A target tracking composite control method

Abstract

The present invention discloses a target tracking composite control method based on angular acceleration information reuse, which sequentially comprises the steps of acquiring the angular acceleration of the optical axis and the target relative to the optical axis of a device, angular acceleration feedback control, target motion state estimation, target trajectory prediction, and establishing a target tracking composite control system. The present invention acquires information such as the optical axis of a photoelectric sensor and the angular acceleration of the target relative to the optical axis, and uses the angular acceleration information to construct a feedback system on the one hand and to predict the target motion trajectory on the other hand. By reusing the angular acceleration information, based on the new target tracking composite control method, the disturbance suppression capability of the tracking control system is improved, and the target tracking performance is enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of target tracking, and in particular relates to a target tracking composite control method. Background Art

[0002] Currently, the operating environments of optoelectronic detection equipment are becoming increasingly complex, and the targets being tracked are becoming increasingly mobile. Consequently, higher requirements are being placed on the target tracking performance of optoelectronic detection equipment. Specifically, high target tracking performance is reflected in the following two aspects: 1) achieving excellent target tracking accuracy for highly mobile targets; and 2) providing excellent disturbance rejection capabilities under dynamic loading conditions of the vehicle.

[0003] Photoelectric detection equipment uses a tracking and control system to achieve target pointing and tracking. The tracking and control system usually adopts a dual closed-loop structure of angular position + angular velocity in the control structure. The angular velocity loop uses the gyroscope signal as the feedback signal to isolate the disturbance caused by the vehicle in a dynamic environment, while the angular position loop uses the target image off-target amount as the angular position error signal of the tracked target to achieve target tracking.

[0004] However, the target tracking control technologies widely used in current optoelectronic detection equipment are mostly dual-loop feedback control technologies of angular position and angular velocity: the angular velocity loop in this technology is used to isolate dynamic disturbances, but it cannot effectively suppress the disturbance effects of some strong loads and high dynamic conditions; and the angular position loop is used to track the target, but it cannot correspond to the tracking error caused by the delay in the miss distance, which is particularly fatal in the problem of tracking highly maneuverable targets.

[0005] Therefore, there is an urgent need for a target tracking control method that can effectively compensate for the tracking error caused by the miss delay and at the same time has a high suppression capability against the disturbance caused by the dynamic environment of the vehicle. Summary of the Invention

[0006] In order to take into account the two requirements of high-precision tracking of highly maneuverable targets and high-efficiency disturbance suppression in high-dynamic environments, the present invention designs a target tracking composite control method based on angular acceleration information reuse.

[0007] The technical solution adopted by the present invention to solve the technical problem is: a target tracking composite control method, comprising the following steps:

[0008] Step S1, obtaining the angular acceleration of the optical axis and the target relative to the optical axis of the device: obtaining the angular acceleration of the optical axis of the photoelectric sensor in the photoelectric detection device in the geographic system Σ N Angular acceleration under The angular position error E(t) is obtained by the target image miss distance E(t-τ) in the photoelectric detection device. According to the relationship between the angular position command R(t) and the angular position response Y(t) of the optical axis, R(t)=E(t)+Y(t), the second-order differential of the angular position command R(t) of the tracking control system is used as the angular acceleration of the target relative to the optical axis of the device. The angular position command R(t), the angular position response Y(t) of the optical axis, and the target image miss distance E(t) are all in the geographic system Σ N The following representation shows that when the device is stably tracking, the target image miss distance E(t) is much smaller than the angular position response Y(t) of the optical axis, that is, E(t)<<Y(t), so the target angular acceleration The angular acceleration of the optical axis can be approximated by Instead of getting

[0009] Step S2, angular acceleration feedback control: according to the angular acceleration of the optical axis in the geographic system An angular acceleration feedback loop is established within the feedback loop consisting of angular position and angular velocity of the tracking control system, converting the original dual-loop feedback control system into a three-loop feedback control system consisting of angular position, angular velocity, and angular acceleration. The use of an angular acceleration feedback controller further enhances the disturbance suppression capability of the entire tracking control system.

[0010] Step S3, target motion state estimation: In order to eliminate the target tracking error caused by the target image miss distance E(t-τ) delay, it is necessary to first accurately estimate the target motion state; use the optimal estimation algorithm such as Kalman filter to convert the geographic system Σ N The target angular acceleration at the time of τ The target angular position R(t-τ) and target angular velocity at the time of lag τ As the measurement signal of the Kalman filter, the angular position estimate of the target is obtained through Kalman filtering Angular velocity estimate and angular acceleration estimates Introducing target angular acceleration This high-order motion information, as an additional measurement signal for the Kalman filter, can effectively increase the bandwidth of the Kalman filter and enhance the tracking capability of the entire tracking composite control method for highly maneuverable targets.

