A method for realizing high-precision memory tracking of an optoelectronic search and tracking system

Abstract

The application discloses a kind of to realize optoelectronic search tracking system high-precision memory tracking method, it includes the following steps: step 1: initialization optoelectronic search tracking system, capture and track target, system shows after tracking well, emit laser ranging, when laser ranging data is normal, enter step 2;Step 2: optoelectronic search tracking system enters regenerative feedback tracking, extracts servo tracking angular velocity, angular acceleration information under automatic tracking state;Step 3: target is shielded, and infrared or television detector error amount is invalid, and optoelectronic search tracking system uses step 2 regenerative feedback extrapolation target information and enters memory tracking;Step 4: memory tracking time length reaches or sensor reacquires target and sends error amount, end memory tracking.The application realizes high-precision memory tracking, and greatly improves the adaptability and fighting capacity of optoelectronic search tracking system.

Description

Technical Field

[0001] This invention belongs to the field of automatic control technology and relates to a method for achieving high-precision memory tracking in a photoelectric search and tracking system. Background Technology

[0002] An optoelectronic search and tracking system is an opto-mechatronic device that uses photoelectric detectors (such as television, infrared, and laser detectors) to process signals and drive the photoelectric detection unit via a servo system to point the optical axis of the device towards the target being detected. Most of them have memory tracking function. Memory tracking refers to the process in which the photoelectric search and tracking system's servo system continues to move along the target trajectory based on memory when the image detector temporarily loses sight due to obstructions such as smoke during the tracking process. High-precision memory tracking can ensure that after the temporary obstruction disappears, the target is still basically in the center of the field of view. After the detector re-acquires the target, the optoelectronic search and tracking system achieves fast and accurate tracking.

[0003] Consulting relevant materials, the "Memory Tracking Algorithm Based on Kalman Filter" is mentioned in antenna usage. This algorithm extracts servo tracking angle data and establishes an antenna tracking motion model through Kalman filtering, which can effectively improve the memory tracking accuracy of telemetry servo equipment under abnormal flight conditions. This method is suitable for situations where the target distance is long. For photoelectric search and tracking systems, the target distance is short and the speed is high, requiring even higher memory tracking accuracy. Furthermore, while memory tracking through an image processing system is theoretically feasible, it significantly increases the computational load of the image processing system in practical engineering. Secondly, the target tracking process is a real-time dynamic process, and the servo system performs the actions. The image processing system's understanding of this process differs from that of the servo system, further increasing the implementation complexity. Additionally, the computation cycle of the servo system is much longer than that of the image processing system; from a real-time perspective, this method is not optimal.

[0004] Currently, there are no reports in publicly available articles, materials, or patents, both domestically and internationally, regarding the high-precision memory tracking method for the photoelectric search and tracking system described in this invention. Summary of the Invention

[0005] (I) Purpose of the Invention

[0006] The purpose of this invention is to propose a method for high-precision memory tracking in photoelectric search and tracking systems when the target is lost.

[0007] (II) Technical Solution

[0008] In this invention, when the target is partially or completely obscured, the video tracker cannot obtain target information and report tracking deviation. At this time, the motion trajectory of the servo system and laser ranging data are collected, and the α-β-γ filtering algorithm and composite control technology are used to achieve extrapolation to complete closed-loop feedback. The pointer of the photoelectric search and tracking system is controlled to rotate in the direction of target movement, so as to ensure that the target is near the center of the field of view of the photoelectric sensor when it reappears.

[0009] The application of memory tracking technology in this invention enables the dynamic tracking accuracy of the system to meet the technical requirements.

[0010] The method provided by this invention is mainly implemented by the servo control software of the optoelectronic device. Before the servo control software runs, the basic loop of the control system is available, and the required data is stable and reliable, including:

[0011] (1) The speed and position sensor data of the photoelectric equipment are stable and usable.

[0012] (2) The laser data of the optoelectronic equipment is stable and usable.

[0013] (3) Azimuth, pitch velocity loop and position loop are available.

