An unmanned equipment countermeasure integrated laser attack system

Abstract

The present application relates to the technical field of unmanned equipment countermeasure, and particularly relates to an unmanned equipment countermeasure integrated laser attack system. The system obtains the wavefront slope of each grid of a detection surface, obtains a slope mutation mask value based on the difference of adjacent wavefront slopes, obtains a total deformation value based on historical driving voltage and deformation mapping, determines a deformation upper limit occupancy, converts the mechanical angular velocity of a servo base into the relative linear velocity of a target surface, combines the slope mutation mask value to predict the time value of abnormal distortion spread, obtains a spatial anomaly suppression weight based on the slope mutation mask value, the deformation upper limit occupancy and the time value, uses the weight as an observation weight to solve a wavefront phase compensation value based on the wavefront slope, converts the driving voltage to complete the control of the bottom deformation mirror. The present application effectively avoids the surface shape tearing and high frequency scattering caused by hard limiting by introducing the spatiotemporal joint soft weight suppression, and maintains the stable convergence of high energy laser under the premise of ensuring the safety of the bottom hardware.

Description

Technical Field

[0001] This invention relates to the field of unmanned equipment countermeasures technology, specifically to an integrated laser strike system for unmanned equipment countermeasures. Background Technology

[0002] In the field of countering unmanned equipment, integrated laser strike systems rely on adaptive optics systems to compensate for wavefront distortion caused by atmospheric turbulence in real time, thereby maintaining the energy convergence of the high-energy laser beam on the target surface. When the system deals with low-altitude, highly maneuverable UAVs, the operating conditions are characterized by strong turbulence interference and high-speed deflection tracking coupling. Strong turbulence induces localized, severe phase abrupt changes within the laser beam cross-section; simultaneously, the underlying servo base of the system must perform high-speed deflection tracking, which forces these phase abrupt changes to undergo rapid position migration (field-of-view slip) on the optical detection grid. Under such extreme conditions, extreme compensation displacement commands are continuously output to track these abrupt changes and high-speed slip distortion data. However, the mechanical extension and retraction stroke of the piezoelectric actuator at the bottom of the deformable mirror has an inherent physical upper limit; extreme displacement commands will exhaust the actuator's stroke margin in a very short time, leading to stroke saturation of a large area of ​​the actuator.

[0003] To prevent damage from forced pulling of hardware, existing technologies typically set a fixed drive voltage limiting threshold at the output end of the digital-to-analog converter. However, this hard-end truncation operation can cause a huge displacement step between adjacent actuators, tearing apart the originally continuous and smooth reflective surface, introducing severe high-frequency scattering distortion, and causing the complete dissipation of laser focusing energy, severely weakening the destructive effectiveness of the laser strike system. Summary of the Invention

[0004] To address the technical problems of existing technologies that use end-effector truncation to prevent actuator stroke saturation, leading to reflective surface tearing, increased high-frequency scattering, and laser energy dissipation, the present invention aims to provide an integrated laser strike system for countering unmanned equipment. The specific technical solution adopted is as follows: This invention provides an integrated laser strike system for countering unmanned equipment, the system comprising: The slope change mask value acquisition module is used to acquire the wavefront slope of each grid divided by the probe surface under the current control cycle, and to acquire the slope change mask value of each grid based on the difference in wavefront slope between each grid and its preset neighboring grid. The deformation limit occupancy rate acquisition module is used to obtain the total deformation value of each grid based on the historical driving voltage of each actuator and the deformation mapping of each actuator in each grid, and to determine the deformation limit occupancy rate of each grid by combining the deformation limit of the actuator. The consumption era value acquisition module is used to convert the current mechanical angular velocity of the servo base into the relative translational linear velocity on the sensor target surface. Using this as a boundary and combined with the slope change mask value, it predicts the consumption era value of abnormal distortion spreading to each grid. The spatial anomaly suppression weight acquisition module is used to acquire the spatial anomaly suppression weight of each grid based on the slope mutation mask value, deformation upper limit occupancy rate, and consumption time value of each grid. The deformable mirror control module is used to take the spatial anomaly suppression weight of each grid as the observation weight, combine it with the wavefront slope, jointly calculate the wavefront phase compensation value, and convert it into driving voltage to complete the closed-loop control of the bottom deformable mirror.

[0005] Furthermore, the method for obtaining the slope abrupt change mask value is as follows: For any given grid, starting from that grid, traverse its preset neighboring grids in sequence along a preset direction to form a closed path; The sum of the absolute values ​​of the differences in wavefront slopes between adjacent grids on a closed path is used as the slope change analysis value for that grid. When the slope change analysis value is greater than the preset change threshold, the slope change mask value of the grid is set to 0; When the slope change analysis value is less than or equal to the preset change threshold, the slope change mask value of the grid is set to 1.

[0006] Furthermore, the method for obtaining the slope change analysis value is as follows: The preset neighborhood grid consists of three adjacent grids that form a 2×2 sliding window with the grid; the wavefront slope includes the longitudinal wavefront slope and the transverse wavefront slope. Following the order of the closed path, obtain the difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the vertical dimension of each adjacent grid on the closed path. The difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the longitudinal dimension are accumulated along the closed path direction, and the absolute value of the accumulated result is taken as the slope change analysis value of the grid.

[0007] Furthermore, the method for obtaining the total deformation value is as follows: For any given grid, obtain the deformation influence mapping coefficient of each actuator in that grid, and use the product of the historical driving voltage of each actuator and its corresponding deformation influence mapping coefficient as the contribution reference amount of each actuator. The sum of the reference values ​​contributed by all actuators is taken as the total deformation value of the mesh.

[0008] Furthermore, the method for obtaining the deformation upper limit occupancy rate is as follows: For any grid, divide the absolute value of the total deformation of the grid by the upper limit of deformation, compare it with the constant 1 and take the minimum value. Use the minimum value as the upper limit occupancy rate of the deformation of the grid.