[0011] Step S4, target trajectory prediction: using the target angular velocity estimate and acceleration The target angular position at the current moment Make a prediction

[0012] Step S5, establish a target tracking composite control system: Based on the original angular position and angular velocity dual-loop feedback, the angular acceleration information of steps S2 to S4 is reused to construct a new target tracking composite control that includes angular acceleration feedback control and target motion state estimation and trajectory prediction.

[0013] The target tracking composite control method, step S1 is to first measure the optical axis of the photoelectric sensor in the photoelectric detection device in the geographical system Σ N Angular velocity under Then, by difference, we can get the optical axis in the geographical system Σ N Angular acceleration under

[0014] The target tracking composite control method, step S1 is to move the two paired three-axis accelerometers along the optical axis pointing direction (i.e., along the geographic system Σ N The y-axis) is arranged at the M and N points of the photoelectric detection equipment, and the three sensitive axes of the two three-axis accelerometers are ensured to be parallel. At this time, the acceleration value a can be measured by the corresponding sensitive axes of the two three-axis accelerometers. M and a N Divide by the corresponding lever arm L MN To obtain the angular acceleration of the optical axis

[0015] Furthermore, the step S1 is specifically as follows: let the optical axis coordinate system and the geographic coordinate system be Σ E and Σ N , place the two triaxial accelerometers along the optical axis coordinate system Σ E The M and N points of the Y axis are arranged in front and behind, and the sensitive axes of the two three-axis accelerometers are aligned with the optical axis coordinate system Σ E The x, y, and z axes are aligned. At this time, the acceleration values sensed by the accelerometers at points M and N are [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T ;

[0016] When the photoelectric detection device rotates around the azimuth axis, the optical axis coordinate system Σ E Will be relatively geographical department N The z-axis produces θ oz When the photoelectric detection device rotates around the pitch axis, the optical axis coordinate system Σ E Will be relatively geographical department N The x-axis produces θ ox The angular acceleration value can be obtained according to the coordinate projection relationship. and Compared with the accelerometer measurement value [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T The relationship is

[0017]

[0018] Ignore the small amount of optical axis rotation The angular acceleration of the optical axis can be obtained as

[0019]

[0020] Convert the above formula into three-dimensional space and Equivalent to a unified parameter It represents the angular acceleration of the optical axis in inertial space; the angular acceleration of the target Angular acceleration with respect to the optical axis There is a relationship between In the formula is the second-order derivative of the target image miss distance. When the device is stably tracking, according to E(t)<<Y(t), the angular acceleration of the target is The angular acceleration of the optical axis can be approximated by Instead, there is

[0021] In the target tracking composite control method, the disturbance suppression function after adding the angular acceleration feedback control in step S2 is:

[0022]

[0023]

[0024] Where s is the Laplace operator, D is the disturbance, G a is the controlled model, C a 、C v and C p They are the angular acceleration loop, angular velocity loop and angular position loop feedback controllers respectively.

[0025] In the target tracking composite control method, the state equation of the Kalman filter dynamic link in step S3 is: Where s is the Laplace operator, are the estimated state variables, which are the angular position, angular velocity and angular acceleration of the target, is the state transfer matrix, K is the Kalman filter gain, H is the output matrix, and Y(s) is the measurement signal; the target angular acceleration Compared with the original target angular position R(t) and angular velocity Together they form a three-dimensional measurement signal Y. An estimated for:

[0026]

[0027]

[0028]

[0029] Where D(s) is (sI-F+KH) -1 The denominator of (sI-F+KH) -1 KH's molecular matrix.

[0030] The beneficial effects of the present invention are:

[0031] 1. The present invention uses the acquired optical axis angular acceleration information to establish an angular acceleration feedback mechanism, further improving the anti-interference capability of the tracking control system, thereby enhancing the performance of the photoelectric detection equipment under high load and high dynamic environment.

[0032] 2. By reusing angular acceleration information, the present invention expands the bandwidth of optimal estimation at the target motion state estimation level, and further effectively characterizes the target maneuverability within the miss delay time at the target trajectory prediction level, thereby improving the tracking performance of maneuverable targets.

