Maneuvering target tracking method, system, equipment, medium and product

Abstract

The invention relates to the technical field of unmanned aerial vehicle target tracking, and discloses a maneuvering target tracking method, system and device, a medium and a product, and the method comprises the steps: building a plurality of motion models of a maneuvering target, and a volume information filter corresponding to each motion model; filtering state updating is carried out on the multiple cooperative unmanned aerial vehicles, and filtering state information corresponding to each cooperative unmanned aerial vehicle is obtained; carrying out fusion iteration on the filtering state information corresponding to the example cooperative unmanned aerial vehicle and other surrounding cooperative unmanned aerial vehicles to obtain global filtering state information corresponding to the example cooperative unmanned aerial vehicle; and obtaining a global state tracking result of the maneuvering target by the example cooperative unmanned aerial vehicle according to the global filtering state information corresponding to the example cooperative unmanned aerial vehicle. According to the method, the maneuvering target can be dynamically tracked, the tracking precision is high, the tracking result has global observability, and the convergence speed and the robustness are high during collaborative estimation tracking.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) target tracking, and in particular, to a method, system, device, medium, and product for tracking a maneuvering target. Background Art

[0002] Due to the characteristics of light flight and flexible movement, a UAV can estimate the pose and track a maneuvering target by carrying corresponding sensors. When tracking a maneuvering target, establishing an accurate target motion model and constructing a corresponding filter based on this motion model for state estimation are the prerequisites for accurately tracking the maneuvering target. However, since the target will perform uncertain maneuvering actions such as left turns, right turns, and straight-line motions during the maneuvering process, a single motion model is difficult to completely describe its entire motion law. Therefore, the state estimation method based on a single motion model cannot accurately estimate the motion state of the maneuvering target.

[0003] In the related art, first, for the state estimation problem of a non-linear system with a target state dimension greater than 3, most non-linear filtering algorithms such as the Extended Kalman Filter (EKF) and the Unscented Kalman Filter (UKF) are used as local filters, which may have large estimation errors, resulting in problems such as a decrease in tracking accuracy. Second, when using multiple cooperative UAVs to cooperatively track a maneuvering target, there is a lack of analysis of the global observability of the maneuvering target state, and at the same time, there are problems such as slow convergence speed and weak robustness of cooperative estimation and tracking. Third, the cooperative estimation method based on the volume Kalman filter with hybrid consistency uses the volume filter as the local filter and improves the distributed state estimation accuracy in the sensor network by using the hybrid consistency method at the formation level, but does not consider the state estimation problem under the condition of multiple motion models of the maneuvering target. Since a single motion model cannot accurately describe the motion of the maneuvering target during maneuvering, it cannot be used for dynamic tracking of the maneuvering target. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, system, device, medium, and product for tracking a maneuvering target to solve the technical problems in the related art such as the inability to dynamically track a maneuvering target, a decrease in tracking accuracy, a lack of global observability, and slow convergence speed and weak robustness of cooperative estimation and tracking.

[0005] In a first aspect, an embodiment of the present invention provides a method for tracking a maneuvering target, including:

[0006] Establishing a plurality of motion models of the maneuvering target and a volume information filter corresponding to each motion model; wherein, the plurality of motion models include: a straight-line motion model and a turning motion model;

[0007] Based on multiple motion models of a maneuvering target and the cubature information filters corresponding to each motion model, the filtering state updates of multiple cooperative unmanned aerial vehicles (UAVs) are respectively performed to obtain the filtering state information corresponding to each cooperative UAV; wherein, the cooperative UAVs are used to track the maneuvering target; the filtering state information includes the Fisher information state, the information matrix, the contribution vector corresponding to the Fisher information state, and the contribution vector corresponding to the information matrix.

[0008] The filtering state information corresponding to the example cooperative UAV and other surrounding cooperative UAVs is fused and iterated to obtain the global filtering state information corresponding to the example cooperative UAV; the global filtering state information includes the global Fisher information state, the global information matrix, the contribution vector corresponding to the global Fisher information state, and the contribution vector corresponding to the global information matrix.

[0009] Based on the global filtering state information corresponding to the example cooperative UAV, the global state tracking result of the example cooperative UAV for the maneuvering target is obtained.

[0010] In an alternative embodiment, after establishing multiple motion models of the maneuvering target and the cubature information filters corresponding to each motion model, the method further includes:

[0011] Obtain the measurement information of multiple cooperative UAVs at the current moment; wherein, the measurement information of the cooperative UAVs at the current moment includes: the relative distance between the cooperative UAVs and the maneuvering target at the current moment.

[0012] Based on the interacting multiple model (IMM) estimation framework, the target motion model of the maneuvering target and the target cubature information filter corresponding to the target motion model are input interactively to obtain the initial state value and the initial covariance of the target cubature information filter.

[0013] Based on the measurement information of the example cooperative UAV at the current moment, the initial state value and the initial covariance of the target cubature information filter, the independent cubature information filtering update of the target motion model of the maneuvering target is performed to obtain the information for filtering state update; wherein, the information for filtering state update includes the probability weight of the target motion model of the maneuvering target at the current moment and the filtering state information corresponding to the target motion model.

[0014] In an alternative embodiment, based on the interacting multiple model (IMM) estimation framework, inputting the target motion model of the maneuvering target and the target cubature information filter corresponding to the target motion model interactively to obtain the initial state value and the initial covariance of the target cubature information filter includes:

[0015] According to the initialization transition probability between each motion model of the maneuvering target and the probability weight of each motion model of the maneuvering target at the previous moment, calculate the probability weight of each motion model of the maneuvering target after input interaction at the previous moment.

