Optoelectronic pod video tracking switching method based on maneuvering target guidance information estimation
By using motion state estimation and adaptive prediction compensation algorithms for maneuvering targets, the tracking error problem of the electro-optical pod during load switching was solved, and the electro-optical pod achieved accurate and stable tracking of maneuvering targets.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING RES INST OF SPATIAL MECHANICAL & ELECTRICAL TECH
- Filing Date
- 2025-12-12
- Publication Date
- 2026-07-10
Smart Images
Figure CN122363352A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a video tracking switching method for an electro-optical pod based on the estimation of maneuvering target guidance information, belonging to the field of airborne electro-optical pod maneuvering target tracking technology. Background Technology
[0002] Airborne electro-optical pods integrate visible light cameras, infrared thermal imagers, video radars, and laser rangefinders, serving as crucial mission payloads for unmanned aerial vehicles (UAVs). As airborne electro-optical pods evolve towards multi-functionality and intelligence, they are widely used in maritime search and rescue, and high-precision tracking of mobile ground targets. Advances in hardware technology and radar image processing algorithms have enabled Synthetic Aperture Radar (SAR) detection results to be output as image and video streams. However, due to significant differences in the detection principles and information characteristics between optical imaging and radar video imaging within electro-optical pods, the position of mobile targets in the image can easily shift before and after image switching, making it difficult for the electro-optical pod to accurately track mobile targets during these transitions.
[0003] To address the aforementioned issues, most existing visual servo tracking methods rely on maneuvering target features or miss distance information for tracking. However, existing visual servo tracking methods are susceptible to the influence of factors such as lighting variations and atmospheric visibility when extracting maneuvering target features, resulting in indistinct maneuvering target features and thus hindering maneuvering target tracking. Similarly, when using miss distance information for image tracking, the miss distance information is prone to cumulative errors and is easily affected by background interference, leading to poor image tracking performance. Furthermore, random vibrations of the aircraft, insufficient servo tracking control bandwidth, and delays in outputting maneuvering target miss distances also increase the probability of losing maneuvering targets. Summary of the Invention
[0004] The technical problem solved by this invention is to overcome the shortcomings of existing technologies and provide a video tracking and switching method for an electro-optical pod based on the estimation of maneuvering target guidance information. This method uses the motion state information of the maneuvering target itself to assist the electro-optical pod in tracking the maneuvering target during the switching of different load data, thereby improving the tracking performance of the electro-optical pod and solving the problem that the electro-optical pod has difficulty accurately tracking the maneuvering target before and after image switching.
[0005] The technical solution of this invention is: A method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information, comprising the following steps: (1) After the electro-optical pod identifies the target, it calculates in real time the estimated value of the motion state of the maneuvering target tracked by the electro-optical pod in the geodetic coordinate system. ,in , and These represent the maneuvering target in the geodetic coordinate system. k The estimated values of position, velocity, and acceleration at each instant; (2) Using the estimated value of the motion state of the maneuvering target in the geodetic coordinate system Calculate the pitch angular velocity of the inertial stabilized platform equipped with the optoelectronic pod. and angular velocity of the orientation frame And the pitch angular velocity of the inertial stabilized platform and angular velocity of the orientation frame As the feedforward value of the speed loop in the three-loop control of tracking loop, speed loop and current loop, the optoelectronic pod accurately tracks the maneuvering target based on the target's motion state when switching between optical images and radar video data. (3) Calculate the optical image miss deviation angle of the photoelectric pod's line of sight at the current moment. And the off-target deviation angle of video radar data And an adaptive prediction compensation algorithm is used to calculate and Provide compensation; based on the compensation and The initial value compensation algorithm is used to calculate the state variable value of the tracking controller at the tracking switching moment. The state variable value at the tracking switching moment is used as the initial value of the state variable of the tracking loop controller, so that the error of the electro-optical pod pointing the line of sight to the ground maneuvering target is further reduced when the electro-optical pod switches between tracking optical images and radar video data.
[0006] The estimated value of the motion state of the computer moving target in the geodetic coordinate system in step (1) The specific process is as follows: (1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod, the probability density function of the maneuvering target's acceleration is represented by the modified Rayleigh density function, thereby constructing an expression for the acceleration state estimation of the maneuvering target; (1.2) Based on the acceleration state estimation expression of the maneuvering target constructed in step (1.1), construct the initial state variables of the maneuvering target. And establish the state equation and observation equation for the maneuvering target; (1.3) Based on the state equation and observation equation of the maneuvering target, the Kalman filter algorithm is used to process the initial state variables of the maneuvering target. Through iteration, the maneuvering target is finally determined in the geodetic rectangular coordinate system. k Motion state estimate at time 1 .
