Equipment target control data acquisition system and acquisition method thereof

Abstract

The present invention relates to the field of data acquisition technology, specifically to an equipment target control data acquisition system and an acquisition method thereof. The system includes: a state potential energy module, a velocity projection module, a trajectory fusion module, a collaborative preloading module, and an acquisition trigger module. In the present invention, a potential function sequence is constructed by acquiring the angular velocity change rate and angular momentum, identifying the change and comparing the fallback point to determine the characteristic stable segment, improving the attitude stability characterization and feature capture under interference, using inertial navigation velocity projection filtering to calculate the angular rate, comparing and determining the motion mode analysis segment, focusing data quality on key intervals, integrating radar coordinate trajectory identification composite segments, and determining multi-node early warning segments with analysis segment time overlap, combining internal and external information to enhance early warning accuracy, generating unified scheduling instructions for collaborative preloading to improve timeliness, activating sensor recording and storage according to instructions, ensuring accurate and lossless collaborative acquisition during critical periods, and improving data integrity.

Description

Technical Field

[0001] The present invention relates to the technical field of data acquisition, and in particular to an equipment target control data acquisition system and an acquisition method thereof. Background Art

[0002] The field of data acquisition technology encompasses physical signal perception, analog signal conversion, digital signal transmission, and storage processing. Its core areas include optimizing sensor signal acquisition accuracy, integrating multi-source heterogeneous data, improving real-time transmission stability, and storing massive amounts of data. This area systematically integrates sensor hardware design, signal conditioning circuit development, analog-to-digital conversion interface optimization, communication protocol adaptation, and database architecture development. Through hardware circuit anti-interference design, dynamic sampling rate adjustment, embedded data compression algorithms, and transmission bandwidth allocation, it provides a complete technical chain from signal source to storage terminal.

[0003] One of the equipment target control data acquisition systems refers to a special acquisition device designed for the target status monitoring needs in weapon test scenarios. Technical matters include the synchronous acquisition of multiple parameters such as target displacement, shock wave intensity, and temperature gradient. It adopts a dynamic binding mechanism between control instructions and data acquisition channels, establishes multi-channel independent trigger acquisition timing through programmable logic devices, combines differential signal transmission to eliminate electromagnetic interference, uses a data packet verification and retransmission mechanism to ensure transmission integrity, automatically switches the range according to the preset threshold, and caches high-frequency data without loss based on the hardware ring buffer.

[0004] Existing data acquisition technologies have limitations in processing high-speed dynamic targets. Their state perception lacks precision, relying on direct physical quantities such as displacement and impact. They lack monitoring and analysis of rotational dynamic characteristics, including an inability to distinguish between stable acceleration and the early stages of unstable oscillation, which limits risk prediction. Data processing and analysis lags, often relying on post-analysis or passive responses based on fixed thresholds. They lack the ability to predict future motion trends, such as when a temperature exceeds a threshold and an alarm is triggered, the optimal opportunity for intervention is estimated to have passed. Their multi-source information fusion application is limited, focusing on parallel data recording. They lack deep spatiotemporal correlation and fusion analysis of information such as internal state and external trajectory, making it difficult to quickly integrate and identify sensor failures or actual complex motion. Their acquisition triggering mechanism is passive and isolated, making it difficult to proactively prepare for predicted complex events. For brief, unexpected events, passive triggering can miss the acquisition window due to delays. They lack intelligent collaborative acquisition optimization, and their multiple channels lack dynamic coordination based on overall situation prediction. Node parameters cannot be adjusted based on anticipated events, resulting in wasted resources and poor quality of critical data. Summary of the Invention

[0005] In order to address the shortcomings of the prior art, the present invention provides an equipment target control data acquisition system and its acquisition method. The technical solution is as follows:

[0006] In one aspect, an equipment target control data acquisition system is provided, the system comprising:

[0007] The state potential energy module obtains the angular velocity change rate and angular momentum vector value of the projectile through real-time sensors, identifies the change in the projectile's motion state and the landing point, performs difference comparison, determines the characteristic stable segment, and transmits it to the velocity projection module;

[0008] The velocity projection module obtains the real-time velocity vector of the flying projectile through the inertial navigation device, adjusts the Kalman filter gain coefficient based on the covariance matrix of the projectile's angular velocity change rate, calculates the projection angle change rate, compares the motion state changes of the characteristic stable segment, determines the projectile motion mode analysis segment, and transmits it to the trajectory fusion module;

[0009] The trajectory fusion module obtains the projectile coordinates and turning trajectory point sets through radar equipment, identifies the composite segment of the projectile coordinates' movement path, and marks the intersection segment that overlaps with the execution time of the projectile motion pattern analysis segment as a multi-node warning segment, and transmits it to the collaborative preloading module;

[0010] A collaborative preloading module receives the multi-node warning segments to generate multiple prompt instructions, and integrates the instructions into a unified scheduling instruction and outputs it to the acquisition trigger module;

[0011] The acquisition trigger module receives the unified scheduling instruction, activates each acquisition node sensor, and performs buffer initialization and data recording and storage.

[0012] As a further solution of the present invention, the feature stabilization segment includes a stable time window, a state parameter set, and a stability identifier; the projectile motion pattern analysis segment includes a projection vector sequence, angular rate data, and pattern matching results; the multi-node warning segment includes a warning time interval, a key coordinate point set, and a fusion trajectory segment; the unified scheduling instruction includes a target node list, a trigger timing arrangement, and sensor configuration parameters; and the buffer initialization includes an initialization state identifier, a recording data block, and a storage location pointer.

[0013] As a further solution of the present invention, the state potential energy module includes:

[0014] Motion data acquisition submodule: obtains the time series of the projectile's angular velocity change rate and angular momentum vector value provided by the real-time sensor to represent the basic data set of the projectile's rotational motion state in a continuous time period and obtain the instantaneous motion state quantity;

[0015] A state function generation submodule is configured to call the instantaneous motion state quantity, perform a differential operation on the angular velocity change rate time series to obtain the second-order time derivative, combine the second-order time derivative of the angular velocity, perform Z-score normalization on the second-order time derivative of the angular velocity and the angular momentum vector value, and calculate the state index value at each time point to establish a state function time series.

[0016] Stable feature discrimination submodule: Based on the state function time series, the absolute difference of the state index values ​​at adjacent time points in the series is calculated, and the calculated difference is compared with the preset state potential energy difference threshold point by point to determine the motion state change and the falling point of the projectile, and determine the characteristic stable segment;

[0017] The state potential energy difference threshold is calibrated and set by analyzing the wind tunnel test data of the missile body according to the observed fluctuation extreme value or change rate extreme value of the key parameters of angular velocity or attitude angle related to the sudden change of the flight state.

[0018] As a further solution of the present invention, the velocity projection module includes:

[0019] Velocity vector acquisition submodule: uses inertial navigation equipment to record the velocity information of the projectile at each flight moment, extracts the velocity component at each moment, and obtains the velocity vector at a specified moment based on the motion state and position data of the projectile;

[0020] The velocity projection processing submodule selects a projection direction that matches the velocity vector based on the velocity vector at the specified moment, performs a vector projection operation, maps the velocity vector to the corresponding reference axis, calculates the projected vector value at each moment, and obtains a velocity projection vector.

[0021] Kalman filter submodule: Based on the obtained velocity projection vector, by introducing prior information, the Kalman filter is used to estimate the true value of the projection vector at each moment, remove noise and error, and adjust the filtering result according to the difference between the current measurement data and the predicted data to obtain a smooth projection vector;

[0022] The measurement noise statistical characteristics in the prior information are obtained by fitting a Gaussian distribution model based on a sensor noise sampling data set collected when the missile is in a stationary state;

[0023] Motion pattern comparison submodule: Based on the smoothed projection vector, the projection angle change rate at each moment is calculated. Combined with the motion state change of the characteristic stable segment, a comparative analysis of the projection angle change rate is performed through angle calculation and time series to obtain the projectile motion pattern analysis segment.

[0024] As a further solution of the present invention, the difference between the current measured data and the predicted data is processed using the formula:

[0025] K i =P i|i-1 H T (HP i|i-1 H T +R) -1 ;

[0026] where K i is the Kalman gain, H represents the measurement matrix, H T The transpose of the measurement matrix H, P i|i-1 Represents the predicted state covariance matrix, describing the error propagation result of the state estimation in the process of predicting from time step i-1 to time step i, R represents the covariance matrix of the measurement noise, (HP i|i-1 H T + R) is the covariance of the predicted measurements.