[0016] Based on the probability weights after input interaction, the state vector and information matrix corresponding to the target motion model of the maneuvering target at the previous moment, the initial state value and initial covariance of the target cubature information filter are obtained.

[0017] In an alternative implementation, according to the measurement information at the current moment of the exemplary cooperative UAV, the initial state value and initial covariance of the target cubature information filter, the target motion model of the maneuvering target is updated by independent cubature information filtering to obtain the information for filtering state update, including:

[0018] According to the measurement information at the current moment of the exemplary cooperative UAV, the initial state value and initial covariance of the target cubature information filter corresponding to the target motion model, the target cubature information filter corresponding to the target motion model is updated in time and measurement to obtain the filtered state information corresponding to the target motion model;

[0019] Calculate the likelihood function probability value of each motion model, and calculate the probability weight of the target motion model of the maneuvering target at the current moment according to the likelihood function probability value.

[0020] In an alternative implementation, according to multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, the filtering state of multiple cooperative UAVs is updated respectively to obtain the filtering state information corresponding to each cooperative UAV, including:

[0021] Multiply the probability weight of the target motion model of the maneuvering target at the current moment by the filtered state information corresponding to the target motion model to obtain the filtered state information component corresponding to the target motion model of the maneuvering target;

[0022] Add the filtered state information components corresponding to each motion model of the maneuvering target to obtain the filtered state information corresponding to the exemplary cooperative UAV.

[0023] In an alternative implementation, the filtered state information corresponding to the exemplary cooperative UAV and the filtered state information corresponding to other surrounding cooperative UAVs are fused and iterated to obtain the global filtered state information corresponding to the exemplary cooperative UAV, including:

[0024] Establish a multi-UAV tracking target network model based on the communication topology connection structure;

[0025] According to the node position of the exemplary cooperative UAV in the multi-UAV tracking target network model, determine the cooperative UAVs corresponding to other nodes around the node of the exemplary cooperative UAV;

[0026] Perform measurement and information hybrid consistency fusion iteration on the filtering state information of the cooperative UAVs corresponding to other nodes to obtain the global Fisher information state and global information matrix of the example cooperative UAV;

[0027] The measurement and information hybrid consistency fusion iteration includes:

[0028]

[0029]

[0030] Among them, is the adjacent connected domain of node s corresponding to the example cooperative UAV, and π st is the weight factor of the connectivity between node s corresponding to the example cooperative UAV and other nodes t around node s; l is the number of iterations; is the Fisher information state of the cooperative UAV of node t; is the information matrix of the cooperative UAV of node t; is the contribution vector corresponding to the Fisher information state of the cooperative UAV of node t; is the contribution vector corresponding to the information matrix of the cooperative UAV of node t; is the global Fisher information state of the example cooperative UAV of node s; is the global information matrix of the example cooperative UAV of node s; is the contribution vector corresponding to the global Fisher information state of the example cooperative UAV of node s; is the contribution vector corresponding to the global information matrix of the example cooperative UAV of node s.

[0031] In a second aspect, an embodiment of the present invention provides a maneuvering target tracking system, and the system includes:

[0032] A multi-motion model and filter construction module, configured to establish multiple motion models of a maneuvering target and a cubature information filter corresponding to each motion model; wherein, the multiple motion models include: a linear motion model and a turning motion model;

[0033] A filtering state update module, configured to respectively update the filtering states of multiple cooperative UAVs according to the multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model to obtain the filtering state information corresponding to each cooperative UAV; wherein, the cooperative UAVs are used to track the maneuvering target; the filtering state information includes Fisher information state, information matrix, contribution vector corresponding to the Fisher information state, and contribution vector corresponding to the information matrix;

[0034] The fusion and iteration module is used to fuse and iterate the filtered state information corresponding to the example cooperative UAV and other surrounding cooperative UAVs to obtain the global filtered state information corresponding to the example cooperative UAV; the global filtered state information includes the global Fisher information state, the global information matrix, the contribution vector corresponding to the global Fisher information state, and the contribution vector corresponding to the global information matrix.

[0035] The tracking result output module is used to obtain the global state tracking result of the example cooperative UAV for the maneuvering target according to the global filtered state information corresponding to the example cooperative UAV.

[0036] In a third aspect, an embodiment of the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the method according to the first aspect or any corresponding implementation manner thereof.

[0037] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, on which computer instructions are stored. The computer instructions are used to cause a computer to execute the method according to the first aspect or any corresponding implementation manner thereof.

[0038] In a fifth aspect, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the method according to the first aspect or any corresponding implementation manner thereof.

[0039] In the embodiment of the present invention, first, a multi-motion model is established for the maneuvering target, which can accurately describe the complete motion state of the maneuvering target; second, each motion model uses the cubature information filtering algorithm for maneuvering target state estimation and tracking, which can improve the tracking accuracy of the local filter; finally, at the level of UAV formation cooperative tracking, the filtered state information corresponding to other cooperative UAVs around the example cooperative UAV is fused and iterated to obtain the global filtered state information corresponding to the example cooperative UAV, further improving the convergence speed and robustness during multi-cooperative UAV tracking, and thus improving the tracking accuracy of the UAV formation for non-cooperative maneuvering targets. Description of the Drawings

[0040] In order to more clearly illustrate the specific implementation manners of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific implementation manners or the prior art. Obviously, the following drawings are some embodiments of the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 It is a flowchart of a method for tracking a maneuvering target provided by an embodiment of the present invention;

