A method and system for dynamic control of a flexible interceptor net launcher array

Abstract

The present application relates to the field of flexible interception equipment simulation and launch control, in particular to a kind of dynamic control method and system for flexible interception net launch array, the method is first to the time-space alignment of the target state of multiple frames obtained, and the rotational speed is estimated to accurately predict target curve position;Subsequently, the launch control parameters are calculated and corrected in combination with the calibration table;Finally, the trigger time, initial speed and launch angle of each launch unit are dynamically adjusted in combination with the target timing and the measured launch state, and the cooperative scheduling of multiple launch units is realized.The present application closely combines launch control calculation and target motion prediction, significantly improves the accuracy of target position prediction, the accuracy of control parameter calculation and the reliability of overall launch timing control.

Description

Technical Field

[0001] This invention relates to the field of simulation and launch control of flexible interception equipment, specifically to a dynamic control method and system for a launch array of a flexible interception network. Background Technology

[0002] Flexible interception nets are a type of flexible capture device composed of netting and traction mass blocks. Multiple mass blocks pull the folded netting to unfold, forming a capture surface with a certain area and orientation in the space through which the moving target is expected to pass. This is suitable for interception operations that reduce the risk of rigid collisions. With increasing requirements for target speed and capture timeliness, related control has evolved from preset fixed trigger times, traction speeds, and launch angles to using dynamic models to predict the netting's center of mass, normal direction, and effective unfolding area online, and then adjusting launch parameters based on the target arrival time. To describe the unfolding of large-sized netting, existing models typically discretize the netting into numerous nodes and rope elements that only bear tension, and use time integration to calculate node motion and tension changes.

[0003] During actual launch, the net rapidly expands from a contracted state, and tension disturbances propagate along the net lines. Local ropes switch between slack and taut states, with multiple switches potentially occurring consecutively within a millisecond-level attitude output cycle. The controller needs to determine the moment when the net first reaches the specified opening area, centroid position, and normal orientation, aligning it with the moment the target enters the capture envelope. Due to limitations in online computation cycles, current practices often employ large fixed step sizes, updating the force state uniformly based on rope length at the end of each step. When a rope is taut then slack within a step, both ends of the interval may remain in the same state; when adjacent ropes switch sequentially, they may also be concentrated at the end of the step, leading to missed events or distorted sequence of events within the step.

[0004] At this point, the force changes caused by the earlier switching fail to participate in subsequent motion calculations. As the error propagates along the net, it causes deviations in the predicted time when the net's orientation and effective deployment area reach the threshold. The controller adjusts the triggering time and traction state accordingly, which can easily lead to the net not being fully deployed or having deviated from its suitable posture when the target actually enters the capture envelope. This results in a mismatch between the net's effective capture posture and the target's arrival time, leading to a capture effect that falls short of expectations. Therefore, accurately predicting the effective deployment time of the net within the online control cycle and reliably aligning the net's posture with the target's arrival time is a major technical problem that urgently needs to be solved for this type of flexible interception equipment. Summary of the Invention

[0005] To address the aforementioned main technical problems, in the first aspect, this invention proposes a dynamic control method for a flexible interceptor network launch array, comprising: S101, performing spatiotemporal alignment of multiple target states, estimating the rotation rate in the velocity direction and its error; S102, predicting the target curve position based on the rotation rate, querying a flight time calibration table, and obtaining a preliminary flight time through damping iteration; S103, calculating the predicted position on a straight line under the same flight time, using the difference in horizontal displacement between the predicted position and the predicted position on a curve as a maneuver correction, synthesizing an allowable error band characterizing the range of the predicted position error, and then... S104. Based on the final hit point and the position of each launch unit, query the flight time calibration table, deployment height calibration table, and ballistic gun-raising compensation angle calibration table to determine the required deployment height, exit velocity, flight time, and launch attitude. Eliminate unusable launch units and sort them by flight time. S105. According to the candidate order, use the actual position and exit velocity of the launch unit to correct the flight time, hit point, exit velocity, and launch attitude. After passing the check, issue control parameters and trigger launch when preparation is complete and the target status is verified.

[0006] Compared to traditional control methods that employ single-frame states, uniform linear extrapolation, fixed lead, or one-time trajectory solving, this invention can identify the target turning trend and its reliability within each tracking frame period, and smoothly determine the hit point based on the relationship between the maneuver correction and the prediction error. This improves the predictive adaptability to turning targets and avoids overcompensation when measurement fluctuations are large. Simultaneously, by incorporating the deployment height, exit velocity, flight time, launch attitude, and the actual position and available status of the launch unit into a unified solution and correction process, candidate units with actual launch conditions can be selected from multiple launch units. Before triggering, the latest target status is checked, thereby improving the consistency between control parameters and the actual trajectory, mechanism status, and target motion status. This enhances the effective coverage probability, dynamic response capability, and launch safety of the flexible interceptor network against maneuvering targets.

[0007] Furthermore, the target position and target velocity are time-aligned and converted to the same coordinate system based on the timestamp, and the target state with communication delay is compensated; the effective velocity vector is projected onto the horizontal plane, the velocity direction angle crossing the angular boundary is made continuous, the velocity direction angle sequence is fitted according to the actual sampling time, and the fitting slope and its standard error are used as the rotation rate and its error, respectively, and the average level velocity and the standard deviation of the horizontal velocity are calculated.

[0008] Compared to directly calculating motion trends using target data from different sampling times and coordinate systems, this invention reduces false displacements and false turns caused by spatiotemporal asynchrony through time alignment, coordinate unification, and communication delay compensation. It avoids abrupt changes when the angle crosses boundaries by processing the direction angle continuously and fits the data based on the actual sampling time, which can adapt to tracking data with unequal intervals or jitter. At the same time, it uses the fitting standard error and velocity dispersion to evaluate the reliability of the estimation, thereby improving the accuracy and stability of the rotation rate estimation and providing a more reliable data foundation for subsequent maneuver prediction.