[0027] As a further solution of the present invention, the trajectory fusion module includes:

[0028] Coordinate acquisition submodule: obtains the real-time missile body coordinate data provided by the radar equipment, extracts the coordinate points of the missile body by analyzing the signal returned by the radar and records the timestamp, unifies the coordinate data into position data in the global coordinate system, and obtains the missile body coordinate point set;

[0029] Trajectory recognition submodule: Based on the projectile coordinate point set, it identifies continuous trajectory points, analyzes the spatial changes between trajectory points, combines the time interval and path changes, and identifies and determines the composite segment of the projectile motion path through the least squares method;

[0030] Overlap judgment submodule: Based on the obtained composite segment of the projectile motion path and the projectile motion pattern analysis segment, the time interval of the composite segment of the projectile motion path and the projectile motion pattern analysis segment is calculated, the time range of the composite segment and the projectile motion pattern analysis segment is compared and calculated, and a time overlap threshold is performed to extract the overlapping time period to obtain a multi-node warning segment;

[0031] The time overlap threshold is calculated based on the standard deviation of the time distribution of the pattern analysis segment of the projectile motion data, and the window length is set according to the standard deviation to determine whether the intersection length of the time period is greater than or equal to half of the window length.

[0032] As a further solution of the present invention, the difference between the actual observed data points and the predicted values ​​of the fitting model is processed using the formula:

[0033]

[0034] Where E is the sum of squared errors, which is an indicator of the overall difference between the fitted model and the actual data, j represents the index of the data point, ranging from 1 to N, where N is the total number of data points, and z jrepresents the independent variable value of the jth data point, x(z j ) is the dependent variable value of the jth data point, corresponding to the independent variable z j , a0 is the constant term of the polynomial, a1 is the linear term in the polynomial, and a2 is the quadratic term in the polynomial The coefficient of is a quadratic polynomial model with z as the independent variable j The predicted value at time .

[0035] As a further solution of the present invention, the collaborative preloading module includes:

[0036] Node distribution generation submodule: receives the multi-node warning segments, divides the nodes into spatial distributions according to the time stamp and spatial coordinates of each node using a cluster analysis method, calculates the positions and time intervals between multiple nodes, and generates a prompt instruction set based on the node distribution;

[0037] Instruction integration submodule: Based on the obtained prompt instruction set, the time and space ranges of multiple prompt instructions are merged, redundant instructions are eliminated, and unified scheduling instructions are generated by optimizing and adjusting instructions in the same time period, and output to the acquisition trigger module.

[0038] As a further solution of the present invention, the acquisition trigger module includes:

[0039] Node activation submodule: Based on the unified scheduling instruction, according to the time and space information in the instruction, activate the sensor of each collection node, and according to the geographical location and scheduling requirements of each node, control the corresponding sensor to turn on and start data collection. The node sensor responds in time and starts working, and a list of activated collection nodes is obtained;

[0040] Buffer initialization submodule: Based on the activated collection node list, allocate an independent buffer to each activated node, initialize the memory according to the size and storage requirements of the buffer, clear the data, so that the buffer of each node can store the real-time collected data, and obtain the initialized buffer information;

[0041] Data storage submodule: Based on the initialized buffer information, the collected data is partitioned and stored according to the timestamp and node ID, a data write operation is performed and a redundancy check is performed, and the data is saved in a specified location to obtain the stored collected data.

[0042] On the other hand, a method for collecting equipment target control data is provided. The method is applied to an equipment target control data collection system, and the method includes:

[0043] S1: The angular velocity change rate and angular momentum vector values ​​are acquired through real-time sensors, and a state potential energy function sequence is constructed based on the data. The state potential energy of adjacent time periods is compared and analyzed by difference comparison to determine whether a change in motion state has occurred. This is used as a basis for identifying the falling point and determining the characteristic stable segment.

[0044] S2: The real-time velocity vector of the flying projectile is obtained through the inertial navigation device, and the velocity vector is projected. The projection vector is smoothed by the Kalman filter algorithm, and the projection angle change rate representing the second-order derivative of the angle between the velocity vector projection and the projectile axis is calculated. The motion state change of the characteristic stable segment and the projection angle change rate are compared and analyzed to generate a projectile motion mode analysis segment;

[0045] S3: Obtain the projectile coordinates and turning trajectory point set through radar equipment, combine the trajectory point and coordinate information to identify the path composite segment, and compare it with the time overlap range of the projectile motion pattern analysis segment to determine whether the time overlap threshold is met. If so, it is determined to be a multi-node warning segment;

[0046] S4: receiving the multi-node warning segment, generating multiple prompt instructions according to the node distribution, and integrating the instructions into a unified scheduling instruction;

[0047] S5: By receiving the unified scheduling instruction, each collection node sensor is activated, a buffer is initialized, and the real-time collected data is recorded and stored to obtain stored collected data.

[0048] The technical solutions provided by the embodiments of the present invention bring at least the following beneficial effects: by acquiring the angular velocity change rate and angular momentum vector values ​​of the projectile, constructing a state potential function sequence analysis, the system can accurately identify motion state changes and landing points, and performing difference comparison to determine the characteristic stable segment, thereby improving the ability to characterize complex motion postures and stability and enhancing the capture of stable features under interference; using inertial navigation to obtain the velocity vector, and through velocity projection and Kalman filtering smoothing, calculating the projection angle change rate, comparing it with the state change of the characteristic stable segment, accurately locking the motion pattern analysis segment, improving the quality of velocity data and interference resistance, and focusing on the key interval; fusing radar coordinates with the turning trajectory point set to identify the composite segment of the moving path, performing time overlap judgment with the motion pattern analysis segment, determining the multi-node warning segment, and combining internal state and external trajectory to enhance warning accuracy; generating prompt instructions based on the multi-node warning segment and node distribution, integrating them into a unified scheduling instruction output, realizing prediction-based collaborative preloading, and improving the pertinence and timeliness of acquisition tasks; the acquisition triggering link activates the sensor according to the unified scheduling instruction, performs buffer initialization and data recording and storage, ensuring accurate and loss-free collaborative acquisition during the critical warning period, and improving the integrity and effectiveness of data acquisition. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 is a system flow chart of the present invention;

[0051] Figure 2 is a system block diagram of the present invention;

[0052] Figure 3 Schematic diagram of the steps of the method of the present invention. DETAILED DESCRIPTION

[0053] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0054] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as an "exemplary" in the present invention should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of the word "exemplary" is intended to present concepts in a concrete manner. Furthermore, in the embodiments of the present invention, "and / or" can mean both or either of the two.

[0055] In the embodiments of the present invention, the terms "image" and "picture" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same. The terms "of," "corresponding," and "corresponding" may be used interchangeably. It should be noted that, when the distinction between them is not emphasized, their intended meanings are the same.

[0056] In the embodiments of the present invention, sometimes a subscript such as W1 may be written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are the same.

[0057] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0058] The embodiment of the present invention provides an equipment target control data acquisition system, please refer to Figure 1 The present invention provides a technical solution, an equipment target control data acquisition system comprising:

[0059] The state potential energy module obtains the angular velocity change rate and angular momentum vector value of the projectile through real-time sensors, identifies the change in the projectile's motion state and the landing point, performs difference comparison, determines the characteristic stable segment, and transmits it to the velocity projection module;

[0060] The velocity projection module obtains the real-time velocity vector of the flying projectile through the inertial navigation device, adjusts the Kalman filter gain coefficient based on the covariance matrix of the projectile's angular velocity change rate, calculates the projection angle change rate, compares the motion state changes in the characteristic stable segment, determines the projectile motion mode analysis segment, and transmits it to the trajectory fusion module;

[0061] The trajectory fusion module obtains the projectile coordinates and turning trajectory point sets through radar equipment, identifies the composite segment of the projectile coordinates' moving path, and marks the intersection segment that overlaps with the execution time of the projectile motion pattern analysis segment as a multi-node warning segment, and passes it to the collaborative preloading module;

[0062] The collaborative preloading module receives multiple node warning segments to generate multiple prompt instructions, and integrates the instructions into a unified scheduling instruction and outputs it to the acquisition trigger module;

[0063] The acquisition trigger module receives unified scheduling instructions, activates the sensor of each acquisition node, and performs buffer initialization and data recording and storage.

[0064] The feature stability segment includes the stable time window, state parameter set, and stability identification; the projectile motion mode analysis segment includes the projection vector sequence, angular rate data, and pattern matching results; the multi-node warning segment includes the warning time interval, key coordinate point set, and fusion trajectory fragment; the unified scheduling instruction includes the target node list, trigger timing arrangement, and sensor configuration parameters; the buffer initialization includes the initialization state identification, recording data block, and storage location pointer.