[0007] The specific process of constructing the acceleration state estimation expression for the maneuvering target in step (1.1) is as follows: (1.1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod a The probability density function of the acceleration of a maneuvering target is established using the modified Rayleigh density function to obtain the current acceleration. a variance ; When the current acceleration of the maneuvering target a When positive, the probability density function of the acceleration of the maneuvering target P r ( a )for
[0008] in, The known positive upper limit of acceleration for a maneuvering target. It is a constant; When the current acceleration of the maneuvering target is negative, the probability density function of the acceleration is:
[0009] in, This represents the negative lower limit of the acceleration of a known maneuvering target. When the current acceleration of the maneuvering target is 0, the probability density function of the acceleration is:
[0010] in, Represents the Diclave function; Based on the current acceleration of the maneuvering target a Parameters in the probability density function ,calculate a variance When the current acceleration of the maneuvering target a When the variance is not zero, for
[0011] When the current acceleration of the maneuvering target a When the variance is 0, =0; (1.1.2) Based on the current acceleration of the maneuvering target a variance The construction has a mean of 0 and a variance of 0. White noise Based on a non-zero mean time correlation model of maneuver acceleration, using white noise Establish the acceleration iterative equation as follows
[0012] in, Zero-mean colored acceleration noise; The value is the reciprocal of the time constant of the acceleration; solving the acceleration iterative equation yields the acceleration iterative solution as follows: ,in It is the average value of the acceleration. (1.1.3) Based on acceleration iterative solution and past acceleration observations of the maneuvering target The expression for obtaining the acceleration state estimate of a maneuvering target is as follows:
[0013] in, This is an estimate of the acceleration state of a maneuvering target. In order to be in Given the circumstances, The conditional mathematical expectation.
[0014] The specific process of establishing the state equation and observation equation of the maneuvering target in step (1.2) is as follows: (1.2.1) will k Substituting the operating time of the time-sensitive optoelectronic pod into the acceleration state estimation expression of the maneuvering target constructed in step (1.1), the following calculation is obtained: k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z Components of acceleration in the direction , and ;set up k The constant movement of the target along the lower edge of the geodetic rectangular coordinate system x , y and z Position components of direction , and All are 1; set k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z directional velocity components , and All are 1; using k Maneuvering targets in the geodetic rectangular coordinate system x Construction of acceleration, velocity, and position components in the direction x Initial maneuvering target state components in direction m ( k )=[ xx ( k ), v x ( k ), a x ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system y Construction of acceleration, velocity, and position components in the direction y Initial maneuvering target state components in direction n ( k )=[ x y ( k ), v y ( k ), a y ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system z Construction of acceleration, velocity, and position components in the direction z Initial maneuvering target state components in direction n ( k )=[ x z ( k ), v z ( k ), a z ( k )];use m ( k ), n ( k )and p ( k ) build k The initial state variables of the maneuvering target at time t are ; (1.2.2) Based on the initial state variables of the maneuvering target The state equation of the maneuvering target is constructed as follows:
[0015] in, The state transition matrix is obtained based on the maneuvering information of the maneuvering target. (1.2.3) Based on the initial state variables of the maneuvering target The observation equation for the maneuvering target is constructed, and the formula is as follows:
[0016] in,Z ( k ) is the observation vector. For the observation matrix, To observe noise.