[0033] 3. The present invention reuses motion information such as the angular acceleration of the optical axis of the photoelectric sensor in the device and the angular acceleration of the tracked target relative to the optical axis of the device. While achieving high-precision target tracking, it further improves the anti-interference capability of the control system, effectively improving the tracking performance of the photoelectric detection equipment for highly maneuverable targets in high-dynamic environments such as vehicle-mounted and airborne equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram for obtaining angular acceleration of a two-axis photoelectric detection device;

[0035] Figure 2 The following is a block diagram comparison of the tracking control system before and after adding angular acceleration feedback control;

[0036] Figure 3 This is the block diagram of the target tracking composite control system. DETAILED DESCRIPTION

[0037] The implementation process of the present invention is further described below with reference to the accompanying drawings.

[0038] In order to take into account the dual requirements of high-precision tracking of highly maneuverable targets and high-efficiency disturbance suppression in highly dynamic environments, the present invention designs a target tracking composite control method based on the reuse of angular acceleration information. In response to the target tracking requirements of photoelectric detection equipment, the angular acceleration information of the optical axis of the photoelectric sensor in the inertial space is obtained (due to the limited acceleration of the maneuvering target, this angular acceleration information can be approximately equivalent to the angular acceleration of the target relative to the optical axis in the inertial space), and this angular acceleration information is used to build a feedback system on the one hand, and to predict the target motion trajectory on the other hand. By reusing the angular acceleration information, a new target tracking composite control method is designed, which not only improves the disturbance suppression capability of the tracking control system, but also enhances the target tracking performance. Specifically, it includes the following steps.

[0039] A) Acquisition of the optical axis and the angular acceleration of the target relative to the device optical axis.

[0040] Obtain the optical axis of the photoelectric sensor in the photoelectric detection device in the geographic system Σ N Angular acceleration under The angular position error E(t) is obtained by the target image miss distance E(t-τ) (where τ is the miss distance lag time) in the photoelectric detection device. According to the relationship between the angular position command R(t) and the angular position response Y(t) of the optical axis, R(t)=E(t)+Y(t), the second-order differential of the angular position command R(t) of the tracking control system is used as the angular acceleration of the target relative to the optical axis of the device. (For the sake of convenience, it is referred to as target angular acceleration below ), where the angular position command R(t), the angular position response Y(t) of the optical axis, and the target image miss distance E(t) are all in the geographic system Σ N The following representation shows that when the device is stably tracking, the target image miss distance E(t) is much smaller than the angular position response Y(t) of the optical axis, that is, E(t)<<Y(t). Therefore, the target angular acceleration The angular acceleration of the optical axis can be approximated by Instead of getting

[0041] In this step, the gyroscope in the photoelectric detection device can be used to measure the optical axis of the photoelectric sensor in the device in the geographic system. N Angular velocity under Then, by difference, we can get the optical axis in the geographical system Σ N Angular acceleration under

[0042] This step can also be done by placing two triaxial accelerometers in front and behind along the direction of the optical axis, and ensuring that the sensitive axes of the two triaxial accelerometers are parallel to each other, and then dividing the acceleration difference measured in the corresponding direction by the corresponding lever arm to obtain the angular acceleration.

[0043] by Figure 1 The schematic diagram of angular acceleration acquisition of the two-axis photoelectric detection device is shown as an example. E and Σ N The two triaxial accelerometers are placed along the optical axis coordinate system Σ E The M and N points of the Y axis are arranged in front and behind, and the sensitive axes of the two accelerometers are aligned with the optical axis coordinate system Σ E The x, y and z axes are aligned, and the acceleration values sensed by the M and N accelerometers are [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T .

[0044] When the device rotates around the azimuth axis, the optical axis coordinate system Σ E Will be relatively geographical department N The z-axis produces θ oz Angle; when the device rotates around the pitch axis, the optical axis coordinate system Σ E Will be relatively geographical department N The x-axis produces θ ox According to the coordinate projection relationship, the angular acceleration value can be obtained and Compared with the accelerometer measurement value [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T The relationship is

[0045]

[0046] By using formula (1), we can ignore the small amount of optical axis image rotation The angular acceleration of the optical axis can be obtained as

[0047]

[0048] Formula (2) can be expressed in three-dimensional space and Equivalent to a unified parameter It represents the angular acceleration of the optical axis in the inertial space. Angular acceleration with respect to the optical axis There is a relationship between In the formula is the second-order derivative of the target image miss distance. Since when the device is tracking stably, according to E(t)<<Y(t), the angular acceleration of the target is The angular acceleration of the optical axis can be approximated by Instead, there is