[0009] Furthermore, when there is insufficient effective speed data, the average speed is lower than the preset speed threshold, the standard deviation of the horizontal speed is too large relative to the average speed, the standard error exceeds the allowable upper limit, the estimated centripetal acceleration exceeds the reasonable maneuver range of the target, or the target state data is abnormal, curve prediction is disabled and linear prediction is adopted; when a single frame is abnormal, the corresponding frame is removed and estimation continues; when there are consecutive abnormalities, the saved rotation rate estimation results are cleared.

[0010] Compared to continuing to use curve models for extrapolation when data quality deteriorates, this invention can automatically downgrade to linear prediction based on the amount of effective data, speed stability, fitting error, centripetal acceleration rationality, and abnormal state conditions, avoiding the amplification of unreliable rotation rates into large hit point deviations. At the same time, it performs local removal of single-frame anomalies and clears historical estimates for continuous anomalies, which can balance short-term anti-interference capability and long-term state reset capability, preventing abnormal data from continuously polluting subsequent control cycles, thereby improving prediction robustness and launch safety in complex communication and tracking environments.

[0011] Furthermore, the horizontal position is extrapolated using a horizontally coordinated turning method, and the vertical position is extrapolated according to the current vertical speed; when the rotation rate is lower than the small angular rate threshold or the target state does not meet the curve prediction conditions, a straight-line limit form is adopted; the flight time is initialized with the geometric center of the launch array and the nominal exit initial velocity, and the previous flight time is fused with the query results of the flight time calibration table through damping until convergence. When the calibration table is substantially out of bounds, it is marked as not being able to launch, and when it does not converge, it is marked as a low confidence result.

[0012] Compared to simply extrapolating linearly based on the current speed or using a fixed flight time, this invention coordinates the turning model to describe the target's horizontal maneuver while maintaining vertical motion prediction, making the curve hit point more consistent with the target's actual motion trend. When the angular rate is small or the prediction conditions are insufficient, a linear limit form is used to avoid numerical instability caused by small divisions and model switching. By coordinating the coupling relationship between flight time, predicted position, and calibration results through damped iteration, iterative oscillations can be reduced and solution stability can be improved. Furthermore, out-of-bounds and non-convergence situations are clearly identified, which helps to promptly prevent unexecutable or low-reliability launch decisions.

[0013] Furthermore, the allowable error band is obtained by propagating and synthesizing the target position measurement error, target velocity measurement error, and rotation rate estimation error, and the allowable error band is not lower than the position calibration error; no curve correction is applied when the maneuver correction amount does not exceed the allowable error band, the compensation weight is continuous and does not decrease when the maneuver correction amount increases, and the compensation weight does not increase when the prediction error increases; the final hit point is located between the straight line prediction position and the curve prediction position, and its vertical coordinate adopts the vertical position predicted by the curve.

[0014] Compared to rigidly switching between straight-line and curve predictions or making full corrections without considering prediction reliability, this invention propagates position, velocity, and rotation rate errors to the allowable error band and uses the position calibration error as the lower limit, ensuring that minor maneuvering differences are not misjudged as true steering by measurement noise. The compensation weight increases smoothly with the enhancement of dynamic evidence and is suppressed as the prediction error increases, allowing the final hit point to continuously change between the two types of prediction results. This reduces control parameter jumps, repeated mechanism adjustments, and overcorrections, while maintaining the consistency of vertical predictions, thereby improving the stability and noise resistance of hit point generation.

[0015] Furthermore, the slant range, launch azimuth, and geometric pitch angle are determined based on the relative position vector between the final hit point and the actual position of each launch unit, and the required deployment height is determined based on the mounting reference plane of each launch unit. When the deployment height calibration table is locally non-monotonic with the exit velocity, the allowable range of exit velocity is divided into multiple monotonic intervals and inverse interpolation is performed separately to retain the exit velocity that meets the deployment height requirements. When there are multiple feasible exit velocities, the exit velocity with the shorter flight time is given priority.

[0016] Furthermore, launch units that are excluded due to unreachable expected launch elevation angle or launch azimuth, out-of-range calibration table inputs, unmet required deployment height, excessive mechanism rotation time, or being in a faulty, locked, unloaded, or communication abnormal state; the remaining launch units are sorted by flight time, and when the flight time difference meets the preset proximity condition, the candidate order is determined in sequence according to the shorter mechanism rotation time, the lower exit initial velocity, and the launch unit number.

[0017] Furthermore, in each final correction, the predicted position of the curve and the predicted position of the straight line are calculated separately under the same flight time, the compensation weight and the corrected hit point are re-determined, the flight time is updated according to the slant distance from the launch unit to the corrected hit point, and the exit initial velocity is corrected according to the updated required deployment altitude, until the flight time and the exit initial velocity converge together; if the current launch unit does not meet the launch conditions, the next launch unit is processed according to the candidate order, and if all candidate launch units do not meet the conditions, a command to temporarily not launch is output.

[0018] Furthermore, after the final calibration is completed, the launch azimuth and geometric elevation angle are determined based on the calibrated hit point. The ballistic gun-raising compensation angle is obtained by querying the ballistic gun-raising compensation angle calibration table and superimposed on the geometric elevation angle. The launch azimuth, final launch elevation angle, and exit velocity are then sent to the launch unit. Upon receiving the preparation completion signal, the preparation completion time and the latest target status are checked. When both meet the preset recalculation conditions, the triggering is canceled and the dynamic control method is re-executed based on the latest target status.