[0017] In step (1.3), the final iteration yields the maneuvering target in the Cartesian coordinate system. k Motion state estimate at time 1 The specific process is as follows (1.3.1) Based on the initial state variables of the maneuvering target Calculate k Prior estimates of state variables at time 1 sum of error covariance estimates The calculation formula is as follows:
[0018] in, Represents the system's transition matrix;
[0019] in, express k The covariance matrix of the state variable estimates at time -1 This represents the noise driving matrix of the system. The covariance matrix representing the observation noise; (1.3.2) Utilization k Error covariance estimate at time 1 ,calculate k The system Kalman gain at time 1 The calculation formula is:
[0020] in, Represents the system's observation matrix. The covariance matrix representing the measurement noise; (1.3.3) Utilization k Prior estimates of state variables at time 1 and system Kalman gain Obtain the maneuvering target k The estimated state of motion at time t. ; The calculation formula is:
[0021] (1.3.4) Calculation k Covariance matrix of state variable estimates at time 1 The calculation formula is:
[0022] in, I Represents the identity matrix; (1.3.5) Repeat steps (1.3.1) to (1.3.4) 300 times. When repeating step (1.3.1), the estimated value of the motion state is... as initial state variables ,Will k Covariance matrix of state variable estimates at time 1 As k Covariance matrix of state variable estimates at time -1 After executing steps (1.3.1) to (1.3.4) for the 300th time, the estimated value of the motion state is obtained. This is the final estimate of the motion state of the maneuvering target in the geodetic coordinate system, thus completing the calculation of the estimated motion state of the maneuvering target in the geodetic coordinate system.
[0023] In step (2), the pitch angular velocity of the inertial stabilized platform is calculated. and angular velocity of the orientation frame The specific process is as follows: (2.1) Establish a geographic coordinate system with the photoelectric pod as the origin; (2.2) Latitude and longitude coordinates based on the pod base ,in , and The latitude, longitude, and altitude of the pod base, and the estimated motion state of the maneuvering target in the geodetic coordinate system. Estimated values of the position and state variables of the maneuvering target ,in x e , y e and z e For maneuvering targets in the geodetic coordinate system x , y and z The estimated position and state of the direction are used to calculate the coordinates of the maneuvering target in the geographic coordinate system established in step (2.1). ,in x nk , y nk and z nk For the maneuvering target in the geographic coordinate system x , y and z The coordinates of the direction. The calculation process is as follows: First calculate transpose matrix The calculation formula is:
[0024] After that, Transpose to get ; (2.3) Based on the coordinates of the maneuvering target in the geographic coordinate system Cylindrical coordinate system of computer moving target The calculation formula is:
[0025] in, The target angle for the pod's pitch frame maneuver. The azimuth frame of the pod represents the target angle for maneuvering. Relative height; (2.4) Take the partial derivative with respect to the cylindrical coordinate position of the maneuvering target to obtain the pitch angular velocity of the inertial stabilized platform. and angular velocity of the orientation frame The calculation formulas are as follows:
[0026] The specific process for calculating the miss deviation angle of the line of sight of the optoelectronic pod at the current moment in step (3) is as follows: The tracking loop of the control system acquires the number of visible light off-target pixels transmitted by the image tracker through optical payloads, radar payloads, and image processing components. And the number of pixels that missed the target by video radar ; Calculate the off-target deviation angle of the optical image in the current frame. Miss angle of radar video data The calculation formula is as follows:
[0027]
[0028] in, For the focal length of a visible light camera, Visible light pixel size, This represents the number of visible light pixels that are not directly in the target's path. For video radar focal length, For video radar pixel size, This represents the number of pixels that the video radar missed the target.
[0029] The formula for calculating the state variable value of the tracking controller at the tracking switching time in step (3) is as follows:
[0030]
[0031]
[0032] in, X As a state variable, it is calculated based on the miss angle of optical image and the miss angle of video radar data; A For the system matrix, For index equations, This is the weight matrix. It is a positive definite equation; state variables X It consists of the platform's angular velocity, angular position, and controller state variables; the dimension of the controller variables depends on the controller type; initial values. Set as the initial state variables, including current, velocity, and position information.
[0033] The beneficial effects of this invention compared to the prior art are: (1) The present invention uses the motion state information of the mobile target itself to assist the optoelectronic pod in tracking the mobile target when different load data are switched, ensuring that the ground mobile target is always within the effective field of view of the optoelectronic load, and effectively improving the tracking effect of the optoelectronic pod.
[0034] (2) Based on the probability density model of the acceleration of the maneuvering target and combined with the time correlation model of the non-zero mean of the maneuvering acceleration, the present invention estimates the motion state of the maneuvering target, which can accurately estimate the real-time state of the maneuvering target and ensure the accuracy of the acquisition of the information of the maneuvering target itself.
[0035] (3) The present invention uses data filtering processing to locate and estimate the motion of ground maneuvering targets, providing more accurate external guidance angular velocity information for the feedforward control of the tracking loop, thereby improving the continuous tracking capability of the electro-optical pod for maneuvering targets.