[0009] Furthermore, the method for obtaining the relative translational linear velocity is as follows: The mechanical pitch and azimuth angular velocities of the servo base within the current control cycle are obtained. Combined with the system's preset optical scaling constant, the mechanical pitch angular velocity is converted into a longitudinal translational velocity component on the sensor target surface, and the azimuth angular velocity is converted into a lateral translational velocity component on the sensor target surface. The relative translational linear velocity is formed by the lateral translational velocity component and the longitudinal translational velocity component.

[0010] Furthermore, the method for obtaining the value of the era is as follows: Mesh points with a slope mutation mask value of 0 are marked as anomalous source points, and their time-consuming value is initialized to 0. Mesh points with a slope mutation mask value of 1 are initialized to the preset maximum reference value. Based on the relative translation linear velocity, the step cost for the anomaly source point to spread to its preset adjacent grids is determined; if the magnitude of the relative translation linear velocity is less than the preset zero-speed anti-jitter dead zone threshold, the same preset default balanced step cost is uniformly adopted in all expansion directions, or the remaining spread expansion calculation for the anomaly source point is directly terminated. If the magnitude of the relative translational velocity is greater than or equal to the preset zero-speed anti-shake dead zone threshold, then when the dot product of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational velocity is greater than or equal to 0, a first preset step cost is set; when the dot product of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational velocity is less than 0, a second preset step cost is set; wherein, the first preset step cost is less than the second preset step cost; Starting from the anomaly source, the time consumption value of each grid is updated layer by layer outward according to the step cost. When the time consumption value on a certain spreading branch reaches the preset maximum prediction time threshold, the branch expansion is stopped, and the time consumption value of the remaining untraversed grids is uniformly assigned to the preset maximum prediction time threshold to obtain the time consumption value of each grid.

[0011] Furthermore, the method for obtaining the spatial anomaly suppression weights is as follows: For any grid, the maximum value between the decay function value of the grid's time consumption period and the deformation upper limit occupancy rate is taken as the spatiotemporal joint risk factor of the grid. The result of negatively correlated mapping of spatiotemporal joint risk factors is used as the safety margin factor of the grid. The product of the slope abrupt change mask value of the grid and the safety margin factor is used as the spatial anomaly suppression weight of the grid.

[0012] Furthermore, the method for obtaining the wavefront phase compensation value is as follows: The two-dimensional wavefront slopes of each grid are concatenated into a one-dimensional global slope observation vector. The spatial anomaly suppression weights of each grid are filled into the corresponding main diagonal elements in the matrix where the horizontal wavefront slope arrangement sequence index position and the vertical wavefront slope arrangement sequence index position of the grid are located. The remaining positions are filled with zeros. In this way, a diagonal weighted matrix that matches the dimension of the long sequence of global slope observation vector is generated. The diagonal weighted matrix is ​​used as the observation weight matrix to calculate the corresponding covariance matrix. If the covariance matrix is ​​rank deficient and exhibits singular matrix characteristics, a preset micro-regularization scalar constant is injected into the full permutation of the main diagonal of the covariance matrix, and then the matrix inversion is performed to output a one-dimensional wavefront phase compensation value. If the covariance matrix is ​​not rank deficient, matrix inversion is performed directly to output a one-dimensional wavefront phase compensation value.

[0013] Furthermore, the method for closed-loop control of the bottom deformable mirror is as follows: Multiply the wavefront phase compensation value by the preset voltage conversion gain constant of the deformable mirror to calculate the single-drive voltage increment signal; Obtain the historical drive voltage values ​​of each underlying actuator retained from the previous control cycle, and accumulate them with the single drive voltage increment signal according to the actuator correspondence to obtain the currently issued drive voltage control signal. The drive voltage control signal is converted into an analog high voltage signal and applied to each actuator at the bottom layer.

[0014] The present invention has the following beneficial effects: This invention first obtains the slope abrupt change mask value of each grid based on the difference in wavefront slope between each grid and its preset neighboring grids. This accurately reflects the data corruption state and distortion step characteristics of each grid, which is beneficial for accurately stripping anomalous turbulent data that disrupts spatial continuity. To provide a precise physical deformation baseline, the invention further obtains the total deformation value of each grid based on the historical driving voltage of each actuator and the deformation mapping of each actuator in each grid. This accurately reflects the actual total physical deformation height of the underlying reflector surface corresponding to each grid. To quantify the remaining compensation capability of the underlying hardware without damage, the invention further combines the upper limit of actuator deformation to determine the upper limit occupancy rate of each grid, facilitating an accurate assessment of the degree of danger of the actuators below each grid approaching their mechanical and physical limits. To quantify the unequal cost of distortion intrusion during high-speed deflection and aiming, the current mechanical angular velocity of the servo base is converted into the relative translational linear velocity on the sensor target surface. This ensures the accurate construction of a pseudo-wind field model characterizing the field-of-view slip trend and provides topological boundary conditions for anomalous propagation. This is a boundary combined with slope abrupt change mask value, predicting the time-consuming value of anomalous distortion spreading to each grid, accurately reflecting the urgency of each grid being covered and contaminated by the surrounding anomalous region; in order to comprehensively consider the dual threats of hardware exhaustion and external distortion approach, the spatial anomaly suppression weight of each grid is obtained based on the slope abrupt change mask value, deformation upper limit occupancy rate and time-consuming value of each grid, accurately reflecting the dynamic algebraic confidence of each grid in the dual critical state; in order to maintain the smooth continuity of the global compensation surface shape while weakening high-risk input, the spatial anomaly suppression weight of each grid is used as the observation weight, combined with the wavefront slope, to jointly calculate the wavefront phase compensation value, actively avoiding the interference of extreme errors at the source of wavefront reconstruction calculation, in order to achieve the final safe optical correction response, and convert it into driving voltage to complete the closed-loop control of the bottom deformable mirror, completely eliminating the surface tearing and high-frequency scattering phenomenon caused by traditional end hard limiting, thereby ensuring the safety of the bottom actuator hardware and continuously maintaining the stable convergence of the high-energy laser beam on the target surface. Attached Figure Description

[0015] Figure 1 This is a structural block diagram of an integrated laser strike system for countering unmanned equipment, provided as an embodiment of the present invention. Detailed Implementation

[0016] The following description, in conjunction with the accompanying drawings, details a specific solution provided by the present invention for an integrated laser strike system for countering unmanned equipment.