[0049] B) Angular acceleration feedback control: According to the angular acceleration of the optical axis in the geographic system obtained in step A) An angular acceleration feedback loop is established within the angular position and angular velocity feedback loops of the tracking control system. The original dual-loop feedback control system consisting of angular position and angular velocity is converted into a three-loop feedback control system consisting of angular position, angular velocity, and angular acceleration. The block diagrams of the tracking control system before and after adding angular acceleration feedback control are shown below. Figure 2 shown. Figure 2 In this example, s is the Laplace operator, D is the perturbation, and G is a is the controlled model, C a 、C v and C p They are the angular acceleration loop, angular velocity loop and angular position loop feedback controllers respectively. Figure 2 It can be concluded that the disturbance rejection transfer function before and after angular acceleration feedback control is (for the purpose of simple expression, the comparison here temporarily ignores the delay characteristic of the target image miss distance e -τs )

[0050]

[0051]

[0052] Comparing equations (3) and (4), it can be seen that after adding angular acceleration feedback control, the controlled model G is eliminated in the denominator of equation (4). a The influence of |G a |<1, so at high frequencies, the disturbance rejection rate of formula (4) is significantly higher than that of formula (3), that is, the angular acceleration feedback control in step B) here improves the anti-disturbance capability of the control system under high dynamic conditions.

[0053] C) Target motion state estimation: Due to the target angular acceleration obtained in step A) Contains large sensor noise, so it is necessary to use the optimal estimation method to estimate the target angular acceleration with better signal-to-noise ratio The Kalman filter, which is widely used in engineering, is used to illustrate the optimal estimation method here. The state equation of its dynamic link is:

[0054] In formula (5), s is the Laplace operator. are the estimated state variables, which are the angular position, angular velocity and angular acceleration of the target respectively. is the state transfer matrix, K is the Kalman filter gain, H is the output matrix, and Y(s) is the measurement signal.

[0055] The method of the present invention sets the target angular acceleration Compared with the angular position R(t) and angular velocity of the target in traditional technology Together, they form a three-dimensional measurement signal Y. An estimated for:

[0056]

[0057] In formula (6), D(s) is (sI-F+KH) -1 The denominator of (sI-F+KH) -1 The molecular matrix of KH. Since the target angular acceleration is added to the measurement signal Y Therefore, the numerators of each term in equation (6) are all third-order polynomials. That is to say, the target angular acceleration is As the measurement signal of the target state estimation, the Kalman filter is used to calculate the target angular acceleration An additional zero point is added to the estimated transfer function to improve the target angular acceleration The estimation bandwidth is improved, which enhances the adaptability of the tracking control system to highly maneuverable targets.

[0058] D) Target trajectory prediction: In step C), the target angular position, angular velocity and angular acceleration estimates before the lag time τ can be obtained and The predicted value of the target angular position at the current moment can be obtained using Equation (7). In the prediction of the target angular position, the angular acceleration information, a high-order motion information, is introduced as a compensation term. This reflects the changes in angular position and angular velocity caused by the target's maneuverability within the delay time [t-τ,t], further eliminating the impact of the delay and improving the trajectory prediction accuracy.

[0059]

[0060] E) Establishing a target tracking composite control system: By combining steps B), C), and D), and reusing the angular acceleration information in the above steps, a new target tracking composite control system can be constructed. The typical control block diagram is as follows: Figure 3 shown. Figure 3 In addition to the angular acceleration feedback control described in step B), steps C) and D) are further integrated. F is a Kalman filter used to estimate the angular position of the target at the lag time in step C). Angular velocity and angular acceleration and At the same time, feedback letter Multiplying it with the angle response Y(s) completes the target trajectory prediction in step D), that is,

[0061]

[0062] from Figure 3 As can be seen, the target tracking composite control system of the present invention combines the angular acceleration feedback control described in step B) with the target motion state estimation and trajectory prediction techniques described in steps C) and D). This target tracking composite control system not only enhances the anti-interference capability of photoelectric detection equipment under high loads and high-dynamic environments, but also improves the tracking performance of maneuvering targets.

[0063] The angular acceleration information reuse involved in the present invention refers to repeatedly applying the acquired angular acceleration information to multiple links such as acceleration feedback control, target motion state estimation, and target trajectory prediction, thereby constructing a new target tracking composite control system to achieve high-precision tracking of the target while enhancing the control system's anti-interference capability.

[0064] The protection scope of the claims of the present invention is not limited to the above embodiments.