[0036] (4) The present invention uses an adaptive prediction compensation algorithm to process the image miss amount, which reduces the negative impact of system delay on the image miss amount. At the same time, the initial value compensation algorithm can effectively smooth the parameter discontinuity phenomenon that occurs during the switching process of the tracking controller, further ensuring the continuous and stable tracking of ground mobile targets. Attached Figure Description
[0037] Figure 1 This is a flowchart of the electro-optical pod video tracking switching method based on maneuvering target guidance information estimation according to the present invention; Figure 2 This is an overall structural diagram of the electro-optical pod video tracking switching method based on maneuvering target guidance information estimation of the present invention; Figure 3 This is a schematic diagram illustrating the switching between optical payload tracking and video radar payload tracking in this invention. Detailed Implementation
[0038] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings.
[0039] like Figure 1 , 2 As shown in Figure 3, the present invention provides a method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information, comprising the following steps: (1) After the electro-optical pod identifies the target, it calculates in real time the estimated value of the motion state of the maneuvering target tracked by the electro-optical pod in the geodetic coordinate system. ,in , and These represent the maneuvering target in the geodetic coordinate system. k The estimated values of position, velocity, and acceleration at each moment; the geodetic rectangular coordinate system is based on the center of the Earth's ellipse as the origin, with the intersection of the initial meridian and the equatorial plane as... X Axis, on the equatorial plane X The direction perpendicular to the axis is Y The axis, the earth axis is Z axis; (1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod, the probability density function of the maneuvering target's acceleration is represented by the modified Rayleigh density function, thereby constructing an expression for estimating the maneuvering target's acceleration state. The specific process is as follows: (1.1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod a The probability density function of the acceleration of a maneuvering target is established using the modified Rayleigh density function to obtain the current acceleration. a variance ; When the current acceleration of the maneuvering target a When positive, the probability density function of the acceleration of the maneuvering target P r ( a )for
[0040] in, The known positive upper limit of acceleration for a maneuvering target. It is a constant; When the current acceleration of the maneuvering target is negative, the probability density function of the acceleration is:
[0041] in, This represents the negative lower limit of the acceleration of a known maneuvering target. When the current acceleration of the maneuvering target is 0, the probability density function of the acceleration is:
[0042] in, Represents the Diclave function; Based on the current acceleration of the maneuvering target a Parameters in the probability density function ,calculate a variance When the current acceleration of the maneuvering target a When the variance is not zero, for
[0043] When the current acceleration of the maneuvering target a When the variance is 0, =0; (1.1.2) Based on the current acceleration of the maneuvering target a variance The construction has a mean of 0 and a variance of 0. White noise Based on a non-zero mean time correlation model of maneuver acceleration, using white noise Establish the acceleration iterative equation as follows
[0044] in, Zero-mean colored acceleration noise; The value is the reciprocal of the time constant of the acceleration; solving the acceleration iterative equation yields the acceleration iterative solution as follows: ,in It is the average value of the acceleration. (1.1.3) Based on acceleration iterative solution and past acceleration observations of the maneuvering target The expression for obtaining the acceleration state estimate of a maneuvering target is as follows:
[0045] in, This is an estimate of the acceleration state of a maneuvering target. In order to be in Given the circumstances, The conditional mathematical expectation; through the above process, such as Figure 3As shown, this invention estimates the motion state of a maneuvering target based on a probability density model of the target's acceleration and combined with a time-dependent model of the non-zero mean of the target's acceleration. This allows for accurate estimation of the target's real-time state and ensures the accuracy of acquiring information about the target itself.
[0046] (1.2) Based on the acceleration state estimation expression of the maneuvering target constructed in step (1.1), the state equation and observation equation of the maneuvering target are established. The specific steps are as follows: (1.2.1) will k Substituting the operating time of the time-sensitive optoelectronic pod into the acceleration state estimation expression of the maneuvering target constructed in step (1.1), the following calculation is obtained: k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z Components of acceleration in the direction , and ;set up k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z Position components of direction , and All are 1; set k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z directional velocity components , and All are 1; using k Maneuvering targets in the geodetic rectangular coordinate system x Construction of acceleration, velocity, and position components in the direction x Initial maneuvering target state components in direction m ( k )=[ x x ( k ), v x ( k ), a x ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system y Construction of acceleration, velocity, and position components in the direction y Initial maneuvering target state components in direction n ( k )=[ xy ( k ), v y ( k ), a y ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system z Construction of acceleration, velocity, and position components in the direction z Initial maneuvering target state components in direction n ( k )=[ x z ( k ), v z ( k ), a z ( k )];use m ( k ), n ( k )and p ( k ) build k The initial state variables of the maneuvering target at time t are ; (1.2.2) Based on the initial state variables of the maneuvering target The state equation of the maneuvering target is constructed as follows:
[0047] in, The state transition matrix is obtained based on the maneuvering information of the maneuvering target.
