheme that can minimize the average CRLB trace of multi-target passive positioning as the optimal scheme at the current moment.

[0118] The passive positioning performance of the target has been improved through the two angles of drone swarm path optimization and node selection, thereby achieving high-precision positioning and tracking.

[0119] By using the real-time configuration method of drone swarm node positions based on CRLB, a drone swarm path optimization plan in continuous time is formed to improve target positioning accuracy; by using the drone swarm node selection method based on CRLB, the multi-target passive positioning accuracy and resource utilization efficiency are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0121] Figure 1 It is a schematic diagram of the steps of a CRLB-based passive positioning path optimization and node selection method for a drone swarm of the present invention;

[0122] Figure 2 It is a geometrical diagram of the TDOA positioning model of the present invention;

[0123] Figure 3 This is a flow chart of the node selection algorithm based on CRLB in the present invention;

[0124] Figure 4 This is a schematic diagram of the turning radius of the UAV of the present invention;

[0125] Figure 5 This is a flow chart of the path optimization algorithm based on CRLB in the present invention;

[0126] Figure 6 This is a schematic diagram of a multi-target passive positioning scenario of a drone swarm according to the present invention;

[0127] Figure 7 A schematic diagram for node selection of the present invention;

[0128] Figure 8 Selecting simulation error comparison results for the node of the present invention;

[0129] Fig. 9 It is a V-shaped drone group configuration of the present invention;

[0130] Fig.10 The error comparison of the UAV swarm with and without node optimization in the V-shaped configuration of the present invention;

[0131] Fig.11 It is a horizontally arranged UAV group configuration of the present invention;

[0132] Fig.12 The error comparison of the UAV group with and without node optimization in the horizontal row configuration of the present invention;

[0133] Fig.13 It is a rhombus-shaped drone group configuration of the present invention;

[0134] Fig.14 The error comparison of the diamond-shaped UAV group configuration with and without node optimization of the present invention;

[0135] Fig.15 A group of drones and positioning targets after the nodes are selected for the simulation of the present invention;

[0136] Fig.16 This is a simulation diagram of the UAV track without path planning in the present invention;

[0137] Fig.17 This is a simulation diagram of the UAV track under the path planning of the present invention;

[0138] Fig.18 Comparison of RMSE and CRLB of passive positioning with and without path optimization in the present invention. DETAILED DESCRIPTION

[0139] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0140] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0141] Embodiment 1:

[0142] A CRLB-based passive positioning path optimization and node selection method for a drone swarm includes the following steps:

[0143] S101, establishing a target passive positioning signal model;

[0144] like Figure 2 As shown, establishing the target passive positioning signal model includes the following steps:

[0145] S11, set the number of targets to m, and the position coordinates of the jth target to [x j ,y j ](j=1,2,3,......m); Assume that the number of drones carrying radar is n, and the position coordinates of the i-th drone are [x i ,y i ](i=0,1,2…n-1), where [x0,y0] represents the coordinates of the main drone;

[0146] S12. Calculate the distance from the i-th UAV to the j-th target. The calculation expression is as follows:

[0147]

[0148] Among them, r i is the distance from the i-th UAV to the j-th target, x j is the horizontal coordinate of the jth target position, y j is the ordinate of the jth target position, x i is the horizontal coordinate of the position of the i-th UAV, y i is the ordinate of the position of the i-th UAV;

[0149] Calculate the distance difference between the drone and the main drone. The calculation expression is as follows:

[0150] r i -r0=c(τ i -τ0)

[0151] i=1,2,3,...,n-1

[0152] Among them, τ i is the time it takes for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation, r i is the distance from the i-th UAV to the j-th target, r0 is the distance from the main UAV to the j-th target;

[0153] S13. Set the covariance matrix of TDOA measurement values:

[0154]

[0155] Where Q is the covariance matrix of TDOA measurements, σ i,1 is the variance of the distance difference measurement between the ith UAV and the benchmark UAV.