[0042] Figure 2 It is a schematic flow chart of another maneuvering target tracking method provided by an embodiment of the present invention;

[0043] Figure 3 It is a curve graph of the motion trajectory of a maneuvering target obtained by 100 Monte Carlo simulations in an embodiment of the present invention;

[0044] Figure 4 It is a probability curve graph of a motion model obtained by 100 Monte Carlo simulations;

[0045] Figure 5 It is a comparison curve graph of position errors obtained by 100 Monte Carlo simulations in an embodiment of the present invention;

[0046] Figure 6 It is a comparison curve graph of velocity errors obtained by 100 Monte Carlo simulations in an embodiment of the present invention;

[0047] Figure 7 It is a structural block diagram of a maneuvering target tracking system provided by an embodiment of the present invention;

[0048] Figure 8 It is a schematic structural diagram of a computer device provided by an embodiment of the present invention. Detailed implementation manners

[0049] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0050] Figure 1 It is a schematic flow chart of a maneuvering target tracking method according to an embodiment of the present invention. As Figure 1 shown, the flow includes the following steps:

[0051] Step S101, establish multiple motion models of the maneuvering target and a cubature information filter corresponding to each motion model; wherein, the multiple motion models include: a linear motion model and a turning motion model.

[0052] In step S101, the motion direction of the maneuvering target may include straight line, right turn and left turn; the motion speed of the maneuvering target may include uniform speed, acceleration, deceleration and variable speed. The motion models may include a uniform linear motion model, accelerating linear motion, uniform left turn motion, decelerating left turn motion, uniform right turn motion, etc., which are not specifically limited herein.

[0053] In one example, a maneuvering target moves in a two-dimensional plane, including three types of motion patterns: uniform linear motion, uniform left-turning motion, and uniform right-turning motion, which are respectively denoted as motion model 1, motion model 2, and motion model 3;

[0054] The state vector x corresponding to the maneuvering target is composed of the position and velocity in the two-dimensional plane, and is expressed as:

[0055]

[0056] where x and y are respectively the abscissa and ordinate in the two-dimensional plane, and respectively represent the velocities in the abscissa and ordinate directions.

[0057] Based on the state vector x corresponding to the maneuvering target, a uniform linear motion model f 1 (x k ), and a uniform turning motion model f 2 (x k ) are established for each cooperative unmanned aerial vehicle.

[0058] x k = f r (x k-1 ) + w k , r = 1, 2,..., m

[0059] where m is the total number of motion models describing the motion of the maneuvering target; w k is the system noise; k is the time sequence of the data acquisition period.

[0060] Specifically, the uniform linear motion model f 1 (x k ) and the uniform turning motion model f 2 (x k ) can be expressed as:

[0061]

[0062]

[0063] where ω is the angular rate; when ω > 0, it represents left-turning motion; when ω < 0, it represents right-turning motion; Δt represents the data acquisition period, and preferably Δt = 1.

[0064] Step S102: According to multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, perform filtering state updates on multiple cooperative unmanned aerial vehicles (UAVs) respectively to obtain the filtering state information corresponding to each cooperative UAV; wherein, the cooperative UAVs are used to track the maneuvering target; the filtering state information includes the Fisher information state, the information matrix, the contribution vector corresponding to the Fisher information state, and the contribution vector corresponding to the information matrix.

[0065] Step S103: Perform fusion iteration on the filtering state information corresponding to the exemplary cooperative UAV and other surrounding cooperative UAVs to obtain the global filtering state information corresponding to the exemplary cooperative UAV; the global filtering state information includes the global Fisher information state, the global information matrix, the contribution vector corresponding to the global Fisher information state, and the contribution vector corresponding to the global information matrix.

[0066] In step S103, for any one of the multiple cooperative UAVs, the value error of observing the maneuvering target is relatively large, which leads to inaccurate filtering state information generated. This is because the observation position, observation angle, etc. of the cooperative UAV are limited. Therefore, by performing fusion iteration on the filtering state information corresponding to the exemplary cooperative UAV and other surrounding cooperative UAVs, information fusion is performed on multiple cooperative UAVs at the formation level, making the sensor information of the exemplary cooperative UAV richer, and further improving the tracking accuracy of the exemplary cooperative UAV for the maneuvering target.

[0067] Step S104: Obtain the global state tracking result of the exemplary cooperative UAV for the maneuvering target according to the global filtering state information corresponding to the exemplary cooperative UAV.

[0068] In the embodiment of the present invention, first, multiple motion models are established for the maneuvering target, which can accurately describe the complete motion state of the maneuvering target; second, each motion model uses the cubature information filtering algorithm for maneuvering target state estimation and tracking, which can improve the tracking accuracy of the local filter; finally, at the level of UAV formation cooperative tracking, fusion iteration is performed on the filtering state information corresponding to other cooperative UAVs around the exemplary cooperative UAV to obtain the global filtering state information corresponding to the exemplary cooperative UAV, further improving the convergence speed and robustness during multi-cooperative UAV tracking, and thereby improving the tracking accuracy of the UAV formation for non-cooperative maneuvering targets.

[0069] In an alternative embodiment, after establishing multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, the method further includes:

[0070] Step a: Obtain the measurement information of multiple cooperative UAVs at the current moment; wherein, the measurement information of the cooperative UAVs at the current moment includes the relative distance between the cooperative UAV at the current moment and the maneuvering target.

[0071] Among them, the information of the relative distance can be obtained by using the cameras equipped on the cooperative UAVs as relative position measurement devices.