Claims

1. A target tracking composite control method, characterized by: The steps include: Step S1, obtaining the optical axis of the photoelectric sensor in the photoelectric detection device in the geographic system Σ N Angular acceleration under The angular position error E(t) is obtained by the target image miss distance E(t-τ) in the photoelectric detection device. According to the relationship between the angular position command R(t) and the angular position response Y(t) of the optical axis, R(t)=E(t)+Y(t), the second-order differential of the system's angular position command R(t) is used as the angular acceleration of the target relative to the optical axis of the device. The angular position command R(t), the angular position response Y(t) of the optical axis, and the target image miss distance E(t) are all in the geographic system Σ N The target angular acceleration is represented as follows: Approximate the angular acceleration of the optical axis Instead of getting Step S2, according to the angular acceleration An angular acceleration feedback loop is established based on the dual-loop feedback control consisting of the system's angular position and angular velocity, converting it into a three-loop feedback control consisting of angular position, angular velocity, and angular acceleration. The angular acceleration feedback controller is used to enhance the entire system's ability to suppress disturbances. Step S3, the geographical system N The target angular acceleration at the time of τ The target angular position R(t-τ) and target angular velocity at the time of lag τ As the measurement signal of the Kalman filter, the angular position estimate of the target is obtained through Kalman filtering Angular velocity estimate and angular acceleration estimates Step S4, using the target angular velocity estimate and target angular acceleration The target angular position at the current moment Make a prediction Step S5, reusing the angular acceleration information in steps S2 to S4 to implement feedback control of the angular acceleration and estimation of the motion state and trajectory prediction of the target.

2. A target tracking composite control method according to claim 1, characterized in that: The step S1 is to first measure the optical axis of the photoelectric sensor in the geographical system Σ N Angular velocity under Then, by difference, we can get the optical axis in the geographical system Σ N Angular acceleration under 3. The target tracking composite control method according to claim 1, characterized in that: The step S1 is to arrange two pairs of three-axis accelerometers at points M and N of the photoelectric detection device in the direction of the optical axis, and ensure that the three sensitive axes of the two three-axis accelerometers are parallel. The acceleration values a measured by the corresponding sensitive axes of the two three-axis accelerometers are M and a N Divide by the corresponding lever arm L MN To obtain the angular acceleration of the optical axis 4. A target tracking composite control method according to claim 3, characterized in that: The step S1 is specifically as follows: Let the optical axis coordinate system and geographic coordinate system be Σ E and Σ N , place the two triaxial accelerometers along the optical axis coordinate system Σ E The M and N points of the Y axis are arranged in front and behind, and the sensitive axes of the two three-axis accelerometers are aligned with the optical axis coordinate system Σ E The x, y, and z axes are aligned, and the acceleration values sensed by the accelerometers at points M and N are [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T ; After the photoelectric detection device rotates around the azimuth axis, the optical axis coordinate system Σ E Relative Geography Department Σ N The z-axis produces θ oz The angle of rotation of the photoelectric detection device is Σ after it rotates around the pitch axis. E Relative Geography Department Σ N The x-axis produces θ ox The angle of rotation, get the angular acceleration value and Compared with the accelerometer measurement value [a Mx a My a Mz ] T and [a Nx a Ny a Nz ] T relationship Ignore the small amount of optical axis rotation Get the angular acceleration of the optical axis The three-dimensional space and Equivalent to a unified parameter That is, it represents the angular acceleration of the optical axis in inertial space; Angular acceleration of the target Angular acceleration with respect to the optical axis Relationship In the formula is the second-order derivative of the target image miss distance. According to E(t)<<Y(t), the angular acceleration of the target Approximate the angular acceleration of the optical axis Instead, that is 5. A target tracking composite control method according to claim 1, 2, 3 or 4, characterized in that: The disturbance suppression function after adding angular acceleration feedback control in step S2 is: Where s is the Laplace operator, D is the disturbance, G a is the controlled model, C a 、C v and C p They are the angular acceleration loop, angular velocity loop and angular position loop feedback controllers respectively.

6. A target tracking composite control method according to claim 1, characterized in that: The Kalman filter state equation in step S3 is Where s is the Laplace operator, are the estimated state variables, which are the angular position, angular velocity and angular acceleration of the target, is the state transfer matrix, K is the Kalman filter gain, H is the output matrix, and Y(s) is the measurement signal; the target angular acceleration With angular position R(t) and angular velocity Together they form a three-dimensional measurement signal Y, and we get Estimated for: Where D(s) is (sI-F+KH) -1 The denominator of (sI-F+KH) -1 KH's molecular matrix.

Images