[0019] In a second aspect, the present invention provides a dynamic control system for a flexible interceptor network launch array, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a dynamic control method for a flexible interceptor network launch array according to the present invention is implemented.

[0020] The technical advantages of this invention are as follows: This invention focuses on the dynamic launch control of maneuvering targets using a flexible interceptor network. Based on the rotation rate and prediction error, it implements continuous and adaptively reliable compensation between straight-line and curved predictions, and incorporates the interceptor network deployment height, exit velocity, flight time, launch attitude, and the actual state of each unit into the solution. Compared to fixed lead and one-time calculation methods, this invention improves the accuracy of maneuvering target prediction, parameter feasibility, multi-unit scheduling efficiency, and launch safety. Attached Figure Description

[0021] Figure 1 This is a flowchart of a dynamic control method for a flexible interceptor network transmitting array according to an embodiment of the present invention;

[0022] Figure 2 This is a flowchart of the candidate order generation and final correction switching of the transmitting unit in this invention; Figure 3 This is a block diagram of a dynamic control system for a flexible interceptor network transmitting array according to an embodiment of the present invention. Detailed Implementation

[0023] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] Example 1: like Figure 1 As shown, a dynamic control method for a flexible interceptor network transmitting array according to the present invention includes: This embodiment executes the target tracking frame cycle in a loop, and in each tracking frame, it sequentially completes target motion state estimation, hit position prediction, launch unit selection, control parameter correction, and launch triggering.

[0025] In this embodiment, the radar and photoelectric sensors are synchronized in time and fused with data to achieve approximately... The equivalent frame rate outputs the target's 3D position, 3D velocity, and corresponding measurement errors. The solution control unit uses a processor supporting double-precision floating-point operations, with an operating frequency of [missing information]. The flexible interceptor network launch array includes... Each launch unit has a pre-calibrated and stored spatial position, mechanical elevation range, azimuth range, and exit muzzle velocity range. The adjustable range of the exit muzzle velocity is... to The mechanically feasible range of the pitch angle is to .

[0026] The calculation and control unit pre-stores flight time calibration tables, deployment altitude calibration tables, and ballistic gun-raising compensation angle calibration tables. Each calibration table uses the slant distance from the launch unit to the point of impact. and exit initial velocity As input, the values ​​output are the flight time, deployment altitude, and ballistic firing compensation angle of the flexible interceptor net. The calibration table above is determined by both ballistic simulation results and actual launch test data, and the values ​​between adjacent calibration nodes are obtained through bilinear interpolation.

[0027] S101. Obtain the target state in multiple frames and estimate the rotation rate in the velocity direction.

[0028] The target position and velocity are time-aligned according to the timestamps carried by each sensor data, and then uniformly converted to the same coordinate system. For data with fixed communication delays, the target state is compensated to the current solution time based on the timestamp, avoiding the direct combination of position and velocity data from different sampling times.

[0029] Within each tracking frame, retain the most recent... Frame effective velocity vector. At the tracking frame rate, Frame data coverage approximately This timeframe covers the continuous turning processes common in small drones, while avoiding significant lag in estimation results due to excessively long data windows. The data used for estimation can be adjusted to the most recent values ​​based on the maneuverability of the intercepted target and the sensor noise level. to frame.

[0030] Projecting the velocity vectors of each frame onto the horizontal plane yields the horizontal velocity and velocity direction angle. The velocity direction angle sequence is then processed to be continuous. When the direction angles of two adjacent frames cross... At the boundary, increase or decrease the direction angle of the next frame. This eliminates numerical jumps caused by the range of angle representation.

[0031] Based on the actual sampling times of each frame, a first-order least-squares fit is performed on the continuous velocity direction angle sequence. The slope of the fitted line is used as the estimated rotational rate of the target velocity direction, denoted as the turning angular rate estimate. The standard error of the fitted slope is denoted as . .in, The absolute value reflects how quickly the target's velocity direction changes. The positive or negative sign reflects the turning direction of the target in the horizontal plane. This is used to reflect the degree to which the current angular rate estimate is affected by measurement noise.

[0032] The average horizontal velocity is obtained by averaging the horizontal velocities involved in the fitting. Simultaneously, the standard deviation of the horizontal velocity is calculated. The standard deviation of the horizontal velocity is a general statistic, used only to determine whether the target velocity has been relatively stable over recent frames; no new evaluation index is constructed based on the average horizontal velocity and the standard deviation of the horizontal velocity.

[0033] If any of the following conditions occur, the current frame is determined not to meet the curve prediction conditions, and the turning angular rate of the current frame will not be used for curve compensation: Effective speed data is less than Frames; average speed is lower than This makes the velocity direction susceptible to measurement noise; the standard deviation of the horizontal velocity exceeds the average level velocity. ; angular rate fitting standard error The target's centripetal acceleration, estimated based on average speed and turning rate, exceeds the reasonable maneuvering range of the intercepted target; or the target's position, speed, or sampling time exhibits anomalies that exceed preset physical ranges.

[0034] For a single frame with missing or abnormal data, that frame is removed from the fitted sequence, and the remaining valid frames are used to continue estimation. A series of abnormal frames are then counted. At frame time, previously saved turning rate estimates are cleared, and only constant speed predictions are retained in this frame to prevent outdated target maneuver information from affecting the hit position.

[0035] S102. Iteratively determine the initial flight time and the predicted position of the curve based on the rotation law in the velocity direction.

[0036] This embodiment uses a horizontal coordinated turn as the basic form of target curvilinear motion. This motion form is suitable for situations where the target's horizontal velocity magnitude is basically stable for a short period of time, while the velocity direction changes continuously. The target's vertical position is extrapolated briefly based on the current vertical velocity.