[0048] The specific steps for calculating the state transition matrix are as follows: (1.2.2.1) The computer-generated moving target along the lower edge of the geodetic rectangular coordinate system x State transition matrix in direction ,for
[0049] in T The sampling period for the photoelectric pod. It is the reciprocal of the time constant of the maneuvering acceleration; (1.2.2.2) The computer-generated moving target along the lower edge of the geodetic rectangular coordinate system y direction and z State transition matrix in direction and Calculation formula and The calculation formula is consistent; (1.2.2.3) Position the maneuvering target along the lower edge of the geodetic rectangular coordinate system. x direction, y direction and z The state transition matrix in the direction will , and By concatenating the matrices along their main diagonal, we obtain the state transition matrix. ; (1.2.3) Based on the initial state variables of the maneuvering target The observation equation for the maneuvering target is constructed, and the formula is as follows:
[0050] in, Z ( k ) is the observation vector. For the observation matrix, To observe noise.
[0051] (1.3) Based on the state equation and observation equation of the maneuvering target, the Kalman filter algorithm is used to process the initial state variables of the maneuvering target. Through iteration, the maneuvering target is finally determined in the geodetic rectangular coordinate system. k Motion state estimate at time 1 The specific process is as follows: (1.3.1) Based on the initial state variables of the maneuvering target Calculate k Prior estimates of state variables at time 1 sum of error covariance estimates The calculation formula is as follows:
[0052] in, Represents the system's transition matrix;
[0053] in, express k The covariance matrix of the state variable estimates at time -1 This represents the noise driving matrix of the system. The covariance matrix representing the observation noise; (1.3.2) Utilization k Error covariance estimate at time 1 ,calculate k The system Kalman gain at time 1 The calculation formula is:
[0054] in, Represents the system's observation matrix. The covariance matrix representing the measurement noise; (1.3.3) Utilization k Prior estimates of state variables at time 1 and system Kalman gain Obtain the maneuvering target k The estimated state of motion at time t. ; The calculation formula is: ; (1.3.4) Calculation k Covariance matrix of state variable estimates at time 1 The calculation formula is:
[0055] in, I Represents the identity matrix; (1.3.5) Repeat steps (1.3.1) to (1.3.4) 300 times. When repeating step (1.3.1), the estimated value of the motion state is... as initial state variables ,Will k Covariance matrix of state variable estimates at time 1 As k Covariance matrix of state variable estimates at time -1 After executing steps (1.3.1) to (1.3.4) for the 300th time, the estimated value of the motion state is obtained. This is the final estimate of the motion state of the maneuvering target in the geodetic coordinate system, thus completing the calculation of the estimated motion state of the maneuvering target in the geodetic coordinate system.
[0056] (2) Using the estimated value of the motion state of the maneuvering target in the geodetic coordinate system Calculate the pitch angular velocity of the inertial stabilized platform equipped with the optoelectronic pod. and angular velocity of the orientation frame ,like Figure 2 As shown, the pitch angular velocity of the inertial stabilized platform is... and angular velocity of the orientation frame As the feedforward value of the speed loop in the three-loop control of tracking loop, speed loop and current loop, the optoelectronic pod accurately tracks the maneuvering target based on the target's motion state when switching between optical images and radar video data. The pitch angular velocity of the computationally stable inertial platform and angular velocity of the orientation frame The specific process is as follows: (2.1) Establish a geographic coordinate system with the photoelectric pod as the origin; the established geographic coordinate system... X The axis points in the direction the pod is moving. Y axis perpendicular to X The axis, according to the right-hand rule, points to the right of the direction the pod is moving. Z Axis perpendicular to X shaft and Y The plane formed by the axis; (2.2) Latitude and longitude coordinates based on the pod base ,in , and The latitude, longitude, and altitude of the pod base, and the estimated motion state of the maneuvering target in the geodetic coordinate system. Estimated values of the position and state variables of the maneuvering target ,in x e , y e and z e For maneuvering targets in the geodetic coordinate system x , y and z The estimated position and state of the direction are used to calculate the coordinates of the maneuvering target in the geographic coordinate system established in step (2.1). ,in x nk , y nk and z nk For the maneuvering target in the geographic coordinate system x , y and z The coordinates of the direction. The calculation process is as follows: First calculate transpose matrix The calculation formula is:
[0057] After that, Transpose to get ; (2.3) Based on the coordinates of the maneuvering target in the geographic coordinate system Cylindrical coordinate system of computer moving target The calculation formula is:
[0058] in, The target angle for the pod's pitch frame maneuver. The azimuth frame of the pod represents the target angle for maneuvering. Relative height; (2.4) Take the partial derivative with respect to the cylindrical coordinate position of the maneuvering target to obtain the pitch angular velocity of the inertial stabilized platform. and angular velocity of the orientation frame The calculation formulas are as follows: .