[0065] See also Figure 2 , the state potential energy module includes:

[0066] Motion data acquisition submodule: obtains the time series of the projectile's angular velocity change rate and angular momentum vector value provided by the real-time sensor to represent the basic data set of the projectile's rotational motion state in a continuous time period and obtain the instantaneous motion state quantity;

[0067] The time series of the projectile angular velocity change rate and the angular momentum vector value provided by the real-time sensor are obtained. The execution process is as follows: first, the data acquisition system is initialized and the sampling frequency is set to 100 Hz, that is, data is collected every 0.01 seconds. Then, the gyroscope sensor group and angular momentum measurement unit installed inside the projectile or at key locations are started. The sensors monitor the angular velocity of the projectile around three mutually orthogonal axes (defined as the roll axis, pitch axis, and yaw axis) in real time. x ,ωy ,ω z ] (unit: rad / s) and the angular momentum vector (Unit: kg·m 2 / s), the system continuously reads the analog or digital signal output by the sensor, performs time difference processing on the read angular velocity signal ω(t), and uses first-order backward difference to calculate the angular velocity change rate At time t i The rate of change of angular velocity, ω(t i ) at time t i The collected angular velocity, ω(t i-1 ) at time t i-1 The collected angular velocity, t i is the i-th sampling moment, Δt is the sampling time interval, i.e. 0.01 seconds, i represents the i-th sampling moment, and the corresponding moment t is recorded synchronously i The angular momentum vector The three component values ​​of , at t = 1.00 seconds, the collected angular velocity is ω(1.00s) = [2.0, 0.5, -1.0] rad / s, at t = 1.01 seconds, the collected angular velocity is ω(1.01s) = [2.1, 0.4, -1.2] rad / s, then the angular velocity change rate at t = 1.01 seconds is calculated At the same time, the angular momentum vector value recorded at t = 1.01 seconds is Each sampling time point t i Obtained angular velocity change rate vector and the angular momentum vector Together with the timestamp t i The data are stored together to form a time series data stream. A data structure or table including time, angular velocity change rate X component, angular velocity change rate Y component, angular velocity change rate Z component, angular momentum X component, angular momentum Y component, and angular momentum Z component can be established. The continuous collection and calculation process covers the entire flight period of the projectile that needs to be analyzed, including from the 5th second to the 25th second after launch, forming a data set including (25-5)×100=2000 data points. The following Table 1 shows a small piece of sample data.

[0068] Table 1 Sampling data fragments of missile motion sensor

[0069]

[0070] As shown in Table 1, the angular velocity change rate and angular momentum data collected and calculated from t = 5.00 seconds to t = 5.04 seconds are a complete record of the dynamic information of the projectile's rotational motion within the specified time period. The set of data points arranged in chronological order generates a basic data set that represents the projectile's rotational motion state within the continuous time period, and the instantaneous motion state quantity is obtained.

[0071] State function generation submodule: Calls the instantaneous motion state quantity, performs differential operation on the angular velocity change rate time series to obtain the second-order time derivative, combines the second-order time derivative of angular velocity, performs Z-score normalization on the second-order time derivative of angular velocity and the angular momentum vector value, and calculates the state index value at each time point to establish the state function time series;

[0072] Calling the instantaneous motion state is to call the angular velocity change rate time series recorded in the basic data set generated by the previous submodule and the angular momentum vector time series Time series of angular velocity change rate Perform a difference operation to obtain the second-order time derivative The operation uses the numerical difference method, second-order backward difference Represents the time t i The second-order time derivative of the angular velocity, ω(t i ) represents the time t i The collected angular velocity, ω(t i-1 ) represents the time t i-1 The collected angular velocity, ω(t i-2 ) represents the time t i-2 The collected angular velocity, Δt represents the sampling time interval, t i-1 , t i-2 Represents the discrete sampling moment. Taking the second-order backward difference as an example, take Δt = 0.01 seconds, and use the data in Table 1 to calculate the second-order time derivative of the angular velocity at t = 5.02 seconds. It is necessary to use t = 5.02s, t = 5.01s, and t = 5.00s. value: [5.0,-2.1,8.3], then Then calculate the square value of the second-order time derivative vector of angular velocity, that is, the square of the modulus For the example of t=5.02s, Then, get the current time point t i The angular momentum vector And calculate the square value of the size Indicates that at the current time point t i When the angular momentum vector L(ti ) square value, L(t i ) represents the discrete time point t i The instantaneous angular momentum vector, L x (t i ) represents the discrete time point t i When the angular momentum vector L(t i ) on the x-axis, L y (t i ) represents the discrete time point t i When the angular momentum vector L(t i ) on the y-axis, L z (t i ) represents the discrete time point t i When the angular momentum vector L(t i ) component on the z-axis, t i Represents a discrete time point or the i-th sampling moment in the time series. For the example of t = 5.02s, we can take but Sum the square value of the second-order time derivative of angular velocity and the square value of the current angular momentum vector to calculate the state index value at each time point For the example with t = 5.02s, S(5.02s) = 27 × 10 6 +533.34=27000533.34, repeat the above difference, square, and sum calculation process for all time points in the basic data set (the difference calculation requires the previous and next data points, and the few starting and ending points cannot be calculated). The obtained state index value S(t i ) are arranged in chronological order to establish a state function time series.

[0073] Stable feature discrimination submodule: Based on the state function time series, the absolute difference of the state index values ​​at adjacent time points in the sequence is calculated, and the calculated difference is compared with the preset state potential energy difference threshold point by point to determine the motion state change and landing point of the projectile, and determine the characteristic stable segment;

[0074] The state potential energy difference threshold is calibrated and set by analyzing the missile wind tunnel test data and based on the extreme fluctuation or rate of change of the key parameters of angular velocity or attitude angle related to the sudden change of flight state.

[0075] Based on the state function time series S(t i ), that is, the state indicator value sequence covering the observation time period calculated by the previous submodule, including the sequence S=[S(t1),S(t2),...,S(t N )], calculate the absolute difference ΔS(ti )=|S(t i+1 )-S(t i )|, the calculation is performed for all adjacent point pairs in the sequence from i 1 to N-1, including, if S(5.02s)=27000533.34 has been calculated, and S(5.03s)=28500700.00 is calculated, then the absolute difference is ΔS(5.02s)=|S(5.03s)-S(5.02s)|=|28500700.00-27000533.34|=1500166.66, the calculated difference ΔS(t i ) and the preset state function difference threshold ΔS th Compare point by point, the state function difference threshold ΔS th The setting of ΔS(t) needs to be based on the flight test data or simulation data of the missile body, and the degree of change of the state function S(t) of the missile body in the known stable flight stage and the stage of maneuvering and changing trajectory or being disturbed should be analyzed. The maximum value and fluctuation range of ΔS(t) in the stable stage and the typical value of ΔS(t) in the unstable stage should be counted, and a value that can distinguish the two states should be selected as the threshold. Including, by analyzing the data, ΔS(t) does not exceed 1.0×10 6 , and when the attitude is adjusted or disturbed, ΔS(t) will jump to 5.0×10 6 In order to identify significant changes while filtering out normal fluctuations during stable flight, the threshold can be set between the two, and can be 1.5 times the maximum fluctuation value during the stable phase. If the maximum ΔS observed during the stable phase is 0.8×10 6 , then set ΔS th =1.5×0.8×10 6 =1.2×10 6 Now compare the calculated difference with the threshold. For the example, ΔS(5.02s)=1500166.66, since 1500166.66>1.2×10 6 , it is determined that a change in motion state occurs between t = 5.02s and t = 5.03s, and point t k The difference ΔS(t k )>ΔS th , then mark t k The point near the moment is the point where the motion state changes. Continue to check the sequence backward. If at t m At the moment (m>k), the difference falls back to ΔS(t m )≤ΔS th , then t m The vicinity can be regarded as a falling point after a state change. By comparing the state change and the falling point of the projectile point by point, ΔS(t i )≤ΔSth Time interval identification, including time series t a ,t a+1 ,...,t b All corresponding differences ΔS(t a ),ΔS(t a+1 ),...,ΔS(t b-1 ) are less than or equal to ΔS th =1.2×10 6 , then the time period [t a ,t b ] is determined as a characteristic stable segment, and the characteristic stable segment is determined.

[0076] See also Figure 2 , the velocity projection module includes:

[0077] Velocity vector acquisition submodule: uses inertial navigation equipment to record the velocity information of the projectile at each flight moment, extracts the velocity component at each moment, and obtains the velocity vector at a specified moment based on the motion state and position data of the projectile;

[0078] The system uses the inertial navigation device (INS) to record the velocity information of the missile at each moment of flight to execute the action. The system is connected to the inertial navigation unit installed on the missile. The unit has a built-in gyroscope and accelerometer. The velocity information of the missile in the predetermined coordinate system (including the launch point ground-fixed coordinate system or the missile body coordinate system) is output in real time through integration operation. At each preset sampling time point t i (The sampling frequency is consistent with the above data acquisition, including 100 Hz, Δt = 0.01 seconds), the speed measurement value is read from the INS data stream, and the measurement value is given in the form of three orthogonal components, including t i The speed at the moment is where v x (t i ),v y (t i ),v z (t i ) are the velocity components along the X, Y, and Z axes of the coordinate system, in meters per second (m / s), including the velocity component v output by the inertial navigation device at time t = 15.20 seconds. x (15.20s) = 850.5m / s, v y (15.20s) = 210.8m / s, v z (15.20s) = -75.3m / s, then extract these three component values, and according to the current motion state of the projectile (including whether it is in the boost or gliding phase, which affects the selection or conversion of the coordinate system) and position data (also provided by INS for coordinate conversion or reference), combine these three velocity components to construct the velocity vector at the moment, that is, For the reference coordinate system used for internal calculations of the INS, if the velocity in the target reference system is required, it is necessary to perform coordinate transformation operations in combination with the attitude data provided by the INS (including Euler angles or quaternions). The current value is set to the velocity in the desired reference system, and this extraction and combination process is repeated to cover the entire time period required for analysis to obtain the velocity vector at the specified time.