[0014] With the above work completed, when the servo control software receives control information from the upper-level system, it will execute the following steps:

[0015] Step 1: Initialize all data and ensure the servo system is powered on. Acquire and track the target. Once the device displays successful tracking, emit a laser. If the laser data is normal, proceed to Step 2. Laser ranging should be functioning correctly in this method; fault conditions are not described here.

[0016] Step 2: After the laser ranging stabilizes, the tracker enters regenerative feedback tracking and extracts information such as servo tracking angular velocity and angular acceleration in automatic tracking mode;

[0017] Step 3: Target occlusion renders the infrared or television detector error invalid. The servo system uses the regenerative feedback from step two to extrapolate the target information and enter memory tracking.

[0018] Step 4: End memory tracking when the memory tracking time expires or the sensor re-acquires the target and sends the error amount.

[0019] (III) Beneficial Effects

[0020] The method for achieving high-precision memory tracking in an electro-optical search and tracking system provided by the above technical solution utilizes the basic characteristics of a servo system. By collecting the motion trajectory of the servo system and laser ranging data, and employing an α-β-γ filtering algorithm and composite control technology, it achieves high-precision memory tracking with an extrapolation time of up to 3.5 seconds. This method fully utilizes the characteristics of the rate / position sensor and laser ranging of the electro-optical search and tracking system. Furthermore, the α-β-γ filtering algorithm has better real-time performance than Kalman filtering, while the application of the composite control algorithm ensures that the memory tracking accuracy error is no greater than 0.1 milliradians and the tracking time is 3.5 seconds. Achieving high-precision memory tracking will greatly improve the adaptability and combat capability of the electro-optical search and tracking system. Attached Figure Description

[0021] Figure 1 Flowchart of the processing method of this invention.

[0022] Figure 2 Block diagram of the composite control system in this invention.

[0023] Figure 3 A schematic diagram of a closed-loop control system with regenerative feedback in this invention.

[0024] Figure 4 A schematic diagram illustrating the application of regenerative feedback technology in this invention.

[0025] Figure 5 Graph showing measured data of a photoelectric search and tracking system in this invention. Detailed Implementation

[0026] To make the objectives, contents, and advantages of the present invention clearer, the specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples.

[0027] A preferred embodiment of the method for achieving high-precision memory tracking in a photoelectric search and tracking system according to the present invention is implemented by photoelectric device servo control software. The servo control software runs on a servo control board with a DSP as the main control chip. The prerequisites for implementing this method are as follows:

[0028] (1) The speed and position sensor data of the photoelectric equipment are stable and usable.

[0029] (2) The laser data of the optoelectronic equipment is stable and usable.

[0030] (3) Azimuth, pitch velocity loop and position loop are available.

[0031] With the above work completed, when the servo control software receives control information from the upper-level system, it will proceed according to... Figure 1 The workflow shown follows these steps:

[0032] Step 1: Initialize all data and ensure the servo system is powered on. Acquire and track the target. Once the device displays successful tracking, emit a laser rangefinder. If the laser rangefinder data is normal, proceed to step 2. In this method, the laser rangefinder should be functioning correctly; fault conditions are not described here.

[0033] When electro-optical search and track systems are applied in battlefield environments, incoming missiles or aircraft approach from a distance. The electro-optical tracking equipment receives target indication from the search radar, points to the target, and then acquires the target solely based on positional error. After successful acquisition, it enters the tracking process. Currently, to ensure high-precision tracking of high-speed targets, the servo system of electro-optical search and track systems uses composite control technology, as shown in the block diagram below. Figure 2 R(s) represents the target's trajectory. Currently, in automatic target tracking, R(s) is the target's motion parameters in the inertial coordinate system obtained by the servo system using parameters measured by the photoelectric search and tracking system (azimuth, pitch, laser distance). These parameters are then transformed to a geodetic rectangular coordinate system via inertial attitude angle transformation, and α-β-γ filtering is used to obtain the target's motion parameters in the inertial frame. Finally, these parameters are transformed back to the photoelectric search and tracking system platform coordinate system through a series of coordinate transformations and introduced into the servo system's velocity loop input Gv(S). If the measured target motion parameters are accurate, the aiming line of the photoelectric search and tracking system can also rotate accurately according to the target's motion, effectively reducing the burden on the servo system's position error channel and significantly decreasing the servo system's dynamic hysteresis error.

