A radar multi-target tracking and imaging echo simulation method and system

Abstract

The application provides a radar multi-target tracking and imaging echo simulation method and system, and belongs to the technical field of radar echo simulation. The method pushes the position and motion state of a radar platform and RCS point set data of multiple targets to an impact function generation FPGA. When a radar signal is detected, the FPGA latches data to form a multi-target data set. The point set data, attitude and world coordinates of the targets in the multi-target data set are split into multiple calculation instances, the sub-range image data of each calculation instance is calculated in parallel, and the sub-range image data of the multiple calculation instances is combined and superimposed into the range image data of a single target. Similarly, the range image data of all targets is calculated in parallel, and the multi-target range image data is convolved with the radar signal. After the convolution signal is obtained, signal delay and multi-branch amplitude-phase modulation are performed, and the amplitude-phase modulated convolution signal is output. The application can perform signal echo simulation in parallel and output high-quality large-scene echo signals.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar echo simulation, and in particular to a radar multi-target tracking and imaging echo simulation method and system. BACKGROUND

[0002] SAR radar can provide high resolution in range and azimuth, and is widely used on various platforms, such as airborne platforms, missile platforms, and satellite platforms. Due to the high cost of hanging the radar, the cost of obtaining data for application research or product testing is limited, especially the real-time SAR echo simulation and application testing under high-precision, three-dimensional large scenes, and multi-targets. In order to adapt to the signal processing of different PRIs of the radar, the simulator must also complete real-time data signal processing in a very short time.

[0003] However, most of the current echo simulation techniques are difficult to adapt to large scenes, multi-targets, high mobility, 10us or less level calculation delay, and high-quality imaging results. It is difficult to balance real-time and real accuracy of imaging results during the simulation of echo generation. SUMMARY

[0004] Therefore, the embodiments of the present application provide a radar multi-target tracking and imaging echo simulation method and system to solve the problem that the existing radar echo simulation technique is not real-time and is difficult to adapt to multi-targets, high mobility, and high-quality imaging scenes.

[0005] The technical scheme adopted by the present application is:

[0006] In a first aspect, the present application provides a radar multi-target tracking and imaging echo simulation method applied to a radar multi-target tracking and imaging echo simulation system, and the method comprises:

[0007] Setting the number of target simulations, and initializing the RCS point set data of a plurality of targets according to the target simulation number scene; the RCS point set data contains the RCS point relative centroid local coordinates and RCS gain value inside each target;

[0008] Converting the RCS point relative centroid local coordinates of the plurality of targets into a world coordinate system to obtain the attitude and world coordinates of the plurality of targets;

[0009] Pushing the radar platform position and motion state, and the RCS point set data, attitude, and world coordinates of the plurality of targets to the impact function generation FPGA according to a simulation beat of 1ms;

[0010] Detecting the radar signal through the signal acquisition FPGA, and triggering the signal PRI synchronization pulse to the impact function generation FPGA when the radar signal is detected;

[0011] The impact function generation FPGA latches the position and motion state of the radar platform and the attitude and world coordinates of the multiple targets after receiving the PRI synchronization pulse to form a multi-target data set;

[0012] The impact function generation FPGA splits the RCS point set data, attitude and world coordinates of each target in the multi-target data set into multiple calculation instances respectively, calculates the sub-range image data of each calculation instance in parallel, merges and superimposes the sub-range image data of multiple calculation instances to form single-target range image data, and obtains the multi-target range image data by summarizing all single-target range image data;

[0013] The multi-target signal convolution FPGA performs fast convolution on the multi-target range image data and the radar signal to form the convolution signals of the multiple targets;

[0014] The signal acquisition and output FPGA performs signal delay and multi-branch amplitude-phase modulation on the convolution signals of the multiple targets in parallel, and simultaneously performs signal superposition on the amplitude-phase modulated convolution signals in the same branch in parallel, and outputs to the radar antenna array.

[0015] Further, the motion state of the radar platform includes a velocity vector and an acceleration vector; the radar platform position and motion state, and the RCS point set data, attitude and world coordinates of the multiple targets are pushed to the impact function generation FPGA at a simulation pace of 1ms, and the following steps are performed:

[0016] In the three-dimensional scene, the radar platform transmits radar signals, the scene data is updated at an accurate time of 1ms, and according to the updated scene data, the accurate coordinate position of the radar platform at the time of 1ms is calculated as an initial condition according to the radar motion equation with the initial position, velocity vector and acceleration vector;

[0017] The accurate coordinate position, velocity vector and acceleration vector of the radar platform at the time of 1ms are pushed to the impact function generation FPGA at a simulation pace of 1ms through a position recursive formula, and the RCS point set data, attitude and world coordinates of the multiple targets are simultaneously pushed to the impact function generation FPGA.