[0037] Let the target's three-dimensional position at the start of the prediction be... The current horizontal velocity direction angle is The current vertical velocity is At a given flight time Under the given conditions, the predicted position of the target curve Determined according to the following formula: ; In the formula, This represents the estimated turning radius, including the turning direction. The above expression assumes the target velocity direction changes according to the turning angular rate over the flight time. Because of continuous rotation, the predicted trajectory is an arc corresponding to the target's turning state, rather than a straight line in constant velocity prediction or a fixed-direction parabola in constant acceleration prediction.

[0038] when When the absolute value is less than the preset small angular velocity threshold, the above-mentioned values ​​are not directly calculated. The expression for the denominator, and the use of this expression in The linear limit form approaching zero. In this embodiment, the small angular rate threshold is taken as... This can avoid numerical instability under low angular rate conditions and also ensure that the predicted position remains continuous when the target switches between linear and turning motions.

[0039] When the target's motion state does not meet the curve prediction conditions in S101, the target's predicted position is calculated in the form of a straight line, without coordinating the turning curve extrapolation.

[0040] Since the flight time of the flexible interceptor depends on the slant distance from the launch position to the predicted hit position, and the predicted hit position changes with the flight time, a damped iterative method is used to obtain a consistent flight time and predicted position.

[0041] First, the geometric center of the transmission array is used as a reference position. The nominal exit initial velocity is the median value of the range. The initial flight time is obtained by consulting the flight time calibration table based on the slant distance from the target's current position to the reference position. .

[0042] The flight time is then updated according to the following formula: ; In the formula, For the first The flight time obtained in the next iteration Indicates the flight time calibration table, This is the damping coefficient. The damping coefficient is used to limit the time-of-flight variation between two adjacent iterations, preventing the predicted position and slant range from repeatedly jumping near the calibration table boundaries.

[0043] In this embodiment, the damping coefficient Pick , can to Adjust between them. The difference in flight time between two adjacent iterations should not exceed [a certain value]. When the iteration converges, the maximum number of iterations is set to [value]. Next. Typically, after to A stable result can be obtained in one iteration.

[0044] If the required slant range exceeds the effective range of the time-of-flight calibration table during the iteration process, the iteration is stopped and the frame is marked as unlaunchable. If the input slightly exceeds the calibration table boundary only due to floating-point arithmetic errors, interpolation can be performed according to the nearest boundary node; when the target slant range substantially exceeds the launch capability, the boundary node values ​​are not used to replace the actual ballistic results.

[0045] This step yields the preliminary flight time and the corresponding predicted curve position. If in If the convergence condition is not met after the second iteration, the result of the last iteration is retained for subsequent feasibility assessment, and it is marked as a low-confidence result without applying full curve compensation.

[0046] S103. Determine the final hit point based on the curve correction amount and prediction error.

[0047] Using the preliminary flight time obtained from S102, a constant-speed linear extrapolation is performed based on the target's current position and velocity to obtain the predicted linear position for the same flight time. The horizontal displacement difference between the predicted curved position and the predicted linear position is used as the maneuver correction amount that needs to be applied after considering the target's velocity direction rotation.

[0048] According to the standard error of the angular rate fitting First-order error propagation is performed to convert the angular rate estimation error into a possible lateral position error during the flight time. This lateral position error increases with increasing average horizontal velocity and with the square of the flight time.

[0049] The position error resulting from the target position measurement error, the velocity measurement error during flight time, and the lateral position error resulting from the angular rate estimation error are combined to obtain the allowable error band for this curve prediction. When the sensor outputs the measurement covariance, error propagation and synthesis are performed according to the measurement covariance; when the sensor does not directly output the measurement covariance, the position standard error and velocity standard error obtained from the sensor calibration test are used for calculation.

[0050] The allowable error band is no less than the sensor's position calibration error, to avoid applying full curve compensation incorrectly due to abnormally underestimated measurement error.

[0051] The compensation weight is determined based on the relative magnitude of the maneuver correction amount and the allowable error band. The specific processing method is as follows: When the maneuver correction amount is not greater than the allowable error band, the compensation weight is taken as follows: Position is predicted using a constant velocity linear motion. When the maneuver correction amount is not less than the allowable error band When the compensation weight is multiplied, the compensation weight is taken as follows: A complete curve correction is used; When the maneuver correction is greater than the allowable error band but less than the allowable error band When the compensation weight is multiplied by 1, the compensation weight is at 1. to The numbers increase continuously between them; When the standard deviation of horizontal velocity exceeds the average speed Angular rate fitting standard error If the allowed upper limit is exceeded, or if S102 fails to converge within the specified number of iterations, the compensation weight is adjusted. .

[0052] In the third scenario, this embodiment uses a linearly increasing method to determine the compensation weight, such that the compensation weight is equal to the allowable error band when the maneuver correction amount is equal to the allowable error band. Continuous connection, and when the motor correction amount is equal to the allowable error band. Double time and Continuous connection. Alternatively, a pre-calibrated monotonic smooth curve can be used, but the curve should satisfy the following conditions: the compensation weight does not decrease when the maneuver correction increases, and the compensation weight does not increase when the prediction error increases.

[0053] The final hit point starts from the predicted position on the constant speed straight line and moves in the direction pointing to the predicted position on the curve. The distance moved is equal to the product of the maneuver correction and the compensation weight. Since the coordinated turning prediction in this embodiment mainly acts on the horizontal plane, the vertical coordinates of the final hit point are obtained from the vertical prediction result in S102.

[0054] Through the above processing, curve correction will only fully participate in the hit point calculation when the target velocity direction rotates continuously and the turning rate estimate has sufficient credibility; when the target is basically flying straight, the target velocity changes significantly, the measurement data jumps, or the flight time iteration does not converge, the final hit point will automatically approach or regress to the constant velocity straight line prediction position.