[0034] When the target undergoes constant acceleration, the target state vector X is a three-dimensional vector X = These represent the position, velocity, and acceleration, respectively.

[0035] The state equation is:

[0036] in, w(k) is the state noise, which is a series of Gaussian white noise with zero mean; T is the sampling interval, which is 0.001s in this system.

[0037] The measurement equation is: Y(k)=H(X(k)+v(k)), where H=(1 0 0), v(k) is Gaussian measurement noise, and Y(k) is the current measurement value.

[0038] The α-β-γ filtering equation is:

[0039]

[0040]

[0041] K = (α β / T 2γ / T2);

[0042] in, X(k / k) represents the predicted value of the current target state vector based on the target state vector of the previous moment, and X(k / k) represents the extrapolated value of the current moment. This represents the predicted value for the next moment. α, β, and γ are set according to the tracking accuracy.

[0043] Step 2: After the laser ranging stabilizes, the tracker enters regenerative feedback tracking and extracts information such as servo tracking angular velocity and angular acceleration in automatic tracking mode;

[0044] Regenerative feedback control technology is an approximate feedforward control technique developed based on composite control technology. A schematic diagram of a closed-loop control system with regenerative feedback is shown below. Figure 3 As shown.

[0045] In automatic target tracking mode, regenerative feedback utilizes the output of the photoelectric closed-loop tracking control system as the current value of the control variable, indirectly measuring the target's motion parameters. A computer is used to statistically analyze, calculate, extrapolate, smooth, and filter these changes to obtain the target's position, velocity, and acceleration in the Cartesian coordinate system. Finally, after a series of coordinate transformations, the velocity and acceleration are fed into the input of W2 as control signals, completing approximate feedforward velocity compensation control of the original closed-loop tracking control system. This constitutes a composite control system with a first-order position loop and regenerative feedback, significantly reducing the tracking error of the tracking system.

[0046] In automatic target tracking mode, target motion parameters can only be measured indirectly, and complex mathematical smoothing and prediction estimation techniques, as well as various forms of coordinate transformation, are required to achieve this. Figure 4 This diagram illustrates the principle of a microcomputer-assisted tracking (regenerative feedback) system for an optoelectronic servo system.

[0047] Figure 4 The servo system hardware unit includes an azimuth servo system, an elevation servo system, and a laser rangefinder system (not fully shown in the diagram), with outputs β0, ε0, and R0, respectively, used as input signals for the regenerative feedback branch. The open-loop channels of the two angle servo systems mainly consist of two units, W1(S) and W2(S). W1(S) is the spatial position loop correction circuit; W2(S) is the system velocity stabilization loop; and R0 is the output of the laser rangefinder. (In this example, the vehicle body and turret coordinate systems are the same, and are uniformly represented by the vehicle body.)

[0048] The azimuth output β of the servo system during automatic target tracking o Elevation angle output ε o The distance output R of the laser rangefinder is converted into X0, Y0, and Z0 in the vehicle's Cartesian coordinate system, and then processed by matrix [C]. TConvert M0, N0, and H0 to the geodetic rectangular coordinate system (i.e., M at the input of the α-β-γ digital recursive filter) n N n H m Smoothing and differentiation employ a three-state constant-coefficient α-β-γ digital recursive filter, which outputs three sets of signals, one of which is the target position estimate. (i.e., in the filtered estimation equation system) The other group is the target speed estimate; (i.e., in the filtered estimation equation system) The third group is the target acceleration estimation. (i.e., in the filtered estimation equation system) The above three sets of estimates are transformed into position estimates in the vehicle's Cartesian coordinate system using matrix [C]. Speed ​​valuation Acceleration valuation The distance to the valuation can then be calculated. Shortcut valuation Related to the estimation of azimuth output Related to pitch output Equal values ​​are obtained. Then, they are transformed by matrix [D] into velocity estimates in the bracket's line-of-sight Cartesian coordinate system. and acceleration valuation according to The azimuth angular velocity estimate can be obtained. And pitch rate estimation And azimuth acceleration estimation And pitch acceleration estimation