[0017] Example 1: This invention proposes an integrated laser strike system for countering unmanned equipment. Please refer to [link / reference]. Figure 1The diagram illustrates a structural block diagram of an integrated laser strike system for countering unmanned equipment, provided by an embodiment of the present invention. The system includes: a slope change mask value acquisition module 10, a deformation upper limit occupancy rate acquisition module 20, a consumption era value acquisition module 30, a spatial anomaly suppression weight acquisition module 40, and a deformable mirror control module 50.

[0018] The slope change mask value acquisition module 10 is used to acquire the wavefront slope of each grid divided by the probe surface under the current control cycle, and to acquire the slope change mask value of each grid based on the difference in wavefront slope between each grid and its preset neighboring grid.

[0019] Specifically, considering the strong transient and time-varying characteristics of the integrated laser strike system when dealing with the superposition of strong turbulence and high-speed deflection and aiming, in order to ensure the timeliness of distortion detection, this embodiment first determines the current control cycle through the system master clock triggering mechanism, and then divides the detection surface under the current control cycle into various grids according to the sub-aperture physical array of the wavefront sensor. It should be noted that the divided grids are two-dimensional grid arrays arranged according to row and column rules. In addition, for areas at the boundary where the effective light-gathering area is insufficient to form a complete grid due to the obstruction of the light-gathering aperture, this embodiment marks them as invalid grids and directly discards them to prevent centroid detection noise caused by weak local light intensity from interfering with system control. In order to accurately characterize the spatial morphology of the current wavefront distortion, the wavefront slope of each grid is obtained in real time through the wavefront sensor. The wavefront slope includes the longitudinal wavefront slope and the transverse wavefront slope, which respectively characterize the transient gradient information of the wavefront phase in the orthogonal dimension within each grid.

[0020] Considering that normal atmospheric distortion wavefronts exhibit a continuous and smooth spatial transition, while strong turbulence can cause localized step-like fractures in the wavefront, this study aims to accurately identify anomalous fracture features that could overload the actuator. This ensures the stable operation of the wavefront reconstruction algorithm and the control safety of the underlying deformable mirror. Furthermore, based on the wavefront slope difference between each grid and its preset neighboring grids, slope abrupt change mask values ​​are obtained for each grid. This accurately reflects the data corruption state of each grid, facilitating accurate analysis of the underlying topological boundaries of anomalous distortion propagation and providing reliable data support for subsequent analysis. The preset neighboring grid consists of three adjacent grids forming a 2×2 sliding window with each grid. For example, taking the current grid in a two-dimensional grid coordinate system, its preset neighboring grids are its right neighbor, its lower right neighbor, and its lower neighbor. It should be noted that if the current grid is located at the boundary of the detection surface, resulting in a lack of true preset neighboring grids, the wavefront slope values ​​of the missing preset neighboring grids are directly copied as the wavefront slope values ​​of the current grid. This ensures dimensional alignment of the subsequent overall mask matrix and avoids incorrect anomalous fault detection.

[0021] Preferably, in one feasible embodiment, the method for obtaining the slope change mask value is as follows: For any grid, starting from the grid, traverse its preset neighboring grids sequentially along a preset direction (i.e., clockwise) to form a closed path (i.e., the grid - the grid's right adjacent grid - the grid's lower right adjacent grid - the grid's lower adjacent grid - the grid); according to the order of the closed path, obtain the difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the vertical dimension of each adjacent grid on the closed path; for example, calculate the difference between the transverse wavefront slope of the adjacent grids at the upper edge of the closed path, the difference between the longitudinal wavefront slope of the adjacent grids at the right edge, the difference between the transverse wavefront reverse slope of the adjacent grids at the lower edge, and the difference between the longitudinal wavefront reverse slope of the adjacent grids at the left edge. Considering that the slope loop integral of a continuous light field along a small closed path should approach zero, since the normal wavefront phase is a single-valued and continuously distributed scalar field, the closed path integral of its gradient must be equal to the phase difference between the starting and ending points, i.e., zero. To extract the curl abrupt change that disrupts spatial continuity, the difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the longitudinal dimension are accumulated along the closed path direction, and the absolute value of the accumulated result is taken as the slope change analysis value of the grid, accurately reflecting the severity of the step fracture in the local region enclosed by the grid; for example, taking the grid in the i-th row and j-th column... For example, the formula for calculating the corresponding slope change analysis value is: In the formula, For grid The slope change analysis value; For grid The transverse wavefront slope; For grid The transverse wavefront slope; For grid The longitudinal wavefront slope; For grid The longitudinal wavefront slope; For grid The transverse wavefront slope; For grid The transverse wavefront slope; For grid The longitudinal wavefront slope; For grid The longitudinal wavefront slope; The symbol represents the absolute value. It should be noted that, to prevent large-scale misjudgments caused by the inherent defocusing and converging quadratic surfaces of the wavefront, the wavefront slope calculated using the closed path in this embodiment is essentially the residual data after subtracting the base. That is, before extracting the difference, the least squares method is first used to perform two-dimensional linear fitting on the current global transverse and longitudinal wavefront slope arrays respectively, extracting the global linear trend surface representing normal convergence or defocusing of the system; then, the original wavefront slope of each grid is subtracted from its corresponding partial derivative value on this linear trend surface to obtain the residual data, i.e., the processed wavefront slope, ensuring that a zero-response is only generated for high-frequency fractures caused by strong turbulence; the least squares method is a well-known technique and will not be elaborated further. Considering the inherent hot electron random disturbances within the wavefront sensor, in order to avoid misjudging the device's own background noise as anomalies caused by strong turbulence during dynamic control, this embodiment extracts the baseline slope change analysis value of each grid on the detection surface under turbulence-free conditions. The product of the maximum baseline slope change analysis value and a preset tolerance coefficient (set to 1.5 in this embodiment to ensure that it can effectively contain the inherent thermal noise fluctuations of the sensor without missing the real turbulent step fracture; the implementer can set it according to the signal-to-noise ratio performance of the specific wavefront sensor used, which is not limited here) is used as a preset change threshold. Essentially, this is a fixed benchmark for determining the security of dynamic grid data. When the slope change analysis value is greater than the preset change threshold, it indicates that the grid has a wavefront break caused by strong atmospheric scintillation. In order to prevent this abnormal data from entering the subsequent wavefront reconstruction equation, the slope change mask value of the grid is set to 0, indicating that the detection data at this location is damaged. When the slope change analysis value is less than or equal to the preset change threshold, it indicates that the grid has not produced obvious phase break. In order to retain the ability to follow and compensate for normal distortion observation data, the slope change mask value of the grid is set to 1, indicating that the detection data at this location is normal and valid.