[0018] Further, the impact function generation FPGA splits the RCS point set data, attitude and world coordinates of each target in the multi-target data set into multiple calculation instances respectively, calculates the sub-range image data of each calculation instance in parallel, merges and superimposes the sub-range image data of multiple calculation instances to form single-target range image data, and obtains the multi-target range image data by summarizing all single-target range image data, and the following steps are performed:

[0019] After the scene state is latched, the impact function generation FPGA splits the RCS point set data, the attitude and the world coordinates of each target in the multi-target data set into multiple calculation instances, and distributes the current position of the radar platform according to the motion equation recursion to each calculation instance to perform one-dimensional range image calculation in parallel, to obtain sub-range image data of the corresponding calculation instance; the attitude includes the azimuth of the target and the attitude angle relative to the world coordinate system;

[0020] The sub-range image data of each calculation instance is merged and superimposed into the complete range image data of the corresponding target to obtain single-target range image data, and all single-target range image data is summarized to obtain multi-target range image data.

[0021] Further, the impact function generation FPGA splits the RCS point set data, the attitude and the world coordinates of each target in the multi-target data set into multiple calculation instances, and distributes the current position of the radar platform according to the motion equation recursion to each calculation instance to perform one-dimensional range image calculation in parallel, to obtain sub-range image data of the corresponding calculation instance; the attitude includes the azimuth of the target and the attitude angle relative to the world coordinate system;

[0022] The attitude and the world coordinates of each target are obtained, the RCS point local coordinates relative to the center of mass and the RCS gain value are read from the RCS point set data of each target, and the RCS point local coordinates relative to the center of mass are rotated according to the attitude angle of the target relative to the world coordinate system to obtain the relative coordinates of the RCS point;

[0023] The relative coordinates of each RCS point are added to the world coordinates of the target to obtain the world coordinates P1 of the corresponding RCS point;

[0024] Based on the world coordinates P1 of each RCS point and the position of the radar platform, the slant range between the RCS point and the radar platform is calculated;

[0025] The slant range is subtracted from the reference slant range, and the difference obtained by the subtraction is divided by the signal processing distance resolution to obtain the range image slot;

[0026] According to the signal radio frequency RF obtained by frequency measurement, the wavelength is calculated to obtain 1 / λ, and 1 / λ is multiplied by 2 times the slant range to obtain the wave number;

[0027] The wave number is taken modulo with the quantization value of 2π to obtain the K value;

[0028] The K value is used as an index to perform table lookup in the trigonometric function stored in the impact function generation FPGA to obtain the complex phase gain M of the RCS point;

[0029] The complex phase gain M is multiplied by the RCS gain value of the RCS point, and the multiplication result is accumulated in the range image slot corresponding to the sub-range image to obtain the sub-range image data;

[0030] Until all calculation instances complete one-dimensional range image calculation, to obtain the sub-range image data of the corresponding target calculation instance.

[0031] Further, the signal acquisition and output FPGA performs the following steps on the convolution signals of multiple targets in parallel: signal delay and multi-branch amplitude-phase modulation, and signal superposition of the amplitude-phase modulated convolution signals in the same branch, and output to the radar antenna array.

[0032] The signal acquisition and output FPGA performs signal delay on the convolution signals of each target according to the slant range of each target and the corresponding target;

[0033] The electromagnetic distance attenuation formula is used to perform spatial attenuation modulation on the delayed convolution signals, and the spatial attenuation modulated convolution signals are input into N branches, and the direction pattern gain information and the phase difference information of the physical antenna array are modulated in parallel by using each branch modulation calculation instance.

[0034] In the same branch, the amplitude-phase modulated convolution signals of multiple targets are superimposed and output to the corresponding output branch of the radar antenna array.