[0079] Velocity projection processing submodule: Based on the velocity vector at a specified moment, select the projection direction that matches the velocity vector, perform the vector projection operation, map the velocity vector to the corresponding reference axis, calculate the vector value after projection at each moment, and obtain the velocity projection vector;

[0080] Velocity vector based on a specified time First, you need to select one or more physically meaningful projection directions as reference axes. The selection is based on the analysis requirements, including the longitudinal axis of the projectile (reflecting the velocity component along the direction of the projectile), the tangent direction of the trajectory, or the normal direction perpendicular to the trajectory. Set the longitudinal axis of the projectile itself. As the projection direction, the direction information of the axis (unit vector) is calculated by the attitude measurement system (including part of the INS or an independent attitude sensor) at the same time t i Provided, perform vector projection operation, is to transform the velocity vector Towards the selected reference axis Projection is performed and the projection operation is completed by vector dot product (inner product): First, the scalar projection of the velocity vector in the direction of the reference axis is calculated Represents the discrete time point t i The instantaneous velocity vector, Represents the discrete time point t i The unit vector in the direction of the reference axis, v x (t i ) represents the discrete time point t i When the velocity vector The component on the x-axis, a x (t i ) represents the discrete time point t i When the reference axis unit vector The component on the x-axis, v y (t i ) represents the discrete time point t i When the velocity vector The component on the y-axis, a y (t i ) represents the discrete time point t i When the reference axis unit vector The component on the y-axis, v z (t i ) represents the discrete time point t i When the velocity vector The component on the z-axis, a z (t i ) represents the discrete time point t i When the reference axis unit vector The component on the z-axis, t i Represents a discrete time point or the i-th sampling moment in the time series, and then multiplies the scalar projection value by the unit vector in the reference axis direction to obtain the projected vector value, that is, the velocity projection vector Taking t = 15.20 seconds as an example, the velocity vector obtained is Set the unit vector of the projectile's longitudinal axis obtained at the same time to (This is a unit vector, 0.995 2 +0.087 2 +(-0.052) 2 ≈1), then calculate the scalar projection v proj_scalar (15.20s) = (850.5)(0.995) + (210.8)(0.087) + (-75.3)(-0.052) = 846.2475 + 18.3396 + 3.9156 = 868.5027 m / s, then calculate the projection vector For each specified time t in the time series i The operations of selecting the projection direction, calculating the dot product, and multiplying by the unit vector are repeated to obtain the velocity projection vector.

[0081] Kalman filter submodule: Based on the obtained velocity projection vector, by introducing prior information, the Kalman filter is used to estimate the true value of the projection vector at each moment, remove noise and error, and adjust the filtering result according to the difference between the current measurement data and the predicted data to obtain a smooth projection vector;

[0082] The statistical characteristics of the measurement noise in the prior information are obtained by fitting a Gaussian distribution model based on the sensor noise sampling data set collected when the missile is in a stationary state.

[0083] Project the vector time series according to the obtained velocity is the directly calculated measurement value, i is the velocity measurement error of INS at the i-th sampling moment in the sequence and the projection error introduced by the attitude measurement error. It is necessary to introduce prior information and use the Kalman filter for processing. First, define the state vector x of the filter. i, including the projection vector itself and the estimated rate of change of velocity, including the assumption State transition model x i =Fx i-1 +w i-1 Describes the state from t i-1 Time evolves to t i At this moment, F is the state transfer matrix, w i-1 is the process noise, the measurement model z i =Hx i +v i Describes the measured value (i.e., the directly calculated Relationship with the real state, H is the measurement matrix (if the state is directly the projection vector, then H is the identity matrix I), v i is the measurement noise. The prior information is: the statistical characteristics of the process noise, that is, the covariance matrix Q, which characterizes the uncertainty of the state transfer model, including the maneuver acceleration range setting based on the projectile estimation, and setting the standard deviation of the projectile acceleration in the projection direction to σ a =5m / s 2 , then the diagonal elements of Q can be set to (σ a Δt) 2 =(5×0.01) 2 =0.0025m 2 / s 2 , the statistical characteristics of the measurement noise, namely the covariance matrix R, characterizes the directly calculated projection vector The uncertainty of the INS velocity error (including σ v_ins =0.1m / s) and attitude error (including σ θ =0.05°≈0.00087rad) is obtained by evaluating the projection error, including the diagonal elements of R, which are set to / s 2 The dynamic constraints are reflected in the construction of the state transfer matrix F, including setting the projection velocity to change approximately uniformly in a short period of time, then F = I, and the time synchronization parameter is used to calculate INS data and used to calculate The attitude data timestamps are strictly aligned, and the parameters (Q, R, F) need to be obtained through actual measurement and calibration, including collecting sensor outputs under known inputs through flight tests or simulations, and analyzing the error statistical characteristics to set Q and R. The execution process of the Kalman filter is at each time step t i Perform two steps of prediction and update: the prediction step is based on t i-1 Optimal estimate of time and the state transition model predicts t i State of the moment and its covariance Pi|i-1 =FP i-1|i-1 F T +Q, the update step uses t i Measurement value at time Correct the prediction results and calculate the Kalman gain K i =P i|i-1 H T (HP i|i-1 H T +R) -1 , where K i is the Kalman gain, H represents the measurement matrix, H T The transpose of the measurement matrix H, P i|i-1 Represents the predicted state covariance matrix, describing the error propagation result of the state estimation in the process of predicting from time step i-1 to time step i, R represents the covariance matrix of the measurement noise, (HP i|i-1 H T +R) is the covariance of the predicted measurement value, and the predicted state covariance matrix P i|i-1 After the prediction step, we obtain the covariance of the prediction errors as The units here are (m / s)2 and (m / s 2 ) 2 , representing the variance of the predicted velocity and the variance of the predicted acceleration, respectively, calculate H T (Transpose of H): Calculate P i|i-1 H T : Calculating HP k|k-1 H T , which represents the prediction uncertainty mapped to the measurement space:

[0084] HP i|i-1 H T +R=[0.25]+[0.1]=[0.35], (HP i|i-1 H T +R) -1 =[0.35] -1 ≈[2.85714], finally calculate the Kalman gain Measured value z k is the directly measured projection velocity v k , that is, z k =[v k,measured ]. Therefore, the measurement matrix H, which maps the state vector to the measurement space, is H = [1 0], and then the optimal estimate and the updated covariance process at the current moment are calculated according to the current measurement data z i With forecast data The difference is calculated by Kalman gain K i Dynamically adjust the filtering results for all Perform a filtering operation to obtain a smoothed projection vector.

[0085] Motion pattern comparison submodule: Based on the smoothed projection vector, the projection angle change rate at each moment is calculated. Combined with the motion state changes in the characteristic stable segment, the projection angle change rate is compared and analyzed through angle calculation and time series to obtain the projectile motion pattern analysis segment.

[0086] Smoothed projection vector time series This is the result of the Kalman filter module output. First, calculate the rate of change of the projection angle at each moment. It is necessary to define the projection angle θ(t i ), which is the smoothed projection vector With a fixed reference direction The angle between the reference direction Can be the projection vector direction at the initial moment Or the vertical upward direction of the geodetic coordinate system choose Then the angle θ(t i ) is calculated by dot product: Then The unit is radians or degrees, including, if t = 20.00 seconds, the filtered projection vector is Then it and The dot product is -80.5, and its modulus is Therefore, cos(θ(20.00s))=-80.5 / 769.3≈-0.1046, and we get θ(20.00s)≈arccos(-0.1046)≈1.675rad≈96.0°. Next, we calculate the rate of change of the projection angle. First-order backward difference t i-1 Represents the current time point t i The previous discrete time point, including, if the calculated θ(19.99s)≈95.8°, then Then, combined with the feature stability segment information determined in the stable feature discrimination submodule, including the time period [t a ,t b ] is the stable segment, and the identified motion state change points (including t k Time ΔS(t k )>ΔS th ), execute the projection angle change rate The comparative analysis is to calculate the time period of the characteristic stability period (including t i ∈[t a ,t b ]) and unstable segments (including state change points t k nearby time window) The statistical characteristics of The average absolute value of and the maximum absolute value And compared with the corresponding statistical values ​​of the unstable section, including, if the average absolute angular rate calculated in the stable section [10.0s, 14.5s] is 8° / s and the maximum absolute angular rate is 15° / s, while in the state change area identified around t=15.2s, the average absolute angular rate calculated is 45° / s and the maximum absolute angular rate reaches 70° / s. This comparison is obtained by angle calculation (obtaining θ(t i )) and time series analysis (calculation and segmented statistics) to determine the degree of difference in the rate of change of the projection angle under multiple motion states, including the ability to set a change rate distinction threshold include If the average of a single section It is considered to be a stable mode. If the peak value significantly exceeds this value, it is considered to be a maneuvering or disturbance mode. The comparative analysis results are associated with the corresponding time period to obtain the projectile motion mode analysis segment.