[0022] The deformation limit occupancy rate acquisition module 20 is used to obtain the total deformation value of each grid based on the historical driving voltage of each actuator and the deformation mapping of each actuator in each grid, and to determine the deformation limit occupancy rate of each grid by combining the deformation limit of the actuator.

[0023] Specifically, considering that the calculation instructions for the current control cycle have not yet been generated, forcibly waiting for the calculation results of this cycle to evaluate the mirror state would cause the control loop to get stuck in deadlock and infinite waiting. In order to provide an accurate physical deformation baseline, this embodiment obtains the historical driving voltage of each actuator by sending a status readback instruction to the system's underlying digital-to-analog converter. That is, the voltage value latched and actually applied to the underlying actuator at the end of the immediately preceding control cycle. If the current control cycle is the initial cycle, the historical driving voltage of each actuator is directly initialized to 0 to characterize the deformable mirror as being in an initial static ground state without deformation. In order to establish a spatial dimension conversion bridge from one-dimensional control voltage to two-dimensional optical grid deformation, the deformation influence mapping coefficient of each actuator in each grid is obtained by reading the factory calibration data stored locally in advance. Its physical meaning is that when a unit voltage is applied to a single actuator, the actuator produces a displacement deformation contribution of a corresponding height at a specific two-dimensional grid. In order to quantify the current mechanical overload risk of the system, and then obtain the total deformation value of each grid based on the historical driving voltage of each actuator and the deformation mapping of each actuator in each grid, the total physical deformation height of the underlying reflective surface corresponding to each grid is accurately reflected.

[0024] The specific method for obtaining the total deformation value is as follows: For any grid, the product of the historical driving voltage of each actuator and its corresponding deformation influence mapping coefficient is used as the contribution reference of each actuator to characterize the real absolute displacement generated independently by a single actuator at that grid; In order to calculate the global deformation state under the combined action of all driving voltages, the sum of the contribution references of all actuators is used as the total deformation value of the grid based on the linear superposition principle of thin plate bending.

[0025] Considering the inherent physical tensile and compressive limits of the mechanical extension and retraction stroke of the piezoelectric actuator at the bottom layer of the deformable mirror, in order to accurately assess the remaining tensile or compressive capacity of the bottom hardware without damage, the upper limit of the actuator's deformation is obtained by retrieving the inherent parameters in the deformable mirror hardware manual, i.e., the maximum allowable single-unit extension and retraction length extreme value; then, for any grid, the absolute value of the total deformation value of the grid (only retaining the deformation amplitude) is divided by the upper limit of deformation, and compared with the constant 1 to take the minimum value, and the minimum value is taken as the upper limit of deformation occupancy rate of the grid. It should be noted that, considering the cross-coupling between adjacent actuators, the absolute value of the total deformation obtained by linear superposition may be greater than the upper limit of the deformation of a single unit. In order to prevent the occupancy rate calculation result from being greater than 1, which would lead to a negative value in the subsequent abnormal suppression weight calculated based on the occupancy rate and cause oscillation of the closed-loop system, it is necessary to perform a forced truncation operation by comparing it with a constant 1. The occupancy rate of the upper limit of deformation after truncation and normalization is strictly between 0 and 1. The closer the occupancy rate of the upper limit of deformation is to 1, the closer the actuator below the grid is to the limit of mechanical and physical damage, and the smaller the remaining compensation margin.

[0026] The consumption era value acquisition module 30 is used to convert the current mechanical angular velocity of the servo base into the relative translational linear velocity on the sensor target surface. Using this as a boundary and combined with the slope change mask value, it predicts the consumption era value of abnormal distortion spreading to each grid.

[0027] Specifically, considering that during the high-speed deflection and aiming process of highly maneuverable targets in an integrated laser strike system, the originally static or slowly evolving atmospheric turbulence will undergo rapid field-of-view slip on the two-dimensional target surface of the wavefront sensor, in order to establish a heuristic engineering pseudo-wind field model boundary to quantify the unequal cost of distortion intrusion, the current mechanical angular velocity of the servo base is converted into the relative translational linear velocity on the sensor target surface to accurately reflect the distortion slip trend induced by servo tracking motion. The relative translational linear velocity is not the physical absolute velocity of the atmospheric crosswind, but a boundary parameter used to define the ease of distortion propagation. Thus, by combining the relative translational linear velocity as the boundary with the slope abrupt change mask value that characterizes the current distorted grid state, the urgency of normal grids being contaminated by surrounding abnormal grids is analyzed, and finally, the epoch value of abnormal distortion spreading to each grid is predicted, so as to achieve early warning intervention for grids that are about to experience hardware overload.