[0035] In a second aspect, the radar multi-target tracking and imaging echo simulation system of the present application comprises: an external scene real-time deduction subsystem, a baseband signal processing subsystem, a radio frequency injection and transmission subsystem, and a global high-precision clock subsystem; wherein,

[0036] The external scene real-time deduction subsystem is used to set the number of target simulations, initialize the RCS point set data of multiple targets according to the number of target simulation scenarios, perform attitude conversion of the RCS points of multiple targets relative to the local coordinates of the mass center according to the world coordinate system, obtain the attitude and world coordinates of multiple targets, update the scene data at the accurate time of 1ms after transmitting the radar signal, and calculate the initial position and motion state of the radar platform at the simulation beat trigger time in the three-dimensional scene according to the updated scene data, and simultaneously synchronize the initial position and motion state of the radar platform and the RCS point set data of multiple targets to the baseband signal processing subsystem;

[0037] The baseband signal processing subsystem comprises a signal acquisition and output FPGA, an impact function generation FPGA, a multi-target signal convolution FPGA, and a platform position recursion module; wherein,

[0038] The platform position recursion module is used to push the radar platform position and motion state, and the RCS point set data, attitude, and world coordinates of multiple targets to the impact function generation FPGA according to the simulation beat of 1ms;

[0039] The impact function generation FPGA is used for synchronously locking the position and motion state of the radar platform and the attitude and world coordinates of the multiple targets when receiving the PRI synchronization pulse, forming a multiple target data set, and respectively splitting the RCS point set data, attitude and world coordinates of each target in the multiple target data set into multiple calculation instances, calculating the sub-range image data of each calculation instance in parallel, and merging and superimposing the sub-range image data of the multiple calculation instances to form single target range image data, and collecting all single target range image data;

[0040] The multiple target signal convolution FPGA is used for convolving the multiple target range image data with the radar signal to form the convolution signals of the multiple targets.

[0041] The signal acquisition and output FPGA is used for detecting the radar signal and generating the PRI synchronization pulse, transmitting the PRI synchronization pulse to the impact function generation FPGA, and performing signal delay and multi-branch amplitude and phase modulation on the convolution signals of the multiple targets in parallel, and performing signal superposition on the echo signals obtained through amplitude and phase modulation in the same branch in parallel and outputting the echo signals to the radar antenna array.

[0042] The radio frequency injection and emission subsystem is used for controlling the echo signal injection into the radar antenna array.

[0043] The global high-precision clock subsystem is used for assisting the signal acquisition and output FPGA in performing echo signal delay.

[0044] Therefore, the radar multi-target tracking and imaging echo simulation method and system have the following beneficial effects:

[0045] The radar multi-target tracking and imaging echo simulation method and system can access an external real-time complex electromagnetic countermeasure scene, and calculate the range image data, i.e., the impact response function, of the multiple targets in parallel according to the complex terrain and target RCS in the three-dimensional scene, perform fast convolution on the impact response function and the radar signal to form the echo signal, meet the requirements of low delay and high-quality target imaging, and can optimize the processing of large-scale target scattering points in parallel. Finally, the echo signal is subjected to high-precision delay modulation, and then subjected to amplitude and phase modulation through each radio frequency branch, and the echo signal with the multi-branch amplitude correlation relationship is output in a closed loop, which meets the requirement of accurate calculation of the echo signal of the high-mobility platform, and improves the phase accumulation accuracy of the echo signal between pulses and the imaging quality of the echo signal. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly introduced as follows, and other drawings can be obtained by those skilled in the art without creative labor on the premise that these drawings are within the protection scope of the present application.

[0047] Figure 1 A radar multi-target tracking and imaging echo simulation method flow chart of the present application;

[0048] Figure 2 A radar multi-target tracking and imaging echo simulation system architecture block diagram of the present application;

[0049] Figure 3 A radar platform position recursion process schematic diagram of the present application;

[0050] Figure 4 A data parallel computing pipeline schematic diagram of the present application;

[0051] Figure 5 A data parallel computing flow chart of the present application;

[0052] Figure 6 A parallel computing process schematic diagram of the range image data of the present application;

[0053] Figure 7 A signal delay and amplitude phase modulation principle schematic diagram of the present application;

[0054] Figure 8 A verification test scene schematic diagram of the present application;

[0055] Figure 9 A target RCS reference graph for testing of the present application;

[0056] Figure 10 A schematic diagram of real-time solving of the RCS impact function of a frame for two targets of the present application;

[0057] Figure 11 A schematic diagram of real-time solving of a frame of echo signals of the present application;

[0058] Figure 12 A radar imaging result schematic diagram of the present application. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. It should be noted that, in this document, relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. If there is no conflict, various features in the present application and embodiments can be combined with each other, and are all within the protection scope of the present application.

[0060] FPGA has natural advantages in real-time and parallel computing of signal processing, especially after algorithm pipeline optimization, which can greatly improve the algorithm data throughput rate, and multiple instances of parallel computing of multiple targets one-dimensional range image can also improve the computing rate of target signal.