[0087] See also Figure 2 , the trajectory fusion module includes:

[0088] Coordinate acquisition submodule: obtains the real-time missile body coordinate data provided by the radar equipment, extracts the coordinate points of the missile body by analyzing the signal returned by the radar and records the timestamp, unifies the coordinate data into position data in the global coordinate system, and obtains the missile body coordinate point set;

[0089] Obtain the real-time missile coordinate data provided by the radar equipment. The execution is to start the ground or airborne radar system to scan the target airspace. When the radar beam detects the missile target, it receives the reflected echo signal. The signal processing unit analyzes the echo signal, extracts the time delay information, and calculates the slant range R between the radar station and the missile. At the same time, it measures the antenna beam pointing when the signal returns, and obtains the azimuth angle φ and pitch angle ∈, together with the measurement timestamp t iThe data recorded together include: at 14:35:22.100 UTC (corresponding to flight time t = 30.10 seconds), the data measured by a single ground radar station are: slant range R = 65200 meters, azimuth φ = 125.2°, pitch angle ∈ = 25.8°, the system reads the raw measurement data (R, φ, ∈) and the timestamp t i , based on the radar station's own spherical coordinate system data, it is converted to the specified global coordinate system, including the WGS-84 Earth Centered Fixed Coordinate System (ECEF) or the Northeastern Universe (ENU) coordinate system with the launch point as the origin. The conversion requires the global coordinates of the radar station (x r ,y r ,z r ) and the local coordinate axis direction, the coordinate rotation and translation operations are converted, including conversion to the ENU coordinate system, setting the radar station at the origin of the ENU coordinate system, then x(t i )=Rcos(∈)sin(φ),y(t i )=Rcos(∈)cos(φ),z(t i )=Rsin(∈), substitute the example data: x(30.10s)=65200×cos(25.8°)×sin(125.2°)≈65200×0.899×0.817≈47950 meters, y(30.10s)=65200×cos(25.8°)×cos(125.2°)≈65200×0.899×(-0.576)≈-33830 meters, z(30.10s)=65200×sin(25.8°)≈65200×0.435≈28360 meters, and calculate the global coordinate point P(t i )=[x(t i ),y(t i ),z(t i )] together with the timestamp t i Store together, continue radar tracking and data processing, obtain and record coordinate points at fixed time intervals (including radar data update rate estimated to be 1Hz, set to 10Hz, i.e. Δt = 0.1 seconds), as shown in Table 2.

[0090] Table 2 Radar measurement of missile body coordinate data fragments

[0091] time X coordinate (meters) Y coordinate (meters) Z coordinate (meters) 30.00 47100 -33100 27800 30.10 47950 -33830 28360 30.20 48810 -34570 28920 30.30 49680 -35320 29490 30.40 50560 -36080 30060

[0092] As shown in Table 2, the partially converted global coordinate data of the projectile from t=30.00 seconds to t=30.40 seconds are listed, and the coordinate point set arranged in time sequence is obtained to obtain the projectile coordinate point set.

[0093] Trajectory recognition submodule: Based on the projectile coordinate point set, it identifies continuous trajectory points, analyzes the spatial changes between trajectory points, combines the time interval and path changes, and uses the least squares method to identify and determine the composite segment of the projectile motion path;

[0094] Based on the projectile coordinate point set, which is the time series data output by the coordinate acquisition submodule, the continuous trajectory points are first identified, that is, a data point sequence within a time window is selected from the point set, including the data points from t j to t k All points of the trajectory are analyzed, the spatial changes between the trajectory points are analyzed, the displacement vectors and velocity estimates between adjacent points are calculated, the magnitude and direction change trends of the velocity vectors are monitored, and the time interval Δt and the path change (including determining the curvature change of the trajectory by calculating the angle change between consecutive displacement vectors) are combined to determine whether the projectile motion path is an approximate straight line segment or a curved segment, or a composite segment composed of multiple segments and multiple characteristic paths. For example, if the direction change of the displacement vectors for multiple consecutive time steps is very small (including the angle between adjacent displacement vectors is less than 1°), it is a straight line motion. If the direction change is continuous and significant (including the angle is continuously greater than 5°), it is a curved motion. Based on this analysis, the entire trajectory is divided into several segments, and the path determination operation is performed on each segment (or point in the entire time window). The trajectory model is fitted using the least squares method, and a quadratic polynomial model is set to fit the change of the X coordinate over time x(t)≈a0+a1t+a2t 2 , select the time window [t j ,t k ] within N=(kj)+1 coordinate points (t l ,x(t l )), the goal is to make the error sum squared by coefficients a0, a1, a2 Minimum, where E is the sum of squared errors, which is an indicator of the overall difference between the fitted model and the actual data, j represents the index of the data point, ranging from 1 to N, where N is the total number of data points, and z j represents the independent variable value of the jth data point, x(z j ) is the dependent variable value of the jth data point, corresponding to the independent variable z j , a0 is the constant term of the polynomial, a1 is the linear term in the polynomial, and a2 is the quadratic term in the polynomial The coefficient of is a quadratic polynomial model with z as the independent variable j The predicted value at time , by taking the partial derivatives of E with respect to a0, a1, a2 and setting them equal to zero, we get a linear equation system, and solving the equation system can get the optimal coefficients Similar operations are performed on the Y and Z coordinates to obtain the fitted path function Each component is a polynomial or selected function form with respect to time t, including fitting a quadratic polynomial to the five data points from t = 30.00s to t = 30.40s in Table 2. The function segment obtained by fitting represents a certain section of the projectile motion path within the corresponding time interval. If the entire trajectory is divided into multiple parts for fitting, the resulting composite section of the projectile motion path is composed of multiple fitting function segments.

[0095] Overlap judgment submodule: Based on the acquired composite segment of the projectile motion path and the projectile motion pattern analysis segment, the time interval of the composite segment of the projectile motion path and the projectile motion pattern analysis segment is calculated, the time range of the two is compared and calculated, and the time overlap threshold is determined. The overlapping time period is extracted to obtain the multi-node warning segment;

[0096] The time overlap threshold is calculated based on the standard deviation of the time distribution of the pattern analysis segment of the projectile motion data, and the window length is set according to the standard deviation to determine whether the intersection length of the time period is greater than or equal to half of the window length;

[0097] The composite segment based on the projectile motion path, that is, the time interval output by the trajectory recognition submodule [t start_path,m ,t end_path,m ] and the corresponding fitting path function Call the motion pattern analysis segment obtained previously in the motion pattern comparison submodule, and the analysis segment also corresponds to the time interval [t start_mode,n ,t end_mode,n ], and each interval is associated with a motion mode (including stable flight mode, maneuvering mode, disturbed mode, etc.). By calculating the time interval of the path composite segment and the motion mode analysis segment, the operation is to traverse all path segments m and all mode segments n. For each pair (m, n), the extracted time interval I path,m =[t start_path,m ,t end_path,m ] and I mode,n =[t start_mode,n ,t end_mode,n ], calculate the intersection of these two time intervals I overlap,mn =I path,m ∩I mode,n =[max(t start_path,m ,t start_mode,n ),min(t end_path,m ,t end_mode,n )], compare and calculate the time range of the two, that is, calculate the duration of the intersection interval Δt overlap,mn =max(0,min(t end_path,m ,t end_mode,n )-max(t start_path,m ,t start_mode,n)), and then perform time overlap threshold determination, it is necessary to pre-set a time overlap threshold T overlap_th The threshold is set based on the minimum time required for the missile to perform a typical tactical maneuver or experience a disturbance, as well as the duration characteristics of the stable flight mode. For example, if the typical duration of a maneuver of the missile is about 1 second, and the minimum analysis time for radar track segment fitting and pattern recognition is 0.5 seconds, the threshold can be set to T overlap_th = 0.8 seconds, which is greater than the pseudo overlap caused by noise or short-term fluctuation estimation, and can capture the real correlation with a certain persistence. The calculated overlap duration Δt overlap,mn With threshold T overlap_th Comparison, including Δt overlap,mn ≥T overlap_th , then it is determined that there is a significant time overlap between the path segment m and the pattern segment n, and the overlapping time period I is extracted. overlap,mn , including, if the time interval of identifying a path composite segment is [28.5s, 32.0s], and the motion pattern analysis segment identifies the maneuvering mode occurring in [30.5s, 31.8s], then the intersection is [max(28.5, 30.5), min(32.0, 31.8]) = [30.5s, 31.8s], and the overlap duration Δt overlap =31.8-30.5=1.3 seconds, since 1.3s≥T overlap_th =0.8s, then extract the time period [30.5s, 31.8s] and collect all overlapping time periods that meet the overlapping threshold condition I overlap,mn , and obtain the multi-node warning segment.