[0028] Preferably, in one feasible embodiment of this method, the relative translational linear velocity is obtained as follows: First, encoder feedback data is read through the communication bus to obtain the mechanical pitch and azimuth angular velocities of the servo base in the current control cycle. Combined with the system's preset optical scaling constant, the mechanical pitch angular velocity is linearly converted into a longitudinal translational velocity component on the sensor target surface to characterize the field-of-view sliding rate of distortion in the Y-axis direction; the azimuth angular velocity is linearly converted into a lateral translational velocity component on the sensor target surface to characterize the field-of-view sliding rate of distortion in the X-axis direction. To construct a comprehensive velocity vector in a two-dimensional coordinate system, the lateral and longitudinal translational velocity components are used to form the relative translational linear velocity, i.e., the two-dimensional velocity vector characterizing the field-of-view translation feature in the current control cycle. In this embodiment, the system's preset optical scaling constant is set as a calibration conversion coefficient for the focal length ratio, which strictly corresponds to the telescope magnification of the main optical system. The implementer can set the system's preset optical scaling constant according to the specific system optical path design parameters, which is not limited here.

[0029] Preferably, in one feasible embodiment, the method for obtaining the time consumption value is as follows: A grid with a slope abrupt change mask value of 0 indicates that the probe data at that location has been broken and damaged. To establish the initial expansion boundary of the fast traversal algorithm, the grid with a slope abrupt change mask value of 0 is marked as an anomaly source point, and its time consumption value is initialized to 0, representing that the distortion already exists and can be reached without waiting. Simultaneously, the time consumption value of the grid with a slope abrupt change mask value of 1 (i.e., the grid to be tested in a safe state) is initialized to a preset maximum reference value, serving as the default base for which the actual time consumption has not yet been calculated. In this embodiment, the preset maximum reference value is set to the maximum floating-point constant allowed by the system (e.g., 99999.9) to ensure that the grid to be tested is distinguished from the known anomaly source points during the initialization phase. The implementer can set the preset maximum reference value according to the code compilation environment and data type, which is not limited here. The fast traversal algorithm is well-known and will not be described in detail here. Considering the strong anisotropy of distortion propagation under the influence of relative translational pseudo-wind fields, in order to enable the fast travel algorithm to reflect the slip characteristics under real working conditions, the step cost of the anomaly source spreading to its preset adjacent grids is determined based on the relative translational linear velocity, thereby measuring the algebraic resistance of the anomaly source invading the adjacent region. Considering that in hovering or extremely slow slip states, even slight velocity calculation perturbations can easily cause severe reversals and jitters in the pseudo-wind field direction determination, it is necessary to determine whether the magnitude of the relative translational linear velocity is less than the preset zero-speed anti-jitter dead zone threshold. In this embodiment, the preset zero-speed anti-jitter dead zone threshold is set to 0.05 to ensure effective filtering of mechanically static aiming states. Regarding the encoder quantization noise, the implementer can set a preset zero-speed anti-shake dead zone threshold based on the static stabilization accuracy of the servo base; no limitation is imposed here. If the relative translational linear velocity is less than the preset zero-speed anti-shake dead zone threshold, it indicates that the current state is relatively static, and the distortion propagation shows no significant directional slippage trend and exhibits isotropic behavior. In this case, the same preset default equilibrium step cost (e.g., 1.0, to characterize the uniform resistance of the natural background diffusion of distortion without pseudo-wind field directional intervention; no limitation is imposed here) is uniformly applied in all propagation directions. Alternatively, the remaining propagation calculation for the abnormal source point can be directly terminated to completely escape the calculation dead zone and avoid local extreme value binding that would cause the safety closed loop to lose its adjustment basis. If the relative translational linear velocity is greater than or equal to the preset zero-speed anti-shake dead zone threshold, it indicates that the current field of view directional slip characteristics are significant and the direction vector confidence is high. In order to accurately construct the anisotropic propagation resistance boundary, when the dot product of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational linear velocity is greater than or equal to 0, it indicates that the angle between the corresponding propagation direction and the field of view slip direction caused by the servo is in the downwind range (angle less than or equal to 90 degrees), and the anomaly is easy to spread rapidly. In order to accelerate the propagation calculation in this direction at the algebraic level, a first preset step cost is set at this time, representing extremely low propagation resistance; when the magnitude of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational linear velocity is greater than or equal to 0, it indicates that the angle between the corresponding propagation direction and the field of view slip direction caused by the servo is in the downwind range (angle less than or equal to 90 degrees), and the anomaly is easy to spread rapidly. When the dot product of the line-moving velocities is less than 0, it indicates that the corresponding spread direction is in the upwind range, making it extremely difficult for anomalies to penetrate in that direction. In order to suppress the propagation calculation in this direction at the algebraic level, a second preset step cost is set at this time. The first preset step cost is less than the second preset step cost, so as to accurately construct an anisotropic diffusion field with fast downwind and slow upwind. In this embodiment, the first preset step cost is set to 0.01 and the second preset step cost is set to 100.0 to ensure that the numerical distribution of the spread time presents a significant gradient that conforms to physical common sense. Implementers can set the first preset step cost and the second preset step cost according to the dynamic smoothness of the specific tracking task, which is not limited here. To meet the timing requirements of the integrated laser strike system's millisecond-level hard real-time closed loop and prevent computational timeouts due to full target depth traversal, the system updates the time-consuming value of each grid layer outwards from the anomaly source point according to the step cost. When the time-consuming value of a certain spreading branch reaches the preset maximum prediction time threshold, it indicates that the branch is far enough away from the anomaly source point or is in a difficult-to-reach area due to headwinds, and there is no risk of being covered by distortion in the short term. At this time, the branch expansion is stopped, and the time-consuming value of the remaining untraversed grids is uniformly assigned to the preset maximum prediction time threshold, thus obtaining the time-consuming value of each grid. This truncation mechanism strictly limits the global search to the local high-risk neighborhood, significantly improving computational efficiency. In this embodiment, the preset maximum prediction time threshold is set to 5 control cycles to ensure that sufficient system pressure response margin is reserved 5 control cycles in advance from a controllable perspective. Implementers can set the preset maximum prediction time threshold according to the upper limit of the computing power of the underlying control chip and the servo response latency characteristics, which is not limited here.