[0072] As an example, the relative distance information between each cooperative UAV and the maneuvering target is selected as the measurement vector and a non - linear measurement model h s (x k ) is established:

[0073]

[0074] Among them, h s (x k ) is the non - linear measurement model; N is the total number of cooperative UAVs describing the motion state of the maneuvering target; is the measurement noise.

[0075]

[0076] Among them, are respectively the horizontal position information of the example cooperative UAV where the node s is located; x k and y k are respectively the horizontal position information of the example cooperative UAV at the k - th moment.

[0077] Step b: Based on the interactive multiple model estimation framework, input - interact the target motion model of the maneuvering target and the target volume information filter corresponding to the target motion model to obtain the initial state value and the initial covariance of the target volume information filter.

[0078] Among them, in the embodiment of the present invention, based on the interactive multiple model estimation framework, information fusion is performed between each motion model and the corresponding volume information filter to achieve input - interaction.

[0079] In an alternative embodiment, step b includes:

[0080] Step b1: Calculate the probability weights of each motion model of the maneuvering target after input - interaction according to the initialization transition probability between each motion model of the maneuvering target and the probability weights of each motion model of the maneuvering target at the previous moment.

[0081] Specifically, the probability weights of each motion model of the maneuvering target after input - interaction can be expressed as:

[0082]

[0083] Among them, is the probability normalization constant, m is the total number of motion models, p iris the initialization transition probability from the motion model u to the motion model r without considering the constraint conditions. is the probability weight of the motion model i of the maneuvering target corresponding to the node s at the previous moment.

[0084] It should be noted that the above formula is the probability weight of the motion model r of the maneuvering target at the previous moment after input interaction. The probability weights of other motion models after input interaction are calculated in the same way and will not be elaborated here.

[0085] Step b2: Obtain the initial state value and initial covariance of the target cubature information filter according to the probability weight of the target motion model of the maneuvering target at the previous moment after input interaction, the state vector and information matrix corresponding to the target motion model of the maneuvering target at the previous moment.

[0086] Specifically, the initial state value of the target cubature information filter and the initial covariance are expressed as:

[0087]

[0088] where are respectively the state vector and information matrix corresponding to the motion model r of the maneuvering target corresponding to the node s at the previous moment, is the probability weight of the motion model r of the maneuvering target at the previous moment after input interaction.

[0089] It should be noted that the above formula represents the initial state value and initial covariance of the target cubature information filter corresponding to the motion model r of the maneuvering target. The initial state values and initial covariances of the target cubature information filters corresponding to other motion models of the maneuvering target, and the initial state values and initial covariances of the target cubature information filters corresponding to the motion models of other cooperative UAVs are calculated in the same way and will not be elaborated here.

[0090] Step c: Perform independent cubature information filtering update on the target motion model of the maneuvering target according to the measurement information at the current moment of the exemplary cooperative UAV, the initial state value and initial covariance of the target cubature information filter, and obtain the information for filtering state update; where the information for filtering state update includes the probability weight of the target motion model of the maneuvering target at the current moment and the filtering state information corresponding to the target motion model.

[0091] In an optional implementation manner, step c includes:

[0092] Step c1: Based on the measurement information of the collaborative UAV at the current moment, the initial state value and the initial covariance of the target volume information filter corresponding to the target motion model, perform time update and measurement update on the target volume information filter corresponding to the target motion model to obtain the filtered state information corresponding to the target motion model.

[0093] Specifically, establishing the volume information filter and the time update of the volume information filter include:

[0094]

[0095] Among them, respectively represent the sigma sample points, the weights corresponding to the sigma sample points, the volume point set, the predicted value of the initial state value, and the predicted value of the initial covariance; respectively represent the motion model r and the measurement model of the collaborative UAV corresponding to node s.

[0096] The measurement update of the volume information filter includes:

[0097]

[0098] Among them, the filtered state information corresponding to the motion model r includes: Fisher information state information matrix contribution vector corresponding to the Fisher information state contribution vector corresponding to the information matrix

[0099] Step c2: Calculate the likelihood function probability value of each motion model, and calculate the probability weight of the target motion model of the maneuvering target at the current moment according to the likelihood function probability value.

[0100] Specifically, the probability weight of the target motion model of the maneuvering target at the current moment can be expressed as:

[0101]

[0102] Among them, is the probability weight of the motion model r of the maneuvering target corresponding to node s at the current moment. The other motion models are calculated in the same way and will not be elaborated here.

[0103] In an alternative embodiment, step S102 includes:

[0104] Step S1021: Multiply the probability weight of the target motion model of the maneuvering target at the current moment by the filtered state information corresponding to the target motion model to obtain the filtered state information component corresponding to the target motion model of the maneuvering target.

[0105] Step S1022: Add the filtering state information components corresponding to each motion model of the maneuvering target to obtain the filtering state information corresponding to the exemplary cooperative UAV.

[0106] In the embodiment of the present invention, the probability weight of the target motion model of the maneuvering target at the current moment and the filtering state information corresponding to the target motion model are output and interacted to obtain the filtering state information corresponding to the exemplary cooperative UAV.

[0107] Specifically, use the probability weight of the motion model r of the maneuvering target at the current moment and the filtering state information corresponding to the motion model r: Fisher information state information matrix contribution vector corresponding to the Fisher information state and the contribution vector corresponding to the information matrix for output interaction to obtain the filtering state information corresponding to the exemplary cooperative UAV.