[0098] See also Figure 2 , collaborative preloading modules include:

[0099] Node distribution generation submodule: Receives multiple node warning segments, divides the nodes into spatial distributions based on the time stamp and spatial coordinates of each node using cluster analysis, calculates the positions and time intervals between multiple nodes, and generates a prompt instruction set based on the node distribution.

[0100] Receive multi-node warning segments. Warning segments are time intervals determined in the previous step where the projectile motion path and the target motion mode overlap significantly in time. Two warning segments are received: segment 1 is [30.5s, 31.8s], with associated path segment m = 1 and maneuver mode n = 1; segment 2 is [45.2s, 46.5s], with associated path segment m = 2 and disturbed mode n = 2. First, extract the corresponding spatial coordinate nodes according to the time stamp of each warning segment, and use the fitting path function output by the trajectory recognition submodule to obtain the corresponding spatial coordinate nodes. In the example, sampling is performed according to the time range of the warning segment and the preset time resolution (including δt = 0.1s) to obtain spatial coordinate points with timestamps. For segment 1 [30.5s, 31.8s], the node sequence is obtained by sampling. For segment 2 [45.2s, 46.5s], sampling is obtained Set the node coordinates obtained by calculation, including N 1,1 =(30.5s,[52300,-37600,31200]), N 1,2 =(30.6s,[53150,-38350,31750])…N 2,1 =(45.2s,[85400,-62100,48500]), etc. Then cluster analysis method is used to analyze all the extracted nodes {N j,k}For spatial distribution division, the density-based spatial clustering algorithm (DBSCAN) is selected, and two key parameters are set: neighborhood radius ∈ and minimum number of neighborhood points MinPts. The setting of ∈ refers to the radar positioning accuracy and the typical spatial scale of the missile during maneuvering, including setting ∈=1000 meters, indicating that nodes with a spatial distance less than 1000 meters are considered to be adjacent. The setting basis of MinPts is the minimum number of nodes required to form a meaningful cluster, including setting MinPts=3, indicating that a node’s ∈ neighborhood includes at least 3 nodes (including itself) to be considered as a core point. The algorithm execution process is to traverse all nodes and calculate the N distance between any two nodes. j,k and N p,q Euclidean distance between If d≤∈, then they are neighboring points. Find all core points. Starting from any core point, expand through the neighborhood relationship and divide all density-reachable nodes (core points or boundary points) into the same cluster. Points that cannot be classified into any cluster are marked as noise points. Set through cluster analysis, all nodes in segment 1 form cluster C1, and all nodes in segment 2 form cluster C2. Then calculate the position and time interval between multiple nodes. For nodes in the cluster, calculate the centroid coordinates of the cluster and the time span of the cluster [t start,C ,t end,C ], for multiple clusters C1, C2, calculate the distance between the centroids and time interval (including the difference between the end time and the start time start,C2 -t end,C1), including the calculation of the centroid of cluster C1 is approximately [56500, -41000, 33500], the time span is [30.5s, 31.8s], the centroid of cluster C2 is approximately [89000, -65000, 50000], the time span is [45.2s, 46.5s], the centroid distance is approximately 45 kilometers, and the time interval is 45.2-31.8=13.4 seconds. According to the distribution of nodes (cluster location, size, density, time span and the relationship between clusters), a prompt instruction set is generated. Rule example: If the number of nodes in a cluster C exceeds N dense_th = 10 (high density threshold, refer to the sampling rate and time span setting, (31.8-30.5) / 0.1+1=14, so C1 is high density), then generate a "focus" instruction, if the time span of the cluster ΔT C =t end,C -t start,C Greater than T duration_th = 1.0 seconds (duration threshold, refer to the typical maneuver duration), then generate a "continuous tracking" instruction. If the time interval ΔT between the two clusters C1 and C2 is C1,C2 Less than T interval_th =5.0 seconds (associated time threshold, refer to tactical response time), a "sequence warning" instruction is generated. According to the above rules and calculation results, cluster C1 (number of nodes 14>10, time span 1.3s>1.0s) generates the instruction: "Instruction 1: Focus on, continue tracking, area [approximate coordinates of cluster C1 center of mass], time [30.5s, 31.8s]", and cluster C2 (set number of nodes 14>10, time span 1.3s>1.0s) generates the instruction: "Instruction 2: Focus on, continue tracking, area [approximate coordinates of cluster C2 center of mass], time [45.2s, 46.5s]". Since the inter-cluster time interval is 13.4s>5.0s, no sequence warning instruction is generated, and a prompt instruction set is formed.

[0101] Instruction integration submodule: Based on the acquired prompt instruction set, it merges the time and space ranges of multiple prompt instructions, eliminates redundant instructions, optimizes and adjusts instructions in the same time period, generates unified scheduling instructions, and outputs them to the acquisition trigger module;

[0102] The prompt instruction set obtained by the node distribution generation submodule includes instruction 1: "focus on, continue tracking, area A, time [30.5s, 31.8s]" and instruction 2: "focus on, continue tracking, area B, time [45.2s, 46.5s]", as well as other existing instructions, including instruction 3: "general attention, area A, time [31.5s, 32.5s]" generated due to the partial overlap of warning segments generated by multiple analysis paths. Area A partially overlaps with area A in space (including the centroid distance less than the preset spatial merging threshold Dmerge_th =2 km), firstly, the time and space ranges of multiple prompt instructions are combined and checked, and any two instructions I are compared. i and I j The time interval [t start,i ,t end,i ] and [t start,j ,t end,j ] and the spatial region R i With R j , calculate the time overlap Δt overlap =max(0,min(t end,i ,t end,j )-max(t start,i ,t start,j )) and spatial overlap (including determining the distance d between the centroid of the region centroids ), if the time overlap Δt overlap >0 and the spatial overlap meets the conditions (including d centroids <D merge_th ), then merge, compare instruction 1 and instruction 3, the time overlap interval is [31.5s, 31.8s], the duration 0.3 seconds is greater than 0, set the distance between area A and A center of mass to be 1.5 kilometers less than 2 kilometers, then meet the merging conditions, the time range after merging is [min(t start,1 ,t start,3 ),max(t end,1 ,t end,3)]=[min(30.5,31.5),max(31.8,32.5)]=[30.5s,32.5s], the merged spatial region is R1∪R3 (expressed by a larger region covering both or the updated centroid, including region A”), the merged instruction priority takes the higher of the two (focus > general attention), and the merged instruction is obtained: "Instruction 1: focus on, continue tracking, region A, time [30.5s,32.5s]", next, eliminate redundant instructions, in the merging process, if a single instruction is completely included in the time and space range of another instruction, and its priority is not higher than that of the instruction that includes it, it will be eliminated, including, if there is instruction 4: "focus on, region A, time [31.8,32.5s]" 0s, 31.5s]", it will be completely covered by the merged instruction 1, and its priority is not higher than instruction 1, then instruction 4 is eliminated, and the instructions in the same or adjacent time periods and spaces are optimized and adjusted, including adjusting the start and end time of the instructions to align with the standard time grid, or fine-tuning the boundaries of the spatial area, and the logical consistency between the instructions, including, the adjusted instruction set: "Instruction A: focus on, continuous tracking, area A, time [30.5s, 32.5s]", "Instruction B: focus on, continuous tracking, area B, time [45.2s, 46.5s]", format the integrated and optimized instruction content (attention level, operation requirements, spatial range, time interval), generate a unified scheduling instruction, and output it to the acquisition trigger module.

[0103] See also Figure 2 , the acquisition trigger module includes:

[0104] Node activation submodule: Based on the unified scheduling instructions, according to the time and space information in the instructions, the sensor of each collection node is activated. According to the geographical location and scheduling requirements of each node, the corresponding sensor is controlled to turn on and start data collection. The node sensor responds in time and starts working, and a list of activated collection nodes is obtained;

[0105] Based on the unified scheduling instructions output by the instruction integration submodule, including instruction A: "focus on, continue tracking, area A time [30.5s, 32.5s]" and instruction B: "focus on, continue tracking, area B, time [45.2s, 46.5s]", the system first parses the time and space information in instruction A. The time information is the start time t start,A =30.5s and end time t end,A = 32.5s, the spatial information is area A, which is defined by a coordinate range or volume, including a center point (x c ,y c ,z c) and a sphere defined by radius r, or a cube defined by minimum / maximum coordinates, the system queries the pre-established acquisition node database, which stores the unique identifier (NodeID) and geographic location coordinates (x n ,y n ,z n ) and sensor type and capability, perform a filtering operation to convert the geographic location of each node (x n ,y n ,z n ) is compared with the spatial region A of instruction A to determine whether the node position falls within the spatial range specified by the instruction, including, if region A is defined as a sphere with a center of [56500, -41000, 33500] meters and a radius of r = 5000 meters, the position of node NodeID101 is [54800, -42100, 33000] meters, and the distance from the node to the center of the region is calculated.