[0030] The spatial anomaly suppression weight acquisition module 40 is used to acquire the spatial anomaly suppression weight of each grid based on the slope mutation mask value, deformation upper limit occupancy rate and consumption time value of each grid.

[0031] Specifically, considering that single parameter threshold determination is prone to extreme logic failure, and that independent consideration of hardware travel state or external propagation trend cannot fully characterize the critical state of the system under the coupling of strong turbulence and high-speed tracking, in order to accurately determine the degree of soft limiting suppression of each grid, and then obtain the spatial anomaly suppression weight of each grid based on the slope change mask value, deformation upper limit occupancy rate and consumption value of each grid, it is beneficial to weaken high-risk distortion in advance at the input of the reconstruction equation and block the generation of abnormal instructions that lead to actuator travel saturation.

[0032] Preferably, in one feasible embodiment of this method, the spatial anomaly suppression weight is obtained as follows: Considering that both internal actuator exhaustion and external anomaly distortion approach are independent high-risk factors that could cause grid overload, in order to ensure that the two early warning systems can independently limit output performance, for any grid, the maximum value between the attenuation function value of the grid's time-consuming period and the deformation upper limit occupancy rate is used as the spatiotemporal joint risk factor of the grid. This ensures that any single crisis indicator can transmit a limiting attenuation signal downstream, accurately reflecting the grid's critical extreme value under either hardware limits or environmental deterioration. The attenuation function value of the time-consuming period is obtained through... To acquire, where T represents the value of the era. Here, exp is the preset decay adjustment constant; exp is an exponential function with the natural constant as the base; in this embodiment, the specific value of the preset decay adjustment constant is set based on the convergence speed calibration of historical simulation data, and is used to convert the topological time elapsed value, which has no physical second attribute, into a dimensionless decay exponent. This decay function value, according to the negatively biased decay characteristic of the exponent, decreases as the propagation time decreases. A sharp increase tends to 1, and conversely, a decrease tends to 0; In order to transform the high-risk state into a suppression and reduction action on the numerical weight, the result of negatively mapping the spatiotemporal joint risk factor is obtained by subtracting the spatiotemporal joint risk factor from the constant 1, which is used as the safety margin factor of the grid. This makes the safety margin factor drop sharply to 0 under the dual critical state, while maintaining a safety range close to 1 when any single condition is safe. Considering that the most stringent veto mechanism should be implemented for grid data that has already been broken or damaged during the detection phase to prevent broken data from polluting the global restoration matrix, in order to achieve physical isolation of the absolute forbidden zone at the source, the product of the slope change mask value of the grid and the safety margin factor is used as the spatial anomaly suppression weight of the grid. Since the mask value of the confirmed abnormal grid is 0, this multiplication operation directly forces its weight to be cleared to zero, so that the system completely abandons the tracking of the damaged data.

[0033] The deformable mirror control module 50 is used to take the spatial anomaly suppression weight of each grid as the observation weight, combine it with the wavefront slope, jointly calculate the wavefront phase compensation value, and convert it into driving voltage to complete the closed-loop control of the bottom deformable mirror.

[0034] Specifically, considering that directly rigidly limiting or violently zeroing the wavefront slope of the probe would force the restoration algorithm to create drastic flat displacement steps in local areas, thereby tearing apart the original continuous reflective surface and inducing severe high-frequency scattering, in order to weaken the high-risk input while maintaining the smooth continuity of the global compensation surface, the spatial anomaly suppression weight of each grid is used as the observation weight. Combined with the wavefront slope, the wavefront phase compensation value is jointly calculated, actively reducing the contribution of the dangerous value at the algebraic input end. In order to achieve the final optical closed-loop response, the calculated wavefront phase compensation value is converted into a driving voltage to complete the closed-loop control of the underlying deformable mirror. This achieves the suppression of divergent oscillations in the control loop and the maintenance of stable convergence of the main energy without modifying the underlying hardware physical limits.

[0035] Preferably, in one feasible embodiment of this method, the wavefront phase compensation value is obtained as follows: by concatenating the two-dimensional wavefront slopes of each grid in a fixed index order from left to right and from top to bottom, a one-dimensional global slope observation vector is formed. This integrates the optical partial derivative information of all sub-apertures on the detection surface in the orthogonal two dimensions, providing a complete data column vector for global wavefront reconstruction. In order to accurately convert the single scalar soft constraint on the two-dimensional spatial grid into an algebraic penalty operator under the one-dimensional biphasor structure, and to resolve the contradiction between the weight base and the tensor scale mismatch between the observation sequence, the spatial anomaly suppression weights of each grid are arranged along the transverse wavefront slope associated with the grid. The matrix containing the sequence index position and the longitudinal wavefront slope arrangement sequence index position is filled into the corresponding main diagonal elements. That is, the same spatial anomaly suppression weights calculated by a single grid are cloned into two equal copies and placed in the row and column diagonal positions corresponding to the horizontal wavefront slope and the vertical wavefront slope of the grid, respectively. The remaining positions are filled with zeros. This expands and generates a diagonal weighted matrix that matches the dimension of the long sequence of global slope observation vectors. This ensures that in the subsequent weighted solution calculation, the two observation components of the same high-risk grid in the horizontal and vertical directions can be subject to strict homogeneous synchronous weight reduction suppression, avoiding local distortion of the solver caused by the omission of single-axis penalty. To reasonably avoid interference from high-risk observation nodes when searching for the globally optimal wavefront reconstruction solution and prevent a few extreme errors from skewing the overall reconstruction surface shape, the diagonal weighted matrix is ​​used as the observation weight matrix, combined with the system's geometric derivative bias matrix to calculate the corresponding covariance matrix. Considering that setting a large number of high-risk grid weights to zero will cause the covariance matrix to be less than rank, it is determined whether the current covariance matrix is ​​rank deficient and exhibits singular matrix characteristics: if the covariance matrix is ​​rank deficient and exhibits singular matrix characteristics, a preset micro-regularization scalar constant is injected into all permutations of the main diagonal of the covariance matrix. To prevent algebraic deadlock, matrix inversion is performed to output a one-dimensional wavefront phase compensation value. In this embodiment, a preset micro-regularization scalar constant is set to 0.001 to ensure that while breaking the singularity of the covariance matrix, excessive smoothing deviation disturbances are not generated in the normal weighted observation data. Implementers can set the preset micro-regularization scalar constant according to the floating-point operation precision of the underlying main control chip and the tolerance lower limit of the matrix inversion operator, which is not limited here. If the covariance matrix does not have rank deficiency, matrix inversion is performed directly to output a one-dimensional wavefront phase compensation value. Due to the intervention of the diagonal weighting matrix at the covariance level, the error contribution of extreme anomaly coordinates is softly suppressed rather than forcibly zeroed, ensuring that the solved wavefront phase compensation value remains numerically smooth between adjacent spatial nodes, thus avoiding step displacement differences in the underlying reflector surface. The specific formula for calculating the wavefront phase compensation value is as follows: In the formula, P is the output wavefront phase compensation value, and H is the system geometric derivative bias matrix. Let C be the transpose of the system geometric derivative bias matrix, C be the diagonal weighted matrix, and G be the global slope observation vector. It should be noted that the system geometric derivative bias matrix is ​​obtained by calibrating and fixing the spatial topological partial derivative relationship between the discrete slope grid of the wavefront sensor and the continuously reconstructed nodes of the deformable mirror at the factory. It is a fixed physical parameter describing the static mapping relationship of the optical path and can be directly used during the solution process.