[0108] The filtering state corresponding to the exemplary cooperative UAV includes: Fisher information state information matrix contribution vector corresponding to the Fisher information state and the contribution vector corresponding to the information matrix The specific expression form is as follows:

[0109]

[0110] In an alternative embodiment, the filtering state information corresponding to the motion model r of the maneuvering target at the current moment, that is, the Fisher information state information matrix contribution vector corresponding to the Fisher information state and the contribution vector corresponding to the information matrix are used to obtain Then, and are brought into steps b1 and b2 for iteration, thereby completing the filtering state update of multiple cooperative UAV nodes in the interactive multiple model filtering framework.

[0111] In an alternative embodiment, step S103 includes:

[0112] Step S1031: Establish a tracking target network model for multiple cooperative UAVs based on the communication topology connection structure.

[0113] In step S1031, in the multi-vehicle cooperative UAV target tracking network model, each node represents a cooperative UAV, and each cooperative UAV can perform information interaction with other node cooperative UAVs according to its own node position. The node s described above corresponds to any one of the multi-vehicle cooperative UAVs.

[0114] Step S1032: Determine the cooperative UAVs corresponding to the other nodes around the node of the example cooperative UAV according to the node position of the example cooperative UAV in the multi-vehicle cooperative UAV target tracking network model.

[0115] Step S1033: Perform measurement and information mixing consistent fusion iteration on the filtering state information of the cooperative UAVs corresponding to the other nodes to obtain the global Fisher information state and global information matrix corresponding to the example cooperative UAV.

[0116] The measurement and information mixing consistent fusion iteration includes:

[0117]

[0118]

[0119] Among them, is the adjacent connected domain of the node s corresponding to the example cooperative UAV, and π st is the weight factor of the connectivity between the node s corresponding to the example cooperative UAV and other nodes t around the node s; l is the number of iterations; is the Fisher information state of the cooperative UAV of the node t; is the information matrix of the cooperative UAV of the node t; is the contribution vector corresponding to the Fisher information state of the cooperative UAV of the node t; is the contribution vector corresponding to the information matrix of the cooperative UAV of the node t; is the global Fisher information state of the example cooperative UAV of the node s; is the global information matrix of the example cooperative UAV of the node s; is the contribution vector corresponding to the global Fisher information state of the example cooperative UAV of the node s; is the contribution vector corresponding to the global information matrix of the example cooperative UAV of the node s.

[0120] In the embodiment of the present invention, at the UAV formation level, based on the communication topology connection structure, measurement consistency and information consistency fusion iteration are performed on multiple cooperative UAVs to obtain the global Fisher information state and global information matrix corresponding to the example cooperative UAV, so that the example cooperative UAV has global observability, effectively improving the formation cooperative convergence speed and robustness, and further improving the tracking accuracy of the UAV formation for non-cooperative maneuvering targets.

[0121] In an alternative embodiment, the global state tracking result in step S104 can be expressed as:

[0122]

[0123] where N s represents the number of connections of node s, and L is the total number of iterations; is the Fisher information state corresponding to the example cooperative UAV, is the information matrix corresponding to the example cooperative UAV, is the contribution vector corresponding to the Fisher information state of the example cooperative UAV, is the contribution vector corresponding to the information matrix of the example cooperative UAV.

[0124] As Figure 2 shown, to verify the effectiveness of the present invention, the exciting target of the embodiment of the present invention is further described:

[0125] The proposed maneuvering target tracking method of the present invention is simulated 100 times by Monte Carlo in the UAV formation target tracking system using Matlab, and the simulation comparison is carried out with the unscented information filtering based on the interacting multiple models (UIF-IMM) and the cubature information filtering method based on the interacting multiple models (CIF-IMM).

[0126] First, it is assumed that the maneuvering target moves in a two-dimensional plane, including three types of motion modes: uniform linear motion, uniform left turn, and uniform right turn, which are respectively denoted as motion model 1, motion model 2, and motion model 3; the motion state vector of the maneuvering target consists of the position and velocity in the two-dimensional plane. The relative distance information between the cooperative UAV and the maneuvering target is selected as the measurement vector, and a multi-motion model and a measurement model including the linear motion, left turn, and right turn of the maneuvering target are established:

[0127]

[0128] where f 1 (x k ), f 2 (x k ), h s (x k ) represent the uniform linear motion model, the uniform rate turning motion model, and the non-linear measurement model respectively; m and N represent the total number of models describing the motion of the maneuvering target and the total number of cooperative UAVs respectively.

[0129]

[0130] where ω > 0 represents a left turn motion and ω < 0 represents a right turn motion; It represents the horizontal position information of the node s corresponding to the cooperative UAV. Δt is the data acquisition period, and preferably Δt = 1; w k and respectively represent the system noise and the measurement noise, and the corresponding covariances are denoted as Q k 、R k 。

[0131]

[0132]

[0133] In addition, the simulation time is set to 180 s, and the target motion speed The left turning angular rate ω = π / 45 rad / s, and the right turning angular rate ω = -π / 36 rad / s;

[0134] The time series of the three motion models are set as shown in the following table:

[0135]

[0136] And the initial switching probabilities between the three motion models are:

[0137]

[0138] Then, the interactive multiple model cubature information filtering is used to update the local node of the cooperative UAV. The specific process includes:

[0139] (1) Use the input interaction to calculate the initial values of the state and covariance of the cubature information filter and

[0140]

[0141] (2) Perform the time update of the cubature information filtering:

[0142]

[0143]

[0144] (3) Perform the measurement update of the cubature information filtering:

[0145]

[0146] (4) Use the output interaction to update the local node information state information matrix and the contribution vector and

[0147]

[0148] Finally, at the level of the UAV formation, with the aid of the communication topology network, the measurement and information consistency method is used for interactive iterative update, and the number of iterations is set to L = 4.