[0055] S104. Generate the candidate order of launch units based on the final hit point.

[0056] The spatial position of each launching unit is read separately. For any launching unit, the position of the launching unit is subtracted from the position of the final hit point to obtain the relative position vector between the two. The magnitude of the relative position vector is used as the slant distance from the launching unit to the final hit point, the horizontal projection direction of the relative position vector is used as the launching azimuth, and the angle between the relative position vector and the horizontal plane is used as the geometric pitch angle.

[0057] The vertical height difference between the final hit point and the installation position of the corresponding launch unit is used as the required deployment height of the flexible interceptor network. For launch arrays with different installation elevations, each launch unit uses its own installation reference plane as the height reference and performs coordinate conversion based on the pre-measured installation elevation, rather than directly using a uniform absolute height to replace the relative deployment height.

[0058] Within the permissible range of the exit initial velocity, consult the deployment height calibration table to find the exit initial velocity that allows the flexible interceptor net to reach the required deployment height at the corresponding slant distance. When the required deployment height is between adjacent calibration nodes, determine the exit initial velocity using bilinear inverse interpolation.

[0059] When the deployment height calibration table is not globally monotonic with respect to the exit initial velocity, the range of the exit initial velocity is divided into multiple monotonic intervals. Inverse interpolation is performed within each monotonic interval, and all exit initial velocities that meet the deployment height requirements are retained. This avoids missing feasible solutions due to local non-monotonic variations in the calibration table.

[0060] When multiple feasible exit velocities correspond to the same slant range and deployment height, query the flight time corresponding to each exit velocity, prioritizing solutions with shorter flight times. If no exit velocity meets the deployment height requirement, or if the inverse interpolation result exceeds the limit... to If the allowed range is not met, the launching unit is deemed unusable, and a feasible solution is not formed on the surface by forcibly cutting off the exit initial velocity to the boundary.

[0061] Using the determined exit velocity and slant range, the flight time corresponding to this launch unit is obtained from the flight time calibration table; then, the corresponding ballistic gun-raising compensation angle is obtained from the ballistic gun-raising compensation angle calibration table. The geometric elevation angle is added to the ballistic gun-raising compensation angle to obtain the expected launch elevation angle.

[0062] The corresponding transmitting unit shall be removed if any of the following conditions are met: The launch pitch angle is expected to exceed the mechanical pitch range of the launch unit; the launch azimuth is beyond the actual reach of the launch unit; the exit velocity or flight time exceeds the effective range of the corresponding calibration table; the required deployment height cannot be met by the launch unit; the rotation time calculated based on the current mechanism attitude is greater than the allowable preparation time; the launch unit is in a faulty, locked, unloaded, or communication abnormal state.

[0063] The remaining launch units are arranged in ascending order of flight time. Shorter flight times can reduce the positional changes caused by the target's continued maneuvering during flight through the flexible interception net, and also reduce the cumulative impact of turn rate estimation errors during flight.

[0064] When the difference in flight time between the two launch units does not exceed When the time required to rotate from the current mechanism attitude to the expected launch attitude is short, the launch unit with the shortest time is selected. If the rotation time is still the same, the launch unit with the lower exit initial velocity is selected. If they still cannot be distinguished, the fixed order is determined according to the launch unit number.

[0065] By using the above step-by-step determination method, a stable and unique candidate order can be formed when the ballistic conditions of multiple launch units are similar, thus avoiding frequent switching of launch units between adjacent tracking frames.

[0066] S105. The final correction and launch are completed using the actual position of the candidate launch unit and the initial velocity of the exit.

[0067] According to the candidate order determined in S104, the first-ranked launch unit is selected. The actual position of this launch unit replaces the array reference position in S102, and the exit initial velocity corresponding to this launch unit replaces the nominal exit initial velocity. The flight time and the predicted target position are then recalculated.

[0068] During the final calibration process, each iteration calculates the predicted position on the curve and the predicted position on the constant velocity straight line under the same flight time, and re-determines the compensation weights according to the judgment method in S103. The calibrated hit point is determined based on the compensation weights, and the flight time is updated by querying the flight time calibration table based on the actual slant distance from the launch unit to the calibrated hit point.

[0069] Therefore, the final flight time is consistent with the hit point after the compensation weight correction, rather than just with the curve prediction position without confidence constraints.

[0070] Recalculate the slant range and required deployment height based on the updated hit point, and correct the exit initial velocity using the deployment height calibration table. The change in exit initial velocity between two consecutive corrections should not exceed [a certain value]. And the flight time variation does not exceed At that time, the final correction is determined to be converged. In this embodiment, the joint correction of the exit initial velocity and flight time is performed at most once. wheel.

[0071] After final calibration, the initial velocity, flight time, deployment altitude, launch azimuth, and launch elevation angle are checked again to ensure they are within acceptable ranges. If the current launch unit no longer meets the launch conditions after final calibration, it is removed from the candidate sequence, and the same process is applied to the next launch unit. If all candidate launch units fail to meet the conditions, the launch command is temporarily withheld from the tracking frame output.

[0072] For a firing unit that meets the conditions, the firing azimuth and geometric elevation angle are redefined based on the final hit point. The ballistic lift-off compensation angle obtained by querying the final slant range and initial velocity is then superimposed on the geometric elevation angle to form the final firing elevation angle. The solution control unit then sends the firing azimuth, firing elevation angle, and initial velocity to the firing unit.

[0073] After the launching mechanism reaches the target attitude and returns a ready-to-complete signal, the calculation and control unit verifies the latest target status. If the difference between the actual ready-to-complete time and the estimated trigger time used in the calculation does not exceed [the specified value], then [the calculation is complete]. If the latest target status does not cause a significant change in the hit position, then the launch is triggered according to the final control parameters.