[0049] Estimating angular acceleration Multiply by the acceleration compensation coefficient G respectively β G ε Then, the corresponding angular velocity estimate Adding them together gives the required auxiliary tracking composite angular velocity. By adding speed control to the W2(S) input terminals of the azimuth and elevation branches of the servo system respectively, it can be ensured that the aiming line of the photoelectric search and tracking system rotates accurately in space according to the target's motion law. That is, the aiming line accurately targets the target throughout the entire tracking process, thereby effectively reducing the dynamic lag error of the servo system during automatic target tracking.

[0050] Once the servo system has successfully tracked the data, it extracts information such as servo tracking angular velocity and angular acceleration, and updates this information in real time.

[0051] Step 3: Target occlusion renders the infrared or television detector error invalid, and the servo system uses the regenerative feedback from the second step to extrapolate the target information and enter memory tracking;

[0052] The key to high-precision memory tracking lies in obtaining sufficient target information to predict the target's position. Composite control technology ensures reliable target motion information because the target's motion has inertia, and its velocity and acceleration do not change abruptly in a short period. Composite control technology guarantees the reliability of the measurement data before memory tracking. After filtering, the position, angular velocity, angular acceleration, and other information are extrapolated using an extrapolation algorithm to obtain high-precision memory tracking prediction data, thus achieving high-precision memory tracking.

[0053] Step 4: End memory tracking when the memory tracking time expires or the sensor re-acquires the target and sends the error amount.

[0054] Figure 5 This data represents the measured tracking data stored in the memory of a certain photoelectric tracking device. The horizontal axis represents time, with a recording interval of 20 milliseconds, and the vertical axis represents the azimuth and pitch errors, in milliradians. The object being measured is a target drone. After successful tracking, there is a smoke obscuring the target for approximately 3.5 seconds after the drone leaves the barrel. The photoelectric search and tracking system then enters memory tracking mode and resumes normal tracking after approximately 3.6 seconds.

[0055] In the figure, the curves represent the pitch tracking error and azimuth tracking error, respectively. The straight lines represent memory tracking, where the error is cleared to zero in the software. The figure shows that when the electro-optical search and tracking system enters memory tracking after being obscured by smoke (lasting approximately 3.6 seconds), and then switches back to normal tracking, the maximum pitch tracking error is 0.08 milliradians, and the maximum azimuth tracking error is 0.1 milliradians.