[0036] Preferably, in one feasible embodiment, the closed-loop control method for the bottom deformable mirror is as follows: Considering that the wavefront phase compensation value is represented as optical path difference data in the digital system, which lacks electrical properties that directly drive hardware motion, in order to convert it into an actuator-recognizable electronic control quantity, the wavefront phase compensation value is multiplied by the preset voltage conversion gain constant of the deformable mirror to calculate the single-drive voltage increment signal; considering that the wavefront phase compensation value only represents the incremental adjustment requirement for the transient residual error of the system, if the existing data is directly issued to replace it, the original tracking and locking deformation of the deformable mirror will be lost and the actuator will rebound to the zero position and miss the target. Therefore, the historical drive voltage values ​​of each bottom actuator retained in the previous control cycle are obtained and accumulated with the single-drive voltage increment signal according to the actuator correspondence, and the calculated and merged results are obtained to obtain the currently issued drive voltage control signal, which accurately reflects the maximum voltage required to maintain the absolute tracking position. Final voltage amplitude; In this embodiment, the preset voltage conversion gain constant is set to the reciprocal of the factory piezoelectric coefficient of the bottom piezoelectric ceramic material (i.e., the optical path difference generated per unit voltage) to ensure that the digital control signal is strictly matched with the actual required deformation displacement. The implementer can set the preset voltage conversion gain constant according to the hardware electrical specifications of the specific deformable mirror purchased, which is not limited here; In order to ultimately drive the hardware to generate physical deformation, the digital drive voltage control signal is sent to the digital-to-analog converter of the system through the communication bus, converted into an analog high voltage signal, and applied to each actuator of the bottom layer. The actuators are driven to generate corresponding physical expansion and contraction according to the received smooth and safe voltage command, and complete the safe optical correction closed loop of the current control cycle. At this time, the drive voltage received by the bottom actuator is always within its physical stroke safety range, the reflective surface shape remains continuous and smooth, and the main energy of the high-energy laser beam can be continuously and stably focused on the target.

[0037] In summary, this embodiment obtains the wavefront slope of each grid on the detection surface, and obtains the slope abrupt change mask value based on the difference in wavefront slope of neighboring areas; it obtains the total deformation value based on historical driving voltage and deformation mapping, and determines the upper limit occupancy rate of deformation; it converts the mechanical angular velocity of the servo base into the relative translational linear velocity of the target surface, and combines it with the slope abrupt change mask value to predict the epoch value of anomalous distortion propagation; it obtains spatial anomaly suppression weights based on the slope abrupt change mask value, the upper limit occupancy rate of deformation, and the epoch value of epoch value, and uses them as observation weights to jointly calculate the wavefront phase compensation value with the wavefront slope, which is then converted into a driving voltage to complete the control of the underlying deformable mirror. This invention effectively avoids surface tearing and high-frequency scattering caused by hard limiting by introducing spatiotemporal joint soft weight suppression, and maintains the stable focusing of high-energy laser while ensuring the safety of the underlying hardware.