[0074] If the difference between the actual preparation completion time and the expected trigger time exceeds Or, during the preparation of the mechanism, new tracking frames received indicate that the predicted target position has changed by more than [a certain amount]. If the current triggering fails, the method of this embodiment will be re-executed based on the latest target state.

[0075] Pre-trigger verification prevents the use of expired hit points after changes in mechanism rotation time, communication delays, or target motion states. After launch, the actual exit velocity, actual mechanism angle, and actual trigger time are recorded for subsequent offline calibration table correction; data not verified by testing is not directly used for online parameter updates within the current launch cycle.

[0076] like Figure 2 As shown, it illustrates the process of generating the candidate order of launch units and switching to final correction. This mainly includes determining the launch conditions for each launch unit and including those that meet the conditions in the candidate order. After generating the candidate order, the launch unit at the top of the list is selected for final correction. If the conditions are still met after correction, the final control parameters are issued; otherwise, the current launch unit is removed and the process switches to the next candidate launch unit. When all candidate launch units are exhausted, a command to temporarily not launch is output.

[0077] Example 2: This implementation is applicable to application scenarios that require clear control states, facilitate state recording and logic backtracking, or do not want the compensation weights to change continuously within adjacent tracking frames.

[0078] This embodiment follows The tracking frame period is executed cyclically. The transmission array includes... Each launch unit has a known location, and the allowable range of exit muzzle velocity, mechanical pitch range, and azimuth range for each launch unit are pre-stored. The flight time calibration table, deployment altitude calibration table, and ballistic gun-raising compensation angle calibration table were established through ballistic simulation and actual launch tests, respectively.

[0079] S201. Obtain the target state and estimate the turning rate.

[0080] Align the target position, target velocity, and timestamp output from radar and photoelectric sensors, retaining the most recent one. Frame effective velocity vector. Project the velocity vectors of each frame onto the horizontal plane to obtain the horizontal velocity and velocity direction angle, and then perform cross-frame velocity vector calculations. The direction angles of the boundary are made continuous.

[0081] The velocity direction angle sequence is fitted using first-order least squares method based on the sampling time of each frame to obtain the estimated turning angular rate. and the standard error of the fit Simultaneously, the average value of the horizontal velocities involved in the fitting is calculated. and standard deviation.

[0082] There are fewer than 10 valid data points. Frame rate, average speed is lower than The standard deviation of horizontal velocity exceeds the average level velocity. Fit standard error If the target centripetal acceleration exceeds the allowable upper limit, or if the target centripetal acceleration estimated based on the average speed and turning rate exceeds a reasonable range, the frame will be determined as not meeting the complete curve compensation condition.

[0083] Single-frame outliers are removed from the fitted sequence. Continuous When frame data is abnormal, clear previously saved turning rate estimates to prevent the continued use of outdated target maneuver states after tracking is resumed.

[0084] S202. Determine the predicted position of the curve and the initial flight time.

[0085] Let the target's position at the start of the prediction be... The current horizontal velocity direction angle is The current vertical velocity is At a given flight time Under the given conditions, the predicted position of the target curve is determined according to the following formula: ; when The absolute value is less than When using the above expression in The linear limit form approaching zero avoids the risk of division by zero under small angular rate conditions and ensures a continuous transition between linear and curvilinear predictions.

[0086] Using the geometric center of the transmission array as a reference position The nominal exit initial velocity is the median value of the range. The initial flight time is obtained by consulting the flight time calibration table based on the slant distance from the target's current position to the reference position. .

[0087] The flight time is damped iteratively calculated using the following formula: ; In this embodiment, the damping coefficient Pick The time-of-flight convergence threshold is taken as The maximum number of iterations is taken as If the required slant range for the iteration exceeds the valid range of the time-of-flight calibration table, then stop the calculation for this frame and output a command to temporarily not launch.

[0088] If in If the convergence condition is not met within the next iteration, the result of the last calculation is retained for subsequent judgment, and the current frame is determined to not meet the complete curve compensation condition.

[0089] S203. Determine the hit point according to the three-level compensation method of high, medium and low.

[0090] The preliminary flight time obtained from S202 is used for constant-speed linear extrapolation to obtain the predicted position on the straight line. The difference in horizontal displacement between the predicted position on the curve and the predicted position on the straight line is used as the maneuver correction.

[0091] According to the standard error of the angular rate fitting First-order error propagation is performed on the target position measurement error and velocity measurement error to determine the allowable error band for curve prediction. The allowable error band is not lower than the sensor's position calibration error and expands with increasing flight time.

[0092] Based on the maneuver correction amount, allowable error band, angular rate fitting error, and the stability of horizontal velocity across multiple frames, this frame is divided into a high-compensation state, a medium-compensation state, or a low-compensation state.

[0093] The system enters a high-compensation state when the following conditions are met simultaneously: the flight time iteration has converged; and the standard deviation of the horizontal velocity does not exceed the average horizontal velocity. ; angular rate fitting standard error Not exceeding half of the allowable upper limit; the maneuver correction amount is not less than the allowable error band. The target centripetal acceleration estimated based on average speed and turning rate is within a reasonable range. Under high compensation, the predicted position of the complete curve is used as the hit point.

[0094] In the absence of a situation where curve compensation must be turned off, the system enters the intermediate compensation state when any of the following conditions are met: the maneuver correction is greater than the allowable error band, but less than the allowable error band. Times; the standard deviation of horizontal velocity is greater than the average level velocity. But not exceeding the average speed ; angular rate fitting standard error It is greater than half of the allowable upper limit, but does not exceed the allowable upper limit. Under medium compensation state, the horizontal displacement difference between the predicted position on the constant velocity straight line and the predicted position on the curve is moved by half, and the point after the movement is taken as the hit point.