[0156] S102, calculating the estimated position of the target point based on the Chan-TDOA positioning algorithm and deriving the CRLB;

[0157] Calculating the target position based on the Chan-TDOA positioning algorithm includes the following steps:

[0158] S21, constructing a TDOA nonlinear equation system based on the initialization parameters of step S101 and the relationship between the parameters;

[0159] Step S21 specifically includes:

[0160] The expression of the distance difference between drone i and the main drone from the target is as follows:

[0161] r i -r0=c(τ i -τ0)

[0162] i=1,2,3,...,n-1

[0163] Among them, τ i is the time it takes for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation, r i is the distance from the i-th UAV to the j-th target, r0 is the distance from the main UAV to the j-th target;

[0164] Based on the distance difference between UAV i and the main UAV from the target, we get:

[0165]

[0166] in, x i,0 =x i -x0,y i,0 =y i -y0,x i is the horizontal coordinate of the position of the i-th UAV, y i is the ordinate of the position of the i-th UAV, x0 is the abscissa of the position of the main UAV, and y0 is the ordinate of the position of the main UAV;

[0167] By processing the TDOA nonlinear equations converted into the matrix equation form, the expression of the TDOA nonlinear equations converted into the matrix equation form is as follows:

[0168] h=GZ

[0169] in, Z=[x j ,y j , r 0 ] T .

[0170] S22, performing the first weighted least square method of the Chan algorithm based on the TDOA nonlinear equations to obtain a maximum likelihood estimate;

[0171] Step S22 includes the following steps:

[0172] Based on the TDOA nonlinear equations, the error vector of TDOA noise is: ψ = h-GZ 0 ;

[0173] The maximum likelihood estimate obtained by weighted least squares is expressed as follows:

[0174] Z = argmin{(h-GZ) T Ψ -1 (h-GZ)}

[0175] ≈(G T Q -1 G)G T Q -1 h

[0176] Among them, Z is the maximum likelihood estimate, Z=[x j ,y j , r0] T , Q is the measurement covariance matrix, ψ=h-GZ 0 .

[0177] S23, executing the second weighted least square method of Chan algorithm based on maximum likelihood estimation to obtain an initial estimate of the target position;

[0178] Step S23 includes the following steps:

[0179] Construct the residual vector ψ′=h′-G′Z′, and the expression of the second weighted least squares method to estimate the position is as follows:

[0180] Z′=(G′TΨ′ -1 G′)G′TΨ′ -1 h′

[0181] Among them, ψ′=h′-G′Z′, Z′ is the second weighted least squares estimated position;

[0182] The second weighted least squares method of the Chan algorithm is performed based on the maximum likelihood estimation to obtain the initial estimate of the target position, which is expressed as follows:

[0183]

[0184] Among them, x is the initial estimated value of the target horizontal coordinate position, y is the initial estimated value of the target vertical coordinate position, Z' is the second weighted least squares estimated position, x0 is the horizontal coordinate of the main UAV position, and y0 is the vertical coordinate of the main UAV position.

[0185] S24, deriving CRLB based on Chan-TDOA positioning algorithm;

[0186] Step S24 includes the following steps:

[0187] The measurement equation is established, and its expression is as follows:

[0188] s=[cτ 1,0 , cτ 2,0 , ..., cτ i,0 ]

[0189] i=1,2,3,...,n-1

[0190] Among them, s′=s+m, s is the true value, s′ is the distance difference measurement value, τ i is the time for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), c is the speed of signal propagation;

[0191] The expression of the probability density function of the system measurement value is as follows:

[0192]

[0193] Where f is the probability density function of the system measurement value, s is the true value, s′ is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 , δ 2 2,0 , …, δ 2 n-1,0};

[0194] Based on the probability density function of the system measurement value, its likelihood function is obtained, and its expression is as follows:

[0195]

[0196] Where lnf is the probability density likelihood function of the system measurement value, s is the true value, s′ is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 , δ 2 2,0 , …, δ 2 n-1,0};

[0197] Find its second-order partial derivative

[0198]

[0199]

[0200]

[0201] Find its mathematical expectation:

[0202]

[0203]

[0204]

[0205] The Fisher information matrix is:

[0206]

[0207] The expression of the CRLB matrix is ​​as follows:

[0208] CRLB(s)=FIM- 1 (s)

[0209] Among them, s is the true value.