Claims

1. An integrated laser strike system for countering unmanned equipment, characterized in that, The system includes: The slope change mask value acquisition module is used to acquire the wavefront slope of each grid divided by the probe surface under the current control cycle, and to acquire the slope change mask value of each grid based on the difference in wavefront slope between each grid and its preset neighboring grid. The deformation limit occupancy rate acquisition module is used to obtain the total deformation value of each grid based on the historical driving voltage of each actuator and the deformation mapping of each actuator in each grid, and to determine the deformation limit occupancy rate of each grid by combining the deformation limit of the actuator. The consumption era value acquisition module is used to convert the current mechanical angular velocity of the servo base into the relative translational linear velocity on the sensor target surface. Using this as a boundary and combined with the slope change mask value, it predicts the consumption era value of abnormal distortion spreading to each grid. The spatial anomaly suppression weight acquisition module is used to acquire the spatial anomaly suppression weight of each grid based on the slope mutation mask value, deformation upper limit occupancy rate, and consumption time value of each grid. The deformable mirror control module is used to take the spatial anomaly suppression weight of each grid as the observation weight, combine it with the wavefront slope, jointly calculate the wavefront phase compensation value, and convert it into driving voltage to complete the closed-loop control of the bottom deformable mirror. The method for obtaining the slope abrupt change mask value is as follows: For any given grid, starting from that grid, traverse its preset neighboring grids in sequence along a preset direction to form a closed path; The sum of the absolute values ​​of the differences in wavefront slopes between adjacent grids on a closed path is used as the slope change analysis value for that grid. When the slope change analysis value is greater than the preset change threshold, the slope change mask value of the grid is set to 0; When the slope change analysis value is less than or equal to the preset change threshold, the slope change mask value of the grid is set to 1; The method for obtaining the value of the aforementioned consumption era is as follows: Mesh points with a slope mutation mask value of 0 are marked as anomalous source points, and their time-consuming value is initialized to 0. Mesh points with a slope mutation mask value of 1 are initialized to the preset maximum reference value. Based on the relative translation linear velocity, the step cost for the anomaly source point to spread to its preset adjacent grids is determined; if the magnitude of the relative translation linear velocity is less than the preset zero-speed anti-jitter dead zone threshold, the same preset default balanced step cost is uniformly adopted in all expansion directions, or the remaining spread expansion calculation for the anomaly source point is directly terminated. If the magnitude of the relative translational velocity is greater than or equal to the preset zero-speed anti-shake dead zone threshold, then when the dot product of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational velocity is greater than or equal to 0, a first preset step cost is set; when the dot product of the direction vector from the anomaly source point to its preset adjacent grid and the relative translational velocity is less than 0, a second preset step cost is set; wherein, the first preset step cost is less than the second preset step cost; Starting from the anomaly source, the time consumption value of each grid is updated layer by layer outward according to the step cost. When the time consumption value on a certain spreading branch reaches the preset maximum prediction time threshold, the branch expansion is stopped, and the time consumption value of the remaining untraversed grids is uniformly assigned to the preset maximum prediction time threshold to obtain the time consumption value of each grid. The method for obtaining the spatial anomaly suppression weights is as follows: For any grid, the maximum value between the decay function value of the grid's time consumption period and the deformation upper limit occupancy rate is taken as the spatiotemporal joint risk factor of the grid. The result of negatively correlated mapping of spatiotemporal joint risk factors is used as the safety margin factor of the grid. The product of the slope abrupt change mask value of the grid and the safety margin factor is used as the spatial anomaly suppression weight of the grid.

2. The integrated laser strike system for countering unmanned equipment as described in claim 1, characterized in that, The method for obtaining the slope change analysis value is as follows: The preset neighborhood grid consists of three adjacent grids that form a 2×2 sliding window with the grid; the wavefront slope includes the longitudinal wavefront slope and the transverse wavefront slope. Following the order of the closed path, obtain the difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the vertical dimension of each adjacent grid on the closed path. The difference between the transverse wavefront slope in the transverse dimension and the difference between the longitudinal wavefront slope in the longitudinal dimension are accumulated along the closed path direction, and the absolute value of the accumulated result is taken as the slope change analysis value of the grid.

3. The integrated laser strike system for countering unmanned equipment as described in claim 1, characterized in that, The method for obtaining the total deformation value is as follows: For any given grid, obtain the deformation influence mapping coefficient of each actuator in that grid, and use the product of the historical driving voltage of each actuator and its corresponding deformation influence mapping coefficient as the contribution reference amount of each actuator. The sum of the reference values ​​contributed by all actuators is taken as the total deformation value of the mesh.

4. The integrated laser strike system for countering unmanned equipment as described in claim 1, characterized in that, The method for obtaining the upper limit of deformation occupancy rate is as follows: For any grid, divide the absolute value of the total deformation of the grid by the upper limit of deformation, compare it with the constant 1 and take the minimum value. Use the minimum value as the upper limit occupancy rate of the deformation of the grid.

5. The integrated laser strike system for countering unmanned equipment as described in claim 1, characterized in that, The method for obtaining the relative translational linear velocity is as follows: The mechanical pitch and azimuth angular velocities of the servo base within the current control cycle are obtained. Combined with the system's preset optical scaling constant, the mechanical pitch angular velocity is converted into a longitudinal translational velocity component on the sensor target surface, and the azimuth angular velocity is converted into a lateral translational velocity component on the sensor target surface. The relative translational linear velocity is formed by the lateral translational velocity component and the longitudinal translational velocity component.

6. The integrated laser strike system for countering unmanned equipment as described in claim 2, characterized in that, The method for obtaining the wavefront phase compensation value is as follows: The two-dimensional wavefront slopes of each grid are concatenated into a one-dimensional global slope observation vector. The spatial anomaly suppression weights of each grid are filled into the corresponding main diagonal elements in the matrix where the horizontal wavefront slope arrangement sequence index position and the vertical wavefront slope arrangement sequence index position of the grid are located. The remaining positions are filled with zeros. In this way, a diagonal weighted matrix that matches the dimension of the long sequence of global slope observation vector is generated. The diagonal weighted matrix is ​​used as the observation weight matrix to calculate the corresponding covariance matrix. If the covariance matrix is ​​rank deficient and exhibits singular matrix characteristics, a preset micro-regularization scalar constant is injected into the full permutation of the main diagonal of the covariance matrix, and then the matrix inversion is performed to output a one-dimensional wavefront phase compensation value. If the covariance matrix is ​​not rank deficient, matrix inversion is performed directly to output a one-dimensional wavefront phase compensation value.

7. The integrated laser strike system for countering unmanned equipment as described in claim 1, characterized in that, The closed-loop control method for the bottom deformable mirror is as follows: Multiply the wavefront phase compensation value by the preset voltage conversion gain constant of the deformable mirror to calculate the single-drive voltage increment signal; Obtain the historical drive voltage values ​​of each underlying actuator retained from the previous control cycle, and accumulate them with the single drive voltage increment signal according to the actuator correspondence to obtain the currently issued drive voltage control signal. The drive voltage control signal is converted into an analog high voltage signal and applied to each actuator at the bottom layer.

Images