[0095] Low compensation mode is entered when any of the following conditions occur: the maneuver correction is not greater than the allowable error band; the standard deviation of horizontal speed exceeds the average level speed. ; angular rate fitting standard error Exceeding the allowed limit; time-of-flight iterations failing to converge within the specified number of iterations; effective velocity data less than [amount missing]. Frame; The target's motion state does not meet the applicable conditions for horizontal coordinated turning. Under low compensation conditions, no curve correction is applied, and the predicted position of constant velocity straight line is directly used as the hit point.

[0096] To prevent frequent switching between high, medium, and low compensation states between adjacent tracking frames when the target state is near the grade boundary, state upgrades must be continuous. If a frame meets the conditions for a higher level, the state downgrade will be executed immediately when the corresponding downgrade condition occurs in the current frame.

[0097] For example, when transitioning from a medium-compensation state to a high-compensation state, continuous... The frame meets the high compensation condition; when the high compensation condition fails, it can be directly reduced to a medium or low compensation state in the current frame. This process makes the compensation enhancement process relatively smooth and can promptly cancel unreliable curve corrections when measurement quality deteriorates.

[0098] S204. Determine the candidate order of the launch unit based on the hit point.

[0099] For each launch unit, calculate its slant range, launch azimuth, and geometric elevation angle to the impact point. Use the vertical height difference between the impact point and the launch unit as the required deployment height, and perform inverse interpolation on the deployment height calibration table within the allowable range of the exit initial velocity.

[0100] When the deployment altitude calibration table has multiple monotonic intervals relative to the exit initial velocity, inverse interpolation is performed within each monotonic interval, and all exit initial velocities that meet the deployment altitude requirements are retained. If multiple feasible exit initial velocities exist for the same launch unit, the corresponding flight times are queried, and the exit initial velocity with the shorter flight time is retained.

[0101] If a launch unit is found to be in a state where there is no feasible exit velocity, the exit velocity exceeds the allowable range, the hit slant range exceeds the calibration table range, the expected launch elevation angle exceeds the mechanical range, the azimuth angle is unreachable, the mechanism rotation time exceeds the allowable preparation time, or the launch unit is not in a usable state, the corresponding launch unit shall be eliminated.

[0102] All available launch units are arranged in ascending order of flight time. The flight time difference between launch units does not exceed [a certain threshold]. When the mechanism rotation time is the same, the launching unit with the shorter mechanism rotation time is selected first; when the mechanism rotation time is still the same, the launching unit with the lower exit initial velocity is selected first; when they still cannot be distinguished, the fixed order is determined according to the launching unit number.

[0103] S205. Complete the final calibration and triggering under actual launch unit conditions.

[0104] Starting with the first launch unit in the candidate sequence, the flight time is recalculated iteratively using the actual position and initial exit velocity of that launch unit. In each iteration, the predicted curve position, the predicted constant velocity straight line position, and the allowable error band are recalculated, and the compensation state is redefined according to the conditions described in S203.

[0105] When the final correction changes the compensation state, the hit point is immediately re-determined according to the changed compensation state, and the hit point before correction is not used. For example, if the initial calculation shows a high compensation state, but after using the flight time corresponding to the actual launch unit, the allowable error band increases and only the medium compensation condition is met, then half-curve correction is used to determine the final hit point, and the full curve correction is no longer used.

[0106] Recalculate the slant range and required deployment height based on the updated hit point, and correct the exit initial velocity using the deployment height calibration table. The change in exit initial velocity between two consecutive calculations should not exceed [a certain value]. And the flight time variation does not exceed At this point, the final correction is considered complete. The joint correction can be performed at most once. wheel.

[0107] After final calibration, the launch azimuth, launch elevation, exit velocity, deployment altitude, and flight time are checked again. If the current launch unit does not meet any constraint, the next launch unit is selected according to the candidate order and recalculated. If all candidate launch units are unavailable, launch is not triggered in this frame.

[0108] For the final determined launch unit, the ballistic gun lifting compensation angle is superimposed on the geometric elevation angle to form the final launch elevation angle, and the launch azimuth angle, the final launch elevation angle, and the exit initial velocity are sent to the actuator.

[0109] Once the implementing agency has provided feedback on the preparation, it verifies the latest target status and the actual preparation completion time. If the difference between the actual preparation completion time and the expected trigger time exceeds [a certain threshold], [further action will be taken]. Or the latest target prediction position changes more than the previously calculated hit point. If the trigger is not triggered, the current trigger will be cancelled and the calculation will be recalculated; otherwise, the launch will be triggered.

[0110] This embodiment outputs only three explicit states: complete curve compensation, half curve compensation, and no curve compensation. This facilitates the upper-level control device in recording the current prediction method and also reduces the computational load of continuous weights during multi-target parallel calculations. By using a process of confirming the state upgrade for two consecutive frames and taking effect immediately for the state downgrade, it is possible to maintain stable transmission pointing while promptly exiting curve predictions with insufficient reliability.

[0111] Example 3: This invention also provides a dynamic control system for a flexible interceptor network launch array. For example... Figure 3 As shown, a dynamic control system for a flexible interceptor network launch array includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a dynamic control method for a flexible interceptor network launch array according to the first aspect of the present invention.

[0112] A dynamic control system for a flexible interceptor network launch array also includes other components well known to those skilled in the art, such as communication interfaces, the setup and functions of which are known in the art and will not be described in detail here.