[0210] S103, selecting drone swarm nodes based on CRLB, obtaining the best node selection method and grouping scheme to improve the resource utilization efficiency of the drone swarm and the target passive positioning accuracy;

[0211] Step S103 includes the following steps:

[0212] S31. According to the existence of M target signal sources with different positions on the two-dimensional plane, the number of drones carrying radars participating in passive positioning is N, and the number of groups of drone groups after node selection is M;

[0213] S32. Clarify the constraints for selecting drone nodes;

[0214] Step S32 includes the following steps:

[0215] The number of drones locating the same target is the same, and the quantitative association between the number of drones involved in positioning and the number of targets is constructed. The expression is as follows:

[0216] N=λM

[0217] Among them, N is the number of drones carrying radars participating in passive positioning, M is the number of groups in the drone swarm after node selection, and λ is the multiple of the number of drones participating in positioning and the number of targets.

[0218] In order to meet the needs of Chan-TDOA two-dimensional plane positioning solution, each target requires three or more UAVs to locate it, so λ≥3.

[0219] S33, selecting drone swarm nodes based on CRLB, and obtaining the best node selection method and grouping scheme;

[0220] Step S33 includes the following steps:

[0221] The optimization goal can be expressed as the average value of the positioning CRLB trace of each group of drones for target j after the drone group node selection at this moment:

[0222]

[0223] Therefore, the mathematical model of the drone node selection optimization problem can be described as:

[0224]

[0225] stN=λM

[0226] λ ≥ 3

[0227] Among them, N is the number of radar-carrying drones participating in passive positioning, and M is the number of groups in the drone swarm after node selection;

[0228] Calculate the average value of the CRLB trace of multi-target passive positioning under different node selection schemes, and select the node selection (grouping) scheme that can minimize the average value of the CRLB trace of multi-target passive positioning as the optimal scheme at the current moment. The algorithm flow chart is shown in the attached figure. Figure 4 shown.

[0229] S104, optimize the drone group path based on CRLB, realize the real-time and effective configuration of the drone group node position, form the drone group path optimization in continuous time, and improve the target positioning accuracy;

[0230] Step S104 includes the following steps:

[0231] S41. According to the movement of the drone group and the target, let the flight time be t, the path optimization time interval be Δt, the drone flight speed be v, the drone heading angle be X, and the drone turning radius at time t be The minimum turning radius is L min , the minimum approach distance between the UAV and the target is d min ;

[0232] S42. Calculate the turning radius The turning radius of the drone is

[0233] like Figure 3 As shown, step S42 includes the following steps:

[0234] Set up the UAV path optimization constraints;

[0235] Assume the turning radius of the UAV is L i Greater than or equal to the minimum turning radius, The turning arc of the drone is The turning arc is approximately a circular arc, so According to the geometric relationship, a mathematical equation is established, and its expression is as follows:

[0236]

[0237] in, is the turning radius of the i-th UAV, is the turning radius;

[0238] S43, based on CRLB, the UAV swarm path optimization algorithm realizes the real-time and effective configuration of the UAV swarm node positions;

[0239] Step S43 includes the following steps:

[0240] like Figure 4 As shown, the distance between the UAV and the target is r i Greater than or equal to the minimum approach distance:

[0241]

[0242] The optimization goal can be expressed as the positioning CRLB trace of the drone swarm at the next moment under different drone swarm node position configuration schemes:

[0243] Tr(CRLB t )

[0244] Therefore, the mathematical model of the UAV swarm path optimization problem can be described as:

[0245] min arg Tr(CRLB t )

[0246]

[0247] r i t ≥d min

[0248] in, is the turning radius of the i-th UAV, L min is the minimum turning radius, d min The minimum approach distance.