[0056] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for achieving high-precision memory tracking in a photoelectric search and tracking system, characterized in that, Includes the following steps: Step 1: Initialize the photoelectric search and tracking system, acquire and track the target. After the system displays that tracking is successful, it will emit a laser for ranging. If the laser ranging data is normal, proceed to Step 2. Step 2: The photoelectric search and tracking system enters regenerative feedback tracking, and extracts servo tracking angular velocity and angular acceleration information in automatic tracking mode; Step 3: Target occlusion, infrared or television detector error is invalid, photoelectric search and tracking system uses the regenerated feedback from step 2 to extrapolate target information and enter memory tracking; Step 4: End memory tracking when the memory tracking time expires or the sensor re-acquires the target and sends the error value; In step 1, the photoelectric search and tracking system adopts servo composite control technology. During the automatic target tracking process, R(s) is the motion parameters of the target in the inertial system obtained by the servo system through the inertial attitude angle transformation of the photoelectric search and tracking system (azimuth angle, pitch angle, laser distance) and the inertial attitude angle transformation. Then, it is transformed back to the coordinate system of the photoelectric search and tracking system platform through a series of coordinate transformations and introduced into the input terminal of the servo system velocity loop Gv(S). In step 1, when the target is undergoing constant acceleration, the target state vector X is a three-dimensional vector X = ( ), representing position, velocity, and acceleration, respectively; The state equation is: in, , , For state noise, is a series of zero-mean Gaussian white noise; T is the sampling interval. The measurement equation is: ,in , For Gaussian measurement of noise, This is the current measurement value; The α-β-γ filtering equation is: ； in, This represents the predicted value of the current target state vector based on the target state vector at the previous moment. This indicates that the current value is the extrapolated value at the current moment; This represents the predicted value for the next moment, with α, β, and γ set according to the tracking accuracy. In step 2, regenerative feedback tracking uses the output of the photoelectric closed-loop tracking control system as the current value of the control quantity. It uses an indirect method to measure the target motion parameters, that is, it uses a computer to statistically analyze, calculate, extrapolate, smooth and filter the changes to obtain the target's position, velocity and acceleration in the geodetic rectangular coordinate system. Finally, after a series of coordinate transformations, the velocity and acceleration are added to the input of W2 as control signals to complete the approximate feedforward velocity compensation control of the original closed-loop tracking control system. This constitutes a composite control system with a first-order position loop and regenerative feedback, reducing the tracking error of the tracking system. In step 2, under automatic target tracking, the target motion parameters are measured indirectly, and mathematical smoothing, prediction estimation, and coordinate transformation are used to achieve this. In step 2, the photoelectric search and tracking system includes an azimuth servo system, an elevation servo system, and a laser ranging system, whose outputs are respectively... , , The input signal is used as the regenerative feedback branch; the open-loop channels of the azimuth servo system and the elevation servo system are both composed of... , It consists of two major units, among which Spatial position loop correction stage, It is the system speed stability loop; It is the output of the laser rangefinder; The azimuth output of the photoelectric search and tracking system during automatic target tracking Elevation output And the distance output R of the laser rangefinder, converted into a Cartesian coordinate system for the vehicle body. , , Then through the matrix Converted to a geodetic rectangular coordinate system , , That is, the input of the αβγ digital recursive filter. , , Smoothing and differentiation employ a three-state constant coefficient αβγ digital recursive filter, which outputs three sets of signals, one of which is the target position estimate. , , That is, in the filtered estimation equation system , , The other group is the target speed estimate. , , That is, in the filtered estimation equation system , , The third group is the target acceleration estimation. , , That is, in the filtered estimation equation system , , The above three sets of valuations were processed by matrix... Position estimate in the vehicle body Cartesian coordinate system , , Speed ​​valuation , , Accelerated valuation , , Calculate the distance estimate Shortcut valuation Related to the estimation of azimuth output , Related to pitch output , Valuation; then processed separately through matrix Velocity estimate transformed into a Cartesian coordinate system based on the bracket's line of sight , , and acceleration valuation , , ;according to , , , , , Calculate the estimated azimuth angular velocity. And pitch rate estimation And azimuth acceleration estimation And pitch acceleration estimation .

2. The method for achieving high-precision memory tracking in a photoelectric search and tracking system as described in claim 1, characterized in that, In step 2, the angular acceleration is estimated. , Multiply by the acceleration compensation coefficient respectively , Then, the corresponding angular velocity estimate , Add them together to obtain the required auxiliary tracking composite angular velocity. = + , = + The values ​​are respectively added to the azimuth branch and elevation branch speed loop of the servo system. Speed ​​control at the input ensures that the aiming line of the photoelectric search and tracking system rotates accurately in space according to the target's motion law. In other words, the aiming line accurately targets the target throughout the entire tracking process, reducing the dynamic lag error of the servo system during automatic target tracking.

3. The method for achieving high-precision memory tracking in a photoelectric search and tracking system as described in claim 2, characterized in that, In step 3, the filtered position, angular velocity, and angular acceleration information are extrapolated using an extrapolation algorithm to obtain memory tracking prediction data, thereby achieving memory tracking.

Images