[0059] (3) Calculate the optical image miss deviation angle of the photoelectric pod's line of sight at the current moment. And the off-target deviation angle of video radar data And an adaptive prediction compensation algorithm is used to calculate and Provide compensation; such as Figure 3 As shown, based on the compensated and The initial value compensation algorithm is used to calculate the state variable value of the tracking controller at the tracking switching moment. The state variable value at the tracking switching moment is used as the initial value of the state variable of the speed loop controller, so that the error of the electro-optical pod pointing the line of sight to the ground maneuvering target is further reduced when the electro-optical pod switches between optical image and radar video data. The specific steps for calculating the miss deviation angle of the line of sight of the optoelectronic pod at the current moment are as follows: The tracking loop of the control system acquires the number of visible light off-target pixels transmitted by the image tracker through optical payloads, radar payloads, and image processing components. And the number of pixels that missed the target by video radar ; Calculate the off-target deviation angle of the optical image in the current frame. Miss angle of radar video data The calculation formula is as follows:
[0060]
[0061] in, For the focal length of a visible light camera, Visible light pixel size, This represents the number of visible light pixels that are not directly in the target's path. For video radar focal length, For video radar pixel size, This represents the number of pixels that missed the target by the video radar. The formula for calculating the state variable value at the tracking switch moment of the tracking controller is as follows:
[0062]
[0063]
[0064] in, X As a state variable, it is calculated based on the miss angle of optical image and the miss angle of video radar data; A For the system matrix, For index equations, This is the weight matrix. It is a positive definite equation. State variables. X It consists of the platform's angular velocity, angular position, and controller state variables; the dimension of the controller variables depends on the controller type; initial values. The initial state variables are set, including current, velocity, and position information. Through the above process, the present invention uses an adaptive prediction compensation algorithm to process the image miss distance, which reduces the negative impact of system delay on the image miss distance. At the same time, the initial value compensation algorithm can effectively smooth the parameter discontinuity phenomenon that occurs during the switching process of the tracking controller, further ensuring the continuous and stable tracking of ground mobile targets.
[0065] In summary, this invention uses the motion state information of the mobile target itself to assist the electro-optical pod in tracking mobile targets when switching between different load data, ensuring that the ground mobile target remains within the effective field of view of the electro-optical load, and effectively improving the tracking performance of the electro-optical pod.
[0066] The parts of this invention not described in detail are common knowledge to those skilled in the art.
Claims
1. A method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information, characterized in that... The steps are as follows: (1) After the electro-optical pod identifies the target, it calculates in real time the estimated value of the motion state of the maneuvering target tracked by the electro-optical pod in the geodetic coordinate system. ,in , and These represent the maneuvering target in the geodetic coordinate system. k The estimated values of position, velocity, and acceleration at each instant; (2) Using the estimated value of the motion state of the maneuvering target in the geodetic coordinate system Calculate the pitch angular velocity of the inertial stabilized platform equipped with the optoelectronic pod. and angular velocity of the orientation frame and the pitch angular velocity of the inertial stabilized platform and angular velocity of the orientation frame As the feedforward value of the speed loop in the three-loop control of tracking loop, speed loop and current loop, the optoelectronic pod accurately tracks the maneuvering target based on the target's motion state when switching between optical images and radar video data. (3) Calculate the optical image miss deviation angle of the photoelectric pod's line of sight at the current moment. And the off-target deviation angle of video radar data And an adaptive prediction compensation algorithm is used to calculate and Provide compensation; based on the compensation and The initial value compensation algorithm is used to calculate the state variable value of the tracking controller at the tracking switching moment. The state variable value at the tracking switching moment is used as the initial value of the state variable of the tracking loop controller, so that the error of the electro-optical pod pointing the line of sight to the ground maneuvering target is further reduced when the electro-optical pod switches between tracking optical images and radar video data.
2. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 1, characterized in that: The estimated value of the motion state of the computer moving target in the geodetic coordinate system in step (1) The specific process is as follows: (1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod, the probability density function of the maneuvering target's acceleration is represented by the modified Rayleigh density function, thereby constructing an expression for the acceleration state estimation of the maneuvering target; (1.2) Based on the acceleration state estimation expression of the maneuvering target constructed in step (1.1), construct the initial state variables of the maneuvering target. And establish the state equation and observation equation for the maneuvering target; (1.3) Based on the state equation and observation equation of the maneuvering target, the Kalman filter algorithm is used to process the initial state variables of the maneuvering target. Through iteration, the maneuvering target is finally determined in the geodetic rectangular coordinate system. k Motion state estimate at time 1 .
3. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 2, characterized in that: The specific process of constructing the acceleration state estimation expression for the maneuvering target in step (1.1) is as follows: (1.1.1) Based on the current acceleration of the maneuvering target detected by the electro-optical pod a The probability density function of the acceleration of a maneuvering target is established using the modified Rayleigh density function to obtain the current acceleration. a variance ; When the current acceleration of the maneuvering target a When positive, the probability density function of the acceleration of the maneuvering target P r ( a )for in, Given the known positive upper limit of acceleration for a maneuvering target. It is a constant; When the current acceleration of the maneuvering target is negative, the probability density function of the acceleration is: in, This represents the negative lower limit of the acceleration of a known maneuvering target. When the current acceleration of the maneuvering target is 0, the probability density function of the acceleration is: in, Represents the Diclave function; Based on the current acceleration of the maneuvering target a Parameters in the probability density function ,calculate a variance When the current acceleration of the maneuvering target a When the variance is not zero, for When the current acceleration of the maneuvering target a When the variance is 0, =0; (1.1.2) Based on the current acceleration of the maneuvering target a variance The construction has a mean of 0 and a variance of 0. White noise Based on a non-zero mean time correlation model of maneuver acceleration, using white noise Establish the acceleration iterative equation as follows in, Zero-mean colored acceleration noise; The value is the reciprocal of the time constant of the acceleration; solving the acceleration iterative equation yields the acceleration iterative solution as follows: ,in It is the average value of the acceleration. (1.1.3) Based on acceleration iterative solution and past acceleration observations of the maneuvering target The expression for obtaining the acceleration state estimate of a maneuvering target is as follows: in, This is an estimate of the acceleration state of a maneuvering target. In order to be in Given the circumstances, The conditional mathematical expectation.
4. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 2, characterized in that: The specific process of establishing the state equation and observation equation of the maneuvering target in step (1.2) is as follows: (1.2.1) will k Substituting the operating time of the time-sensitive optoelectronic pod into the acceleration state estimation expression of the maneuvering target constructed in step (1.1), the following calculation is obtained: k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z Components of acceleration in the direction , and ;set up k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z Position components of direction , and All are 1; set up k The maneuvering target at any time is located at the lower edge of the geodetic rectangular coordinate system. x , y and z directional velocity components , and All are 1; use k Maneuvering targets in the geodetic rectangular coordinate system x Construction of acceleration, velocity, and position components in the direction x Initial maneuvering target state components in direction m ( k )=[ x x ( k ), v x ( k ), a x ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system y Construction of acceleration, velocity, and position components in the direction y Initial maneuvering target state components in direction n ( k )=[ x y ( k ), v y ( k ), a y ( k )];use k Maneuvering targets in the geodetic rectangular coordinate system z Construction of acceleration, velocity, and position components in the direction z Initial maneuvering target state components in direction n ( k )=[ x z ( k ), v z ( k ), a z ( k )]; use m ( k ), n ( k )and p ( k ) build k The initial state variables of the maneuvering target at time t are ; (1.2.2) Based on the initial state variables of the maneuvering target The state equation of the maneuvering target is constructed as follows: in, The state transition matrix is obtained based on the maneuvering information of the maneuvering target. (1.2.3) Based on the initial state variables of the maneuvering target The observation equation for the maneuvering target is constructed, and the formula is as follows: in, Z ( k ) is the observation vector. For the observation matrix, To observe noise.
5. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 2, characterized in that: In step (1.3), the final iteration yields the maneuvering target in the Cartesian coordinate system. k Motion state estimate at time 1 The specific process is as follows (1.3.1) Based on the initial state variables of the maneuvering target Calculate k Prior estimates of state variables at time 1 sum of error covariance estimates The calculation formula is as follows: in, Represents the system's transition matrix; in, express k The covariance matrix of the state variable estimates at time -1 Represents the noise driving matrix of the system. The covariance matrix representing the observation noise; (1.3.2) Utilization k Error covariance estimate at time 1 ,calculate k The system Kalman gain at time 1 The calculation formula is: in, Represents the system's observation matrix. The covariance matrix representing the measurement noise; (1.3.3) Utilization k Prior estimates of state variables at time 1 and system Kalman gain Obtain the mobile target k The estimated value of the motion state at time t. ; The calculation formula is: (1.3.4) Calculation k Covariance matrix of state variable estimates at time 1 The calculation formula is: in, I Represents the identity matrix; (1.3.5) Repeat steps (1.3.1) to (1.3.4) 300 times. When repeating step (1.3.1), the estimated value of the motion state is... As initial state variables ,Will k Covariance matrix of state variable estimates at time 1 As k Covariance matrix of state variable estimates at time -1 After executing steps (1.3.1) to (1.3.4) for the 300th time, the estimated value of the motion state is obtained. This is the final estimate of the motion state of the maneuvering target in the geodetic coordinate system, thus completing the calculation of the estimated motion state of the maneuvering target in the geodetic coordinate system.
6. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 1, characterized in that: In step (2), the pitch angular velocity of the inertial stabilized platform is calculated. and angular velocity of the orientation frame The specific process is as follows: (2.1) Establish a geographic coordinate system with the photoelectric pod as the origin; (2.2) Latitude and longitude coordinates based on the pod base ,in , and The latitude, longitude, and altitude of the pod base, and the estimated position and state variables of the maneuvering target in the geodetic coordinate system. ,in x e , y e and z e For maneuvering targets in the geodetic coordinate system x , y and z The estimated position and state of the direction are used to calculate the coordinates of the maneuvering target in the geographic coordinate system established in step (2.1). ,in x nk , y nk and z nk For the maneuvering target in the geographic coordinate system x , y and z The coordinates of the direction. The calculation process is as follows: First calculate transpose matrix The calculation formula is: After that, Transpose to get ; (2.3) Based on the coordinates of the maneuvering target in the geographic coordinate system Cylindrical coordinate system of computer moving target The calculation formula is: in, The target angle for the pod's pitch frame maneuver. The azimuth frame of the pod represents the target angle for maneuvering. Relative height; (2.4) Take the partial derivative with respect to the cylindrical coordinate position of the maneuvering target to obtain the pitch angular velocity of the inertial stabilized platform. and angular velocity of the orientation frame The calculation formulas are as follows: 。 7. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 1, characterized in that: The specific process for calculating the miss deviation angle of the line of sight of the optoelectronic pod at the current moment in step (3) is as follows: The tracking loop of the control system acquires the number of visible light off-target pixels transmitted by the image tracker through optical payloads, radar payloads, and image processing components. And the number of pixels that missed the target by video radar ; Calculate the off-target deviation angle of the current frame optical image. Miss angle of radar video data The calculation formula is as follows: in, For the focal length of a visible light camera, Visible light pixel size, This represents the number of visible light pixels that are not directly in the target's path. For video radar focal length, For video radar pixel size, This represents the number of pixels that the video radar missed the target.
8. The method for switching video tracking of an electro-optical pod based on estimation of maneuvering target guidance information according to claim 1, characterized in that: The formula for calculating the state variable value of the tracking controller at the tracking switching time in step (3) is as follows: in, X As a state variable, it is calculated based on the miss angle of optical image and the miss angle of video radar data; A For the system matrix, For index equations, This is the weight matrix. It is a positive definite equation; state variables X It consists of the platform's angular velocity, angular position, and controller state variables; the dimension of the controller variables depends on the controller type; initial values. Set as the initial state variables, including current, velocity, and position information.