[0249] S105, fusing the positioning target position information and the target position prediction information to obtain the final position of the target;

[0250] Step S105 specifically includes:

[0251] Repeat steps S102, S103, and S104 to update the optimized positioning result.

[0252] The following is a computer simulation experiment of the target positioning and tracking algorithm based on the improved TDOA and Kalman prediction model of the present invention:

[0253] The positions of the 9 drones and the 3 targets are shown in the parameters in 6.1. The topological structure and the direction of the drone movement are shown in Figure 7 shown.

[0254] RMSE is also used as the evaluation index to judge the target passive positioning accuracy, and a total of 1000 independent Monte Carlo experiments are conducted. Figure 8 For Figure 7 Simulation results of passive positioning RMSE comparison based on CRLB drone node selection / grouping and random selection / grouping under topological structure. As shown in the figure, when the drone swarm performs node selection and grouping, the passive positioning RMSE of the three targets is lower than that of the random selection and grouping passive positioning. It can be seen that by selecting and grouping effective nodes in the drone swarm, the target passive positioning accuracy can be effectively improved, thereby improving the resource utilization efficiency of the drone swarm.

[0255] In order to further verify the influence of effective node selection and grouping of drone swarm on target positioning accuracy, three groups of different drone configurations were set up to study the effect of node selection and grouping based on CRLB on improving target positioning accuracy. Fig.10 , Attachment Fig.12 and attached Fig.14 By comparing the results of the CRLB-based UAV swarm node selection and grouping, it can be concluded that the target positioning accuracy can be improved and the resource utilization efficiency of the UAV swarm can be improved.

[0256] According to the passive positioning of multiple targets by multiple drone swarms, the number of drone swarm nodes is n=9, and their coordinates are: [0,0]m, [350,1000]m, [800,100]m, [-200,200]m, [-600,700]m, [-800,-150]m, [-250,-350]m, [380,-520]m, [200,-900]m. The number of targets is m=3, and their coordinates are: [3500,4400]m, [1200,10000]m, [4000,12000]m; the flight speed of the drone swarm is v=60m / s, Δt=5s, and the minimum turning radius is L min =60m: variable = value, minimum approach distance is d min =500m.

[0257] In order to simplify the simulation experiment, three UAV nodes are selected at each moment to passively locate a target (i.e., λ=3), and its moving path is optimized. The application scenario diagram is shown in the figure below. Figure 6 shown.

[0258] In order to simplify the simulation experiment, the simulation of the drone swarm path optimization algorithm considers that the drone swarm nodes have completed selection and grouping. Therefore, only one group (three drones passively positioning one target) of the drone swarm path optimization is considered. At this time, the initial positions of the three drone nodes are: [350,1000]m, [380,-520]m, [800,100]m, and the target position is: [3500,4400]m. The schematic diagram of its topological structure is shown in the attached figure. Fig.15 As shown, the positioning time is 12Δt.

[0259] Taking the root mean square error (RMSE) as the evaluation index of positioning error, C independent Monte Carlo repeated experiments (C = 300) were conducted to compare the target positioning accuracy of the UAV swarm with and without path optimization. Fig.16 and 17 . Fig.16 and 17 They represent the movement trajectory of the drone swarm with and without path optimization, Fig.16It means that the UAV group moves in a configuration that minimizes the positioning CRLB. At this time, the target position obtained by passive positioning is closer to the actual position.

[0260] Fig.18 The passive positioning CRLB and RMSE of the drone swarm with and without path optimization are compared. According to the simulation results, as the drone approaches the target, the positioning error RMSE and CRLB in both cases show a decreasing trend. This shows that the drone swarm path optimization algorithm based on CRLB can effectively reduce the positioning error and improve the positioning accuracy.