[0149]

[0150]

[0151] Therefore, after the iteration ends, the cooperative estimation result of the UAV formation is obtained:

[0152]

[0153] Simulation results:

[0154] According to the above simulation conditions, the simulation results are as Figures 3 - 6 shown.

[0155] As can be Figure 3 seen, the maneuvering target has carried out various maneuvering motions such as straight-line motion, left turn, and right turn. The maneuvering target tracking method of the embodiment of the present invention can well track the motion trajectory of the maneuvering target. Further, Figure 4 the estimation results of the motion model probabilities of the maneuvering target at different times by the maneuvering target tracking method of the embodiment of the present invention during the simulation process are given. The curve results are consistent with the simulation settings of the present invention and can accurately estimate the motion model of the maneuvering target. Figures 5 - 6 The position tracking error and speed tracking error results of the method of the present invention, the unscented information filtering based on the interactive multiple model (UIF-IMM), and the cubature information filtering method based on the interactive multiple model (CIF-IMM) are given respectively. It can be seen from this that the maneuvering target tracking method of the embodiment of the present invention has the fastest convergence speed and the smallest tracking error. In summary, the method of the present invention can accurately identify different motion models of the maneuvering target, and at the same time has a faster convergence speed and higher tracking accuracy.

[0156] The embodiment of the present invention has the following advantages:

[0157] In the embodiments of the present invention, a multi-motion model is first established for a maneuvering target, which can accurately describe the complete motion state of the maneuvering target. Secondly, each motion model uses the cubature information filtering algorithm for state estimation and tracking of the maneuvering target, which can improve the tracking accuracy of the local filter. At the same time, based on the interactive multi-model estimation framework, the target motion model of the maneuvering target and the target cubature information filter corresponding to the target motion model are input and interacted. Finally, at the level of UAV formation cooperative tracking, the filtering state information corresponding to other cooperative UAVs around the example cooperative UAV is fused and iterated to obtain the global filtering state information corresponding to the example cooperative UAV, further improving the convergence speed and robustness during multi-cooperative UAV tracking, and thus improving the tracking accuracy of the UAV formation for non-cooperative maneuvering targets.

[0158] An embodiment of the present invention provides a maneuvering target tracking system, as Figure 7 shown, including:

[0159] A multi-motion model and filter construction module 701, configured to establish multiple motion models of a maneuvering target and a cubature information filter corresponding to each motion model; wherein, the multiple motion models include: a straight-line motion model and a turning motion model.

[0160] A filtering state update module 702, configured to respectively update the filtering states of multiple cooperative UAVs according to the multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, to obtain the filtering state information corresponding to each cooperative UAV; wherein, the cooperative UAVs are used to track the maneuvering target; the filtering state information includes Fisher information state, information matrix, contribution vector corresponding to the Fisher information state, and contribution vector corresponding to the information matrix.

[0161] A fusion iteration module 703, configured to fuse and iterate the filtering state information corresponding to the example cooperative UAV and other cooperative UAVs around it, to obtain the global filtering state information corresponding to the example cooperative UAV; the global filtering state information includes global Fisher information state, global information matrix, contribution vector corresponding to the global Fisher information state, and contribution vector corresponding to the global information matrix.

[0162] A tracking result output module 704, configured to obtain the global state tracking result of the example cooperative UAV for the maneuvering target according to the global filtering state information corresponding to the example cooperative UAV.

[0163] In an alternative embodiment, the maneuvering target tracking system further includes:

[0164] A measurement information acquisition module 705, configured to acquire the measurement information of multiple cooperative UAVs at the current moment; wherein, the measurement information of the cooperative UAVs at the current moment includes: the relative distance between the cooperative UAVs and the maneuvering target at the current moment.

[0165] A state and covariance acquisition module 706, configured to input and interact a target motion model of a maneuvering target and a target volume information filter corresponding to the target motion model based on an interactive multiple model estimation framework, so as to obtain an initial state value and an initial covariance of the target volume information filter.

[0166] An independent volume information filtering update module 707, configured to perform independent volume information filtering update on the target motion model of the maneuvering target according to the measurement information at the current moment of the example cooperative unmanned aerial vehicle, the initial state value and the initial covariance of the target volume information filter, so as to obtain information for filtering state update; wherein, the information for filtering state update includes the probability weight of the target motion model of the maneuvering target at the current moment and the filtering state information corresponding to the target motion model.

[0167] In an alternative embodiment, the state and covariance acquisition module 706 includes:

[0168] A probability weight calculation unit 7061, configured to calculate the probability weight of each motion model of the maneuvering target after input interaction according to the initialization transition probability between each motion model of the maneuvering target and the probability weight of each motion model of the maneuvering target at the previous moment.

[0169] A state and covariance acquisition unit 7062, configured to obtain the initial state value and the initial covariance of the target volume information filter according to the probability weight of the target motion model of the example cooperative unmanned aerial vehicle after input interaction at the previous moment, the state vector and the information matrix corresponding to the target motion model of the example cooperative unmanned aerial vehicle at the previous moment.