Claims

1. A dynamic control method for a flexible interceptor network transmitting array, characterized in that, The method includes: S101, performing spatiotemporal alignment of target states across multiple frames, and estimating the rotation rate in the velocity direction and its error; S102. Predict the target curve position based on the rotation rate, query the flight time calibration table, and obtain the preliminary flight time through damping iteration; S103. Calculate the predicted position on a straight line under the same flight time, and use the difference in horizontal displacement between it and the predicted position on a curve as the maneuver correction amount. Synthesize the allowable error band that characterizes the range of error in the predicted position, determine the continuous compensation weight according to the relative size of the two, and obtain the final hit point. S104. Based on the final hit point and the position of each launching unit, query the flight time calibration table, deployment height calibration table and ballistic gun lifting compensation angle calibration table to determine the required deployment height, exit velocity, flight time and launch attitude, eliminate unusable launching units and sort them by flight time. S105. The flight time, impact point, exit velocity and launch attitude are corrected according to the actual position and exit velocity of the launch unit in the candidate order. After the check is passed, the control parameters are issued and the launch is triggered when the preparation is completed and the target status is verified.

2. The dynamic control method for a flexible interceptor network transmitting array according to claim 1, characterized in that, The target position and target velocity are time-aligned based on the timestamp and converted to the same coordinate system to compensate for the target state with communication delay; The effective velocity vector is projected onto the horizontal plane, and the velocity direction angles crossing the angular boundary are made continuous. The velocity direction angle sequence is fitted according to the actual sampling time. The fitting slope and its standard error are used as the rotation rate and its error, respectively. The average horizontal velocity and the standard deviation of the horizontal velocity are calculated.

3. The dynamic control method for a flexible interceptor network transmitting array according to claim 2, characterized in that, When there is insufficient effective speed data, the average speed is lower than the preset speed threshold, the standard deviation of the horizontal speed is too large relative to the average speed, the standard error exceeds the allowable upper limit, the estimated centripetal acceleration exceeds the reasonable maneuver range of the target, or the target state data is abnormal, curve prediction is disabled and straight line prediction is used. When a single frame is abnormal, the corresponding frame is removed and estimation continues. When there are consecutive abnormalities, the saved rotation rate estimation results are cleared.

4. The dynamic control method for a flexible interceptor network transmitting array according to claim 1, characterized in that, The horizontal position is pushed outward using a horizontally coordinated turning method, and the vertical position is pushed outward according to the current vertical speed; When the rotational rate is lower than the small angular rate threshold or the target state does not meet the curve prediction conditions, the linear limit form is adopted. The flight time is initialized with the geometric center of the launch array and the nominal exit initial velocity. The previous flight time is fused with the query results of the flight time calibration table through damping until convergence. If the calibration table is substantially out of bounds, it is marked as not being able to launch. If it does not converge, it is marked as a low confidence result.

5. The dynamic control method for a flexible interceptor network transmitting array according to claim 1, characterized in that, The allowable error band is obtained by propagating and synthesizing the target position measurement error, the target velocity measurement error, and the rotation rate estimation error, and the allowable error band is not lower than the position calibration error; No curve correction is applied when the maneuver correction amount does not exceed the allowable error band; the compensation weight is continuous and does not decrease when the maneuver correction amount increases; and the compensation weight does not increase when the prediction error increases. The final hit point is located between the predicted position on the straight line and the predicted position on the curve, and its vertical coordinates are the vertical position predicted by the curve.

6. The dynamic control method for a flexible interceptor network transmitting array according to claim 1, characterized in that, The slant range, launch azimuth, and geometric elevation angle are determined based on the relative position vector between the final hit point and the actual position of each launch unit, and the required deployment height is determined based on the mounting reference plane of each launch unit itself. When the deployment height calibration table is locally non-monotonic with the exit initial velocity, the allowable range of the exit initial velocity is divided into multiple monotonic intervals and inverse interpolation is performed separately. The exit initial velocity that meets the deployment height requirements is retained, and when there are multiple feasible exit initial velocities, the exit initial velocity with the shorter flight time is retained first.

7. The dynamic control method for a flexible interceptor network transmitting array according to claim 6, characterized in that, Eliminate launch units that are not expected to launch elevation angle or launch azimuth, have calibration table inputs that are out of bounds, cannot meet the required deployment height, have exceeded the mechanism rotation time limit, or are in a faulty, locked, unloaded, or communication abnormal state. The remaining launch units are sorted by flight time. When the flight time difference meets the preset proximity condition, the candidate order is determined in turn according to the shorter mechanism rotation time, the lower exit initial velocity, and the launch unit number.

8. The dynamic control method for a flexible interceptor network transmitting array according to claim 1, characterized in that, In each final correction, the curve prediction position and the straight line prediction position are calculated separately under the same flight time. The compensation weight and the corrected hit point are re-determined. The flight time is updated according to the slant distance from the launch unit to the corrected hit point. The exit velocity is corrected according to the updated required deployment altitude until the flight time and exit velocity converge together. If the current launch unit does not meet the launch conditions, the next launch unit will be processed according to the candidate order. If none of the candidate launch units meet the conditions, a command to temporarily not launch will be output.

9. A dynamic control method for a flexible interceptor network transmitting array according to claim 8, characterized in that, After the final calibration is completed, the launch azimuth and geometric elevation angle are determined based on the calibrated hit point. The ballistic gun-raising compensation angle is obtained by looking up the ballistic gun-raising compensation angle calibration table and superimposed on the geometric elevation angle. The launch azimuth, final launch elevation angle and exit velocity are then sent to the launch unit. Upon receiving the preparation completion signal, the preparation completion time and the latest target state are checked. If both meet the preset recalculation conditions, the triggering is canceled and the dynamic control method is re-executed according to the latest target state.

10. A dynamic control system for a flexible interceptor network launch array, characterized in that, It includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, they implement the dynamic control method for a flexible interceptor network launch array as described in any one of claims 1 to 9.

Images