[0261] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0262] The terms "first", "second" and "third" etc. in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein, for example. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0263] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A CRLB-based passive positioning path optimization and node selection method for drone swarms, characterized in that: The following steps are involved: S101, establishing a target passive positioning signal model; S102, calculating the estimated position of the target point based on the Chan-TDOA positioning algorithm and deriving the CRLB; S103, selecting drone swarm nodes based on CRLB, obtaining the best node selection method and grouping scheme to improve the resource utilization efficiency of the drone swarm and the target passive positioning accuracy; S104, optimize the drone group path based on CRLB, realize the real-time and effective configuration of the drone group node position, form the drone group path optimization in continuous time, and improve the target positioning accuracy; S105: Fusing the positioning target position information and the target position prediction information to obtain the final position of the target.

2. The method for passive positioning path optimization and node selection of unmanned aerial vehicle swarm based on CRLB according to claim 1 is characterized in that: The step S101 comprises the following steps: S11, set the number of targets to m, and the position coordinates of the jth target to [x j ,y j ]; Assume that the number of drones carrying radars is n, and the position coordinates of the i-th drone are [x i ,y i ], the coordinates of the main drone are [x0,y0]; S12. Calculate the distance from the i-th UAV to the j-th target. The calculation expression is as follows: Among them, r i is the distance from the i-th UAV to the j-th target, x j is the horizontal coordinate of the jth target position, y j is the ordinate of the jth target position, x i is the horizontal coordinate of the position of the i-th UAV, y i is the ordinate of the position of the i-th UAV; Calculate the distance difference between the i-th UAV and the main UAV. The calculation expression is as follows: r i -r0=c(τ i -τ0) i=1,2,3,...,n-1 Among them, τ i is the time it takes for the target radiation signal to propagate to each UAV, c is the speed of signal propagation, r i is the distance from the i-th UAV to the j-th target, r0 is the distance from the main UAV to the j-th target; S13. Construct the covariance matrix of TDOA measurement values, which is expressed as follows: Where Q is the covariance matrix of TDOA measurements, σ i,1 is the variance of the distance difference measurement between the ith UAV and the benchmark UAV.

3. The method for passive positioning path optimization and node selection of drone swarm based on CRLB according to claim 1 is characterized in that: The step S102 includes the following steps: S21, constructing a TDOA nonlinear equation system based on the initialization parameters of step S101 and the relationship between the parameters; S22, performing the first weighted least square method of the Chan algorithm based on the TDOA nonlinear equations to obtain a maximum likelihood estimate; S23, executing the second weighted least square method of Chan algorithm based on maximum likelihood estimation to obtain an initial estimate of the target position; S24. Derivation of CRLB based on Chan-TDOA positioning algorithm.

4. The method for passive positioning path optimization and node selection of drone swarm based on CRLB according to claim 3 is characterized in that: The step S23 comprises the following steps: Construct the residual vector ψ'=h'-G'Z', and the expression of the second weighted least squares estimated position is as follows: Z'=(G' T P' -1 G')G' T P' -1 h' Among them, ψ'=h'-G'Z', Z' is the second weighted least squares estimated position; The second weighted least squares method of the Chan algorithm is performed based on the maximum likelihood estimation to obtain the initial estimate of the target position, which is expressed as follows: Among them, x is the initial estimated value of the target horizontal coordinate position, y is the initial estimated value of the target vertical coordinate position, Z' is the second weighted least squares estimated position, x0 is the horizontal coordinate of the main UAV position, and y0 is the vertical coordinate of the main UAV position.