[0170] In an alternative embodiment, the independent volume information filtering update module 707 includes:

[0171] A time and measurement update unit 7011, configured to perform time update and measurement update on the target volume information filter corresponding to the target motion model according to the measurement information at the current moment of the example cooperative unmanned aerial vehicle, the initial state value and the initial covariance of the target volume information filter corresponding to the target motion model, so as to obtain the filtering state information corresponding to the target motion model;

[0172] A probability weight update unit 7012, configured to calculate the likelihood function probability value of each motion model, and calculate the probability weight of the target motion model of the maneuvering target at the current moment according to the likelihood function probability value.

[0173] In an alternative embodiment, the filtering state update module 702 includes:

[0174] The filtering state information component obtaining unit 7021 is configured to multiply the probability weight of the target motion model of the maneuvering target at the current moment by the filtering state information corresponding to the target motion model to obtain the filtering state information component corresponding to the target motion model of the maneuvering target;

[0175] The filtering state information obtaining unit 7022 is configured to add up the filtering state information components corresponding to each motion model of the maneuvering target to obtain the filtering state information corresponding to the example cooperative UAV.

[0176] In an optional implementation manner, the fusion iteration module 703 includes:

[0177] The network construction unit 7031 is configured to establish a multi-UAV cooperative target tracking network model based on the communication topology connection structure;

[0178] The surrounding node determination unit 7032 is configured to determine the cooperative UAVs corresponding to the other nodes around the node of the example cooperative UAV according to the node position of the example cooperative UAV in the multi-UAV cooperative target tracking network model;

[0179] The information fusion iteration unit 7033 is configured to perform measurement and information hybrid consistency fusion iteration on the filtering state information of the cooperative UAVs corresponding to the other nodes to obtain the global Fisher information state and the global information matrix corresponding to the example cooperative UAV.

[0180] Wherein, the measurement and information hybrid consistency fusion iteration includes:

[0181]

[0182] Wherein, is the adjacent connected domain of the node s corresponding to the example cooperative UAV, and π st is the weight factor of the connectivity between the node s corresponding to the example cooperative UAV and the other node t around the node s; l is the number of iterations; is the Fisher information state of the cooperative UAV of the node t; is the information matrix of the cooperative UAV of the node t; is the contribution vector corresponding to the Fisher information state of the cooperative UAV of the node t is the contribution vector corresponding to the information matrix of the cooperative UAV of the node t; is the global Fisher information state of the example cooperative UAV of the node s; is the global information matrix of the example cooperative UAV of the node s; is the contribution vector corresponding to the global Fisher information state of the example cooperative UAV of the node s; is the contribution vector corresponding to the global information matrix of the example cooperative UAV of the node s.

[0183] The further function descriptions of the above-mentioned respective modules and units are the same as those in the corresponding embodiments above, and will not be elaborated here.

[0184] A maneuvering target tracking system in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0185] An embodiment of the present invention further provides a computer device having the above-mentioned Figure 7 shown maneuvering target tracking system.

[0186] Please refer to Figure 8 , this computer device includes: one or more processors 10, a memory 20, and an interface for connecting each component, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as an array of servers, a set of blade servers, or a multi-processor system). Figure 8 Take one processor 10 as an example in

[0187] The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above-mentioned hardware chip can be an application specific integrated circuit, a programmable logic device, or a combination thereof. The above-mentioned programmable logic device can be a complex programmable logic device, a field programmable gate array, a generic array logic, or any combination thereof.

[0188] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0189] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely provided with respect to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0190] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above types of memories.

[0191] The computer device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 can be connected through a bus or other means. Figure 3 Taking connection through a bus as an example.

[0192] The input device 30 can receive input digital or character information, and generate key signal inputs related to the user settings and function controls of the computer device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (such as an LED), and a haptic feedback device (such as a vibration motor), etc. The above-mentioned display device includes but is not limited to a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device may be a touch screen.

[0193] Embodiments of the present invention also provide a computer-readable storage medium. The methods according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code that is originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the methods described herein can be stored as such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the methods shown in the above embodiments are implemented.

[0194] A part of the embodiments of the present invention can be applied as a computer program product, for example, computer program instructions, which when executed by a computer, can call or provide the methods and / or technical solutions according to the present invention through the operations of the computer. Those skilled in the art should be able to understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible by the computer.

[0195] Although the embodiments of the present invention are described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for maneuvering target tracking, characterized in that, The method includes: Establishing multiple motion models of a maneuvering target and a cubature information filter corresponding to each motion model; wherein, the multiple motion models include: a linear motion model and a turning motion model; According to the multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, respectively updating the filtering states of multiple cooperative UAVs to obtain the filtering state information corresponding to each cooperative UAV; wherein, the cooperative UAVs are used to track the maneuvering target; the filtering state information includes Fisher information state, information matrix, contribution vector corresponding to the Fisher information state, and contribution vector corresponding to the information matrix; Performing fusion iteration on the filtering state information corresponding to the example cooperative UAV and other surrounding cooperative UAVs to obtain the global filtering state information corresponding to the example cooperative UAV; the global filtering state information includes global Fisher information state, global information matrix, contribution vector corresponding to the global Fisher information state, and contribution vector corresponding to the global information matrix; According to the global filtering state information corresponding to the example cooperative UAV, obtaining the global state tracking result of the example cooperative UAV for the maneuvering target.