[0106] Since d = 2086m ≤ r = 5000m, node 101 is located in area A and is determined to be a node that needs to be activated. This spatial position comparison process is repeated for all nodes in the database to filter all nodes in area A. A}, according to the scheduling requirements of instruction A (focus on, continuous tracking) and the capabilities of the nodes (whether the sensor type matches the task requirements), to the selected node set {N A Each node in the system sends an activation control instruction, which includes the node ID, the sensor ID to be turned on, and the acquisition start time t start,A and end time t end,A, the command is sent through the preset communication network (including wireless sensor network or dedicated wired link), including sending the activation command {NodeID:101, Command:Activate, Sensor:S1, StartTime:30.5, EndTime:32.5} to node 101 at t=30.4s (slightly earlier than the start time to reserve response time). After receiving the command, the node controls the corresponding sensor S1 to turn on and start data collection on time at t=30.5s. The collected data stream includes sensor readings and timestamps. The node also sends The control center sends back status information verifying successful activation. When the control center receives verification information from node 101 (including {NodeID: 101, Status: Activated, Sensor: S1} received at t = 30.48s), it is considered that the node sensor responds in time and starts working. The node ID is added to the currently activated collection node list, and the same spatial screening, instruction sending, and status verification process is performed on instruction B. The IDs of all successfully activated nodes (regardless of whether they respond to instruction A or instruction B) are summarized to obtain the activated collection node list.

[0107] Buffer initialization submodule: Based on the list of activated collection nodes, an independent buffer is allocated to each activated node. The memory is initialized according to the size and storage requirements of the buffer, and the data is cleared so that the buffer of each node can store the real-time collected data and obtain the initialized buffer information.

[0108] Based on the list of activated acquisition nodes output by the node activation submodule, including the list [Node101, Node105, Node108, Node210, Node215], the system needs to allocate an independent memory buffer for each activated node in the list to temporarily store the data collected by the node in real time. The memory is initialized according to the size of the buffer and the storage requirements. First, the buffer size required for each node is determined. The size is based on the sensor data generation rate R of the node. data (including the unit of bytes per second, Bytes / s) and the duration T during which the node is activated active (by the t in the scheduling instruction end -t start Determine) calculated and multiplied by a safety factor k greater than 1 margin (including k margin =1.2), the calculation formula is BufferSize=R data ×T active ×k margin ,The safety factor sets the accumulation caused by short peak of data transmission or ,network delay. Setting it to 1.2 means that 20% additional space is reserved, including ,the activation time T of node 101 in response to instruction A.active,101 =32.5s-30.5s=2.0s, assuming the data rate of sensor S1 is R data,101 =10000Bytes / s(10KB / s), then the required buffer size BufferSize 101 =10000×2.0×1.2=24000Bytes (about 23.4KB), the system will allocate 24KB of memory space, node 210 responds to instruction B, activation time T active,210 =46.5s-45.2s=1.3s, set the data rate R data,210 =50000Bytes / s(50KB / s), then the required buffer size BufferSize 210 = 50000 × 1.3 × 1.2 = 78000 Bytes (approximately 76.2 KB). The system will allocate 80 KB (rounded up to the matching memory block size) of memory space for it. Repeat this calculation for all active nodes in the list to determine the required buffer size, as shown in Table 3 below.

[0109] Table 3 Activation node buffer allocation table

[0110]

[0111] As shown in Table 3, the calculation and allocation results of the buffer size for each activated node are listed. Based on the calculated allocation size, the system applies for a continuous memory block of a specified size (including 24KB) in the memory for each node (including Node101) and associates the starting address of the memory block with the node ID. After completing the memory allocation, the memory initialization operation is performed, which is to clear all bytes of each allocated memory block to zero (including calling the memset function to fill the memory area with 0). Before starting to receive data, the buffer does not include any old or invalid data, so that the buffer of each node is in a ready state and can store the real-time collected data. The ID of each node, the starting address of the allocated buffer, the buffer size and other information are recorded to obtain the initialized buffer information.

[0112] Data storage submodule: Based on the initialized buffer information, the collected data is partitioned and stored according to the timestamp and node ID, the data is written and redundancy checked, and the data is saved in the specified location to obtain the stored collected data;

[0113] Based on the buffer initialization submodule, the initialization buffer information provided includes the known node 101 corresponding to the buffer address Addr101, the size of which is 24KB; the node 210 corresponding to the buffer address Addr210, the size of which is 80KB, etc. When the activated acquisition node (including Node101) starts working and sends the collected data packets to the data processing center through the network, the data receiving module receives the data packets. Each data packet includes the ID of the source node (including 101), the timestamp of the data acquisition (including t=30.505s) and the sensor reading (including numerical values ​​or binary data blocks). The system searches for the initialized buffer according to the node ID (101) in the data packet. The buffer zone information, the corresponding buffer address (Addr101) and the current position of the written data (initial 0), write (copy) the sensor readings in the data packet to the current position of the node-specific buffer in the order of timestamps, update the position pointer of the written data, partition the collected data according to timestamps and node IDs and store them in the memory buffer, perform redundancy checks while performing data write operations, and check whether the check code (including CRC check code) of the data packet matches the data content. If they do not match, mark the data packet as an error or request retransmission, or check whether the timestamp of the data packet is roughly consistent with the expected order, including, if consecutive data packets are received from N The data packet timestamps of ode101 are 30.505s, 30.515s, 30.535s, and 30.525s respectively. If the timestamp of 30.525s is less than the previous one of 30.535s, an out-of-order event is recorded. The data will still be stored in the buffer for processing and sorting or correction. It can also be checked whether the buffer is about to be full, including when the buffer space of Node101 reaches 95% of the allocated size (24KB) (this is a write threshold, 24×0.95=22.8KB), a buffer overflow warning is triggered, and a data transfer operation is performed to transfer the data in the buffer (including the 22.8KB data starting from Addr101) to the buffer. The data is moved to a permanent storage medium (including a hard disk or database) and the pointer to the written location is reset or adjusted. The data writing operation continues until the acquisition end time of the node is reached (including t = 32.5s). The data is saved in the specified location. The location is a structured file system path or database table. The path structure can be designed based on the task ID, node ID and time period, including / data / mission-XYZ / node-101 / 30.5-32.5.dat, or stored in a database table, including fields: node ID, timestamp, data value, etc. After completing the data writing and transfer of all activated nodes, the stored acquisition data is obtained.

[0114] See also Figure 3 , methods include:

[0115] S1: The angular velocity change rate and angular momentum vector values ​​are acquired through real-time sensors, and a state potential energy function sequence is constructed based on the data. The state potential energy of adjacent time periods is compared and analyzed by difference comparison to determine whether a change in motion state has occurred. This is used as a basis for identifying the falling point and determining the characteristic stable segment.

[0116] S2: The real-time velocity vector of the flying projectile is obtained through the inertial navigation device, and the velocity vector is projected. The projection vector is then smoothed using the Kalman filter algorithm. The rate of change of the projection angle, which represents the second-order derivative of the angle between the velocity vector projection and the projectile axis, is calculated. The motion state change during the characteristic stable segment is compared with the rate of change of the projection angle to generate a projectile motion pattern analysis segment.

[0117] S3: The radar device obtains the projectile coordinates and the turning trajectory point set, combines the trajectory point and coordinate information to identify the path composite segment, and compares the time overlap range with the projectile motion pattern analysis segment to determine whether the time overlap threshold is met. If so, it is determined to be a multi-node warning segment;

[0118] S4: Receive multi-node warning segments, generate multiple prompt instructions based on node distribution, and integrate the instructions into a unified scheduling instruction;

[0119] S5: By receiving the unified scheduling instruction, each collection node sensor is activated, the buffer is initialized, and the real-time collected data is recorded and stored to obtain the stored collected data.

[0120] It should be understood that the term "and / or" as used herein simply describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A alone, A and B together, or B alone. A and B can be singular or plural. Furthermore, the character " / " as used herein generally indicates an "or" relationship between the associated objects, but it may also indicate an "and / or" relationship. For specific understanding, please refer to the context.

[0121] In this disclosure, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, "at least one of a, b, or c" can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or plural.

[0122] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0123] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0124] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0125] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, and can be electrical, mechanical, or other forms.

[0126] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0127] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0128] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and other media that can store program codes.