5. The method for passive positioning path optimization and node selection of unmanned aerial vehicle swarm based on CRLB according to claim 3 is characterized in that: The step S24 comprises the following steps: The measurement equation is established, and its expression is as follows: s=[cτ 1,0 ,cτ 2,0 ,...,cτ i,0 ] i=1,2,3,...,n-1 Among them, s'=s+m, s is the true value, s' is the distance difference measurement value, τ i is the time for the target radiation signal to propagate to each UAV (i=0,1,2……n-1), and c is the speed of signal propagation; The expression of the probability density function of the system measurement value is as follows: Where f is the probability density function of the system measurement value, s is the true value, s' is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 ,δ 2 2,0 ,...,δ 2 n-1,0 }; Based on the probability density function of the system measurement value, its likelihood function is obtained, and its expression is as follows: Among them, lnf is the probability density likelihood function of the system measurement value, s is the true value, s' is the distance difference measurement value, Q is, Q = diag{δ 2 1,0 ,δ 2 2,0 ,...,δ 2 n-1,0 }; Find its second-order partial derivative Find its mathematical expectation: The Fisher information matrix is: The expression of the CRLB matrix is ​​as follows: CRLB(s)=FIM -1 (s) Among them, s is the true value.

6. The method for passive positioning path optimization and node selection of unmanned aerial vehicle swarm based on CRLB according to claim 1 is characterized in that: The step S103 comprises the following steps: S31. According to the existence of M target signal sources with different positions on the two-dimensional plane, the number of drones carrying radars participating in passive positioning is N, and the number of groups of drone groups after node selection is M; S32. Clarify the constraints for selecting drone nodes; S33. Select the drone swarm nodes based on CRLB to obtain the best node selection method and grouping scheme.

7. The method for passive positioning path optimization and node selection of a swarm of unmanned aerial vehicles based on CRLB according to claim 6 is characterized in that: The step S33 comprises the following steps: Optimization target, the average value of the positioning CRLB trace of each group of drones for target j after the drone group node selection at this moment: Therefore, the expression of the mathematical model of the drone node selection optimization problem is as follows: stN=λM λ ≥ 3 Among them, N is the number of radar-carrying drones participating in passive positioning, and M is the number of groups in the drone swarm after node selection; The average CRLB trace values ​​of multi-target passive positioning under different node selection schemes are calculated, and the node selection scheme that can minimize the average CRLB trace value of multi-target passive positioning is selected as the optimal scheme at the current moment.

8. The method for passive positioning path optimization and node selection of drone swarm based on CRLB according to claim 1 is characterized in that: The step S104 includes the following steps: S41. According to the movement of the drone group and the target, let the flight time be t, the path optimization time interval be Δt, the drone flight speed be v, the drone heading angle be χ, and the drone turning radius at time t be The minimum turning radius is L min , the minimum approach distance between the UAV and the target is d min ; S42. Calculate the turning radius The turning radius of the drone is S43. The CRLB-based UAV swarm path optimization algorithm realizes real-time and effective configuration of the UAV swarm node locations.

9. The method for passive positioning path optimization and node selection of a swarm of unmanned aerial vehicles based on CRLB according to claim 8 is characterized in that: The step S42 comprises the following steps: Set up the UAV path optimization constraints; Assume the turning radius of the UAV is L i Greater than or equal to the minimum turning radius, The turning arc of the drone is The turning arc is approximately a circular arc, so According to the geometric relationship, a mathematical equation is established, and its expression is as follows: in, is the turning radius of the i-th UAV, is the turning radius.

10. The method for passive positioning path optimization and node selection of a swarm of unmanned aerial vehicles based on CRLB according to claim 8, characterized in that: The step S43 comprises the following steps: The distance between the drone and the target r i Greater than or equal to the minimum approach distance: r i t ≥d min i=0,1,2,...,n-1 Optimization target, the positioning CRLB trace of the drone swarm at the next moment under different drone swarm node position configuration schemes: Tr(CRLB t ) Therefore, the mathematical model of the UAV swarm path optimization problem is expressed as follows: min arg Tr(CRLB t ) r i t ≥d min in, is the turning radius of the i-th UAV, L min is the minimum turning radius, d min The minimum approach distance.

Citation Information

Cited By

  • Unmanned aerial vehicle group grouping obstacle avoidance path intelligent optimization method based on CRLB

    CN119806205A