2. The method according to claim 1, wherein After establishing the multiple motion models of the maneuvering target and the cubature information filter corresponding to each motion model, the method further includes: Obtaining the measurement information of multiple cooperative UAVs at the current moment; wherein, the measurement information of the cooperative UAVs at the current moment includes: the relative distance between the cooperative UAVs and the maneuvering target at the current moment; Based on the interactive multiple model estimation framework, inputting and interacting the target motion model of the maneuvering target and the target cubature information filter corresponding to the target motion model to obtain the initial state value and initial covariance of the target cubature information filter; According to the measurement information of the example cooperative UAV at the current moment, the initial state value and initial covariance of the target cubature information filter, independently updating the target motion model of the maneuvering target by cubature information filtering to obtain the information for filtering state update; wherein, the information for filtering state update includes the probability weight of the target motion model of the maneuvering target at the current moment and the filtering state information corresponding to the target motion model; 3. The method according to claim 2, wherein The inputting and interacting the target motion model of the maneuvering target and the target cubature information filter corresponding to the target motion model based on the interactive multiple model estimation framework to obtain the initial state value and initial covariance of the target cubature information filter includes: Calculating the probability weight of each motion model of the maneuvering target after input interaction at the previous moment according to the initialization transition probability between each motion model of the maneuvering target and the probability weight of each motion model of the maneuvering target at the previous moment; According to the probability weight of the target motion model of the maneuvering target after input interaction at the previous moment, the state vector and information matrix corresponding to the target motion model of the maneuvering target at the previous moment, obtaining the initial state value and initial covariance of the target cubature information filter.

4. The method according to claim 2 or 3, characterized in that, The independently updating the target motion model of the maneuvering target by cubature information filtering according to the measurement information of the example cooperative UAV at the current moment, the initial state value and initial covariance of the target cubature information filter to obtain the information for filtering state update includes: Based on the measurement information at the current moment of the collaborative UAV according to the example, the initial state value and the initial covariance of the target volume information filter corresponding to the target motion model, perform time update and measurement update on the target volume information filter corresponding to the target motion model to obtain the filtered state information corresponding to the target motion model; Calculate the likelihood function probability value of each motion model, and calculate the probability weight of the target motion model of the maneuvering target at the current moment according to the likelihood function probability value.

5. The method according to claim 4, wherein The filtering state update of multiple collaborative UAVs is respectively performed according to multiple motion models of the maneuvering target and the volume information filter corresponding to each motion model to obtain the filtering state information corresponding to each collaborative UAV, including: Multiply the probability weight of the target motion model of the maneuvering target at the current moment by the filtered state information corresponding to the target motion model to obtain the filtered state information component corresponding to the target motion model of the maneuvering target; Add the filtered state information components corresponding to each motion model of the maneuvering target to obtain the filtered state information corresponding to the collaborative UAV according to the example.

6. The method according to claim 1, wherein The fusion iteration of the filtered state information corresponding to the collaborative UAV according to the example and other surrounding collaborative UAVs to obtain the global filtered state information corresponding to the collaborative UAV according to the example includes: Establish a tracking target network model of multiple collaborative UAVs based on the communication topology connection structure; According to the node position corresponding to the collaborative UAV according to the example in the tracking target network model of multiple collaborative UAVs, determine the collaborative UAVs corresponding to other nodes around the node of the collaborative UAV according to the example; Perform measurement and information mixing consistent fusion iteration on the filtered state information of the collaborative UAVs corresponding to other nodes to obtain the global Fisher information state and the global information matrix corresponding to the collaborative UAV according to the example; The measurement and information mixing consistent fusion iteration includes: Among them, is the adjacent connected domain of node s corresponding to the example cooperative UAV, and π st is the weight factor of the connectivity between node s corresponding to the example cooperative UAV and other nodes t around node s; l is the number of iterations; is the Fisher information state of the cooperative UAV of node t; is the information matrix of the cooperative UAV of node t; is the contribution vector corresponding to the Fisher information state of the cooperative UAV of node t; is the contribution vector corresponding to the information matrix of the cooperative UAV of node t; is the global Fisher information state of the example cooperative UAV of node s; is the global information matrix of the example cooperative UAV of node s; is the contribution vector corresponding to the global Fisher information state of the example cooperative UAV of node s; is the contribution vector corresponding to the global information matrix of the example cooperative UAV of node s.

7. A maneuvering target tracking system, characterized in that, The system includes: A multi-motion model and filter construction module, which is used to establish multiple motion models of the maneuvering target and the volume information filter corresponding to each motion model; among them, the multiple motion models include: a straight-line motion model and a turning motion model; A filtering state update module, which is used to perform filtering state update on multiple collaborative UAVs respectively according to multiple motion models of the maneuvering target and the volume information filter corresponding to each motion model to obtain the filtering state information corresponding to each collaborative UAV; among them, the collaborative UAV is used to track the maneuvering target; the filtering state information includes the Fisher information state, the information matrix, the contribution vector corresponding to the Fisher information state, and the contribution vector corresponding to the information matrix; A fusion iteration module, which is used to perform fusion iteration on the filtered state information corresponding to the collaborative UAV according to the example and other surrounding collaborative UAVs to obtain the global filtered state information corresponding to the collaborative UAV according to the example; the global filtered state information includes the global Fisher information state, the global information matrix, the contribution vector corresponding to the global Fisher information state, and the contribution vector corresponding to the global information matrix; A tracking result output module, which is used to obtain the global state tracking result of the maneuvering target by the collaborative UAV according to the example according to the global filtered state information corresponding to the collaborative UAV according to the example.

8. A computer device, characterized in that, Including: A memory and a processor, which are communicatively connected to each other. Computer instructions are stored in the memory, and the processor executes the computer instructions to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to perform the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, It includes computer instructions, and the computer instructions are used to cause a computer to perform the method according to any one of claims 1 to 6.

Images