[0129] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. An equipment target control data acquisition system, characterized in that: The system comprises: The state potential energy module obtains the angular velocity change rate and angular momentum vector value of the projectile through real-time sensors, identifies the change in the projectile's motion state and the landing point, performs difference comparison, determines the characteristic stable segment, and transmits it to the velocity projection module; The velocity projection module obtains the real-time velocity vector of the flying projectile through the inertial navigation device, adjusts the Kalman filter gain coefficient based on the covariance matrix of the projectile's angular velocity change rate, calculates the projection angle change rate, compares the motion state changes of the characteristic stable segment, determines the projectile motion mode analysis segment, and transmits it to the trajectory fusion module; The trajectory fusion module obtains the projectile coordinates and turning trajectory point sets through radar equipment, identifies the composite segment of the projectile coordinates' movement path, and marks the intersection segment that overlaps with the execution time of the projectile motion pattern analysis segment as a multi-node warning segment, and transmits it to the collaborative preloading module; A collaborative preloading module receives the multi-node warning segments to generate multiple prompt instructions, and integrates the instructions into a unified scheduling instruction and outputs it to the acquisition trigger module; The acquisition trigger module receives the unified scheduling instruction, activates each acquisition node sensor, and performs buffer initialization and data recording and storage.

2. The equipment target control data acquisition system according to claim 1, characterized in that: The characteristic stabilization segment includes a stable time window, a state parameter set, and a stability identifier; the projectile motion pattern analysis segment includes a projection vector sequence, angular rate data, and pattern matching results; the multi-node warning segment includes a warning time interval, a key coordinate point set, and a fusion trajectory segment; the unified scheduling instruction includes a target node list, a trigger timing arrangement, and sensor configuration parameters; the buffer initialization includes an initialization state identifier, a recording data block, and a storage location pointer.

3. The equipment target control data acquisition system according to claim 1, characterized in that: The state potential energy module includes: Motion data acquisition submodule: obtains the time series of the projectile's angular velocity change rate and angular momentum vector value provided by the real-time sensor to represent the basic data set of the projectile's rotational motion state in a continuous time period and obtain the instantaneous motion state quantity; A state function generation submodule is configured to call the instantaneous motion state quantity, perform a differential operation on the angular velocity change rate time series to obtain the second-order time derivative, combine the second-order time derivative of the angular velocity, perform Z-score normalization on the second-order time derivative of the angular velocity and the angular momentum vector value, and calculate the state index value at each time point to establish a state function time series. Stable feature discrimination submodule: Based on the state function time series, the absolute difference of the state index values ​​at adjacent time points in the series is calculated, and the calculated difference is compared with the preset state potential energy difference threshold point by point to determine the motion state change and the falling point of the projectile, and determine the characteristic stable segment; The state potential energy difference threshold is calibrated and set by analyzing the wind tunnel test data of the missile body according to the observed fluctuation extreme value or change rate extreme value of the key parameters of angular velocity or attitude angle related to the sudden change of the flight state.

4. The equipment target control data acquisition system according to claim 1, characterized in that: The velocity projection module includes: Velocity vector acquisition submodule: uses inertial navigation equipment to record the velocity information of the projectile at each flight moment, extracts the velocity component at each moment, and obtains the velocity vector at a specified moment based on the motion state and position data of the projectile; The velocity projection processing submodule selects a projection direction that matches the velocity vector based on the velocity vector at the specified moment, performs a vector projection operation, maps the velocity vector to the corresponding reference axis, calculates the projected vector value at each moment, and obtains a velocity projection vector. Kalman filter submodule: Based on the obtained velocity projection vector, by introducing prior information, the Kalman filter is used to estimate the true value of the projection vector at each moment, remove noise and error, and adjust the filtering result according to the difference between the current measurement data and the predicted data to obtain a smooth projection vector; The measurement noise statistical characteristics in the prior information are obtained by fitting a Gaussian distribution model based on a sensor noise sampling data set collected when the missile is in a stationary state; Motion pattern comparison submodule: Based on the smoothed projection vector, the projection angle change rate at each moment is calculated. Combined with the motion state change of the characteristic stable segment, a comparative analysis of the projection angle change rate is performed through angle calculation and time series to obtain the projectile motion pattern analysis segment.

5. The equipment target control data acquisition system according to claim 4, characterized in that: The difference between the current measured data and the predicted data is processed using the formula: ； in is the Kalman gain, H represents the measurement matrix, The transpose of the measurement matrix H, Represents the predicted state covariance matrix, which describes the error propagation result of the state estimation in the process of predicting from time step i-1 to time step i, and R represents the covariance matrix of the measurement noise. is the covariance of the predicted measurements.

6. The equipment target control data acquisition system according to claim 1, characterized in that: The trajectory fusion module includes: Coordinate acquisition submodule: obtains the real-time missile body coordinate data provided by the radar equipment, extracts the coordinate points of the missile body by analyzing the signal returned by the radar and records the timestamp, unifies the coordinate data into position data in the global coordinate system, and obtains the missile body coordinate point set; Trajectory recognition submodule: Based on the projectile coordinate point set, it identifies continuous trajectory points, analyzes the spatial changes between trajectory points, combines the time interval and path changes, and identifies and determines the composite segment of the projectile motion path through the least squares method; Overlap judgment submodule: Based on the obtained composite segment of the projectile motion path and the projectile motion pattern analysis segment, the time interval of the composite segment of the projectile motion path and the projectile motion pattern analysis segment is calculated, the time range of the two is compared and calculated, and a time overlap threshold judgment is performed to extract the overlapping time period to obtain a multi-node warning segment; the time overlap threshold is calculated based on the standard deviation of the time length distribution of the pattern analysis segment of the projectile motion data, and the window length is set according to the standard deviation to judge whether the intersection length of the time period is greater than or equal to half of the window length.

7. The equipment target control data acquisition system according to claim 6, characterized in that: The difference between the actual observed data points and the predicted values ​​of the fitted model is processed using the formula ； Where E is the sum of squared errors, which is an indicator of the overall difference between the fitted model and the actual data, j represents the index of the data point, ranging from 1 to N, where N is the total number of data points. represents the independent variable value of the jth data point, is the dependent variable value of the jth data point, corresponding to the independent variable , is the constant term of the polynomial, is a linear term in the polynomial, is a quadratic term in the polynomial The coefficient of is a quadratic polynomial model with independent variables The predicted value at time .

8. The equipment target control data acquisition system according to claim 1, characterized in that: The collaborative preloading module includes: Node distribution generation submodule: receives the multi-node warning segments, divides the nodes into spatial distributions according to the time stamp and spatial coordinates of each node using a cluster analysis method, calculates the positions and time intervals between multiple nodes, and generates a prompt instruction set based on the node distribution; Instruction integration submodule: Based on the obtained prompt instruction set, the time and space ranges of multiple prompt instructions are merged, redundant instructions are eliminated, and unified scheduling instructions are generated by optimizing and adjusting instructions in the same time period, and output to the acquisition trigger module.

9. The equipment target control data acquisition system according to claim 1, characterized in that: The acquisition trigger module includes: Node activation submodule: Based on the unified scheduling instruction, according to the time and space information in the instruction, activate the sensor of each collection node, and according to the geographical location and scheduling requirements of each node, control the corresponding sensor to turn on and start data collection. The node sensor responds in time and starts working, and a list of activated collection nodes is obtained; Buffer initialization submodule: Based on the activated collection node list, allocate an independent buffer to each activated node, initialize the memory according to the size and storage requirements of the buffer, clear the data, so that the buffer of each node can store the real-time collected data, and obtain the initialized buffer information; Data storage submodule: Based on the initialized buffer information, the collected data is partitioned and stored according to the timestamp and node ID, a data write operation is performed and a redundancy check is performed, and the data is saved in a specified location to obtain the stored collected data.

10. A method for collecting equipment target control data, characterized in that: The method is implemented using the equipment target control data acquisition system according to any one of claims 1 to 9, and the method includes: S1: Obtain the angular velocity change rate and angular momentum vector value through real-time sensors, and construct a state potential energy function sequence based on the data. Compare and analyze the state potential energy of adjacent time periods through difference comparison to determine whether a change in motion state occurs, and use it as a basis for identifying the fallback point and determining the characteristic stable segment; S2: Obtain the real-time velocity vector of the flying projectile through the inertial navigation device, perform velocity vector projection operation, and smooth the projection vector through the Kalman filter algorithm to calculate the projection angle change rate of the second-order derivative representing the angle between the velocity vector projection and the projectile axial direction, and perform the motion state change and projection angle change rate based on the characteristic stable segment. Comparative analysis is performed to generate a projectile motion pattern analysis segment; S3: the projectile coordinates and the turning trajectory point set are obtained through the radar equipment, the path composite segment is identified by combining the trajectory point and coordinate information, and the time overlap range with the projectile motion pattern analysis segment is compared to determine whether the time overlap threshold is met. If so, it is determined to be a multi-node early warning segment; S4: receiving the multi-node early warning segment, generating multiple prompt instructions according to the node distribution, and integrating the instructions into a unified scheduling instruction; S5: by receiving the unified scheduling instruction, each acquisition node sensor is activated, the buffer is initialized, and the real-time acquired data is recorded and stored to obtain the stored acquisition data.

Images