[0061] Multi-antenna angle measurement of the target in the radar target indication guidance process is also an important means, so the system integrates the superposition output of multiple branches and multiple targets.

[0062] Most of the current echo simulation devices are difficult to adapt to the accurate calculation of delay and generation of high-quality imaging results in the multi-target, high-mobility, 10us and below level large scene, and it is difficult to balance between real-time and real accuracy of imaging results.

[0063] The current radar development trend tends to be multi-antenna array radar, which also requires the simulator to consider the relative amplitude correlation between branches according to the current real-time scene spatial relationship when simulating the echo signal of this type of radar.

[0064] Based on the above technical background, the present application provides a radar multi-target tracking and imaging echo simulation method and system with accurate signal, real-time processing, multi-branch, multi-target parallel pipeline calculation and output of high-quality large scene echo signal. The detailed implementation process of the present application is shown in the following examples.

[0065] Example 1: Referring to Figure 1 and Figure 3 The present application provides a radar multi-target tracking and imaging echo simulation method, which is applied to a radar multi-target tracking and imaging echo simulation system, and the method comprises:

[0066] Setting the number of target simulation, and initializing the RCS point set data of multiple targets according to the target simulation number scene; the RCS point set data contains the RCS point relative to the local coordinates of the mass center and the RCS gain value inside each target;

[0067] The RCS point relative to the local coordinates of the mass center of the multiple targets is posture converted according to the world coordinate system, to obtain the posture and world coordinates of the multiple targets;

[0068] According to the simulation beat of 1ms, the radar platform position and motion state, and the RCS point set data, posture and world coordinates of the multiple targets are pushed to the impact function generation FPGA;

[0069] The signal acquisition FPGA detects the radar signal, and when the radar signal is detected, a signal PRI synchronization pulse is triggered to the impact function generation FPGA;

[0070] The impact function generation FPGA latches the position and motion state of the radar platform and the attitude and world coordinates of the multiple targets after receiving the PRI synchronization pulse, and forms a multiple target data set;

[0071] The impact function generation FPGA splits the RCS point set data, attitude and world coordinates of each target in the multiple target data set into multiple calculation instances respectively, calculates the sub-range image data of each calculation instance in parallel, merges and superimposes the sub-range image data of the multiple calculation instances to form single target range image data, and obtains the multiple target range image data by summarizing all the single target range image data.

[0072] The multiple target signal convolution FPGA performs fast convolution on the multiple target range image data and the radar signal to form the convolution signals of the multiple targets.

[0073] The signal acquisition and output FPGA performs signal delay and multi-branch amplitude-phase modulation on the convolution signals of the multiple targets in parallel, and simultaneously performs signal superposition on the amplitude-phase modulated convolution signals in the same branch in parallel, and outputs to the radar antenna array.

[0074] Specifically, the application sets the target simulation number in the baseband signal processing system, and initializes the RCS point set data of the multiple targets through the scene, that is, sets the RCS point relative centroid local coordinates and RCS gain value and the like in each target.

[0075] Further, as shown in Figure 3 The motion state of the radar platform of the application includes a velocity vector and an acceleration vector, such as missile trajectory and instantaneous motion state vector information.

[0076] The radar platform transmits radar signals in a three-dimensional scene, updates the scene data at 1ms accurate time, and calculates the accurate coordinate position of the radar platform at 1ms according to the radar motion equation with the initial position, velocity vector and acceleration vector as the initial conditions.

[0077] The accurate coordinate position, velocity vector and acceleration vector of the radar platform at 1ms are recursively calculated according to the position recursive formula at 1ms simulation pace, and the RCS point set data, attitude and world coordinates of the multiple targets are synchronously pushed to the impact function generation FPGA at 1ms simulation pace.

[0078] ;

[0079] ;

[0080] ;

[0081] ;

[0082] wherein, represents the previous clock time, is the next clock time, represents the speed, represents the current acceleration of the target, represents the time interval of the FPGA clock cycle, represents the displacement amount caused by the speed at the previous clock time, represents the displacement amount caused by the acceleration at the previous clock time, represents the displacement amount caused by the speed at the next clock time, represents the displacement amount caused by the acceleration at the next clock time, and the initial value is 0. represents the quantization precision of the speed increment, represents the quantization precision of the acceleration, is the position parameter at the initial time of recursion, is the position parameter at the current time of recursion, is the position recursion error, is the real position parameter of the radar platform, represents the initial speed, represents a time point in the recursion process from the initial time of recursion 0s, corresponds to the recursion clock time; is the initial acceleration, and INT represents the INT integer function.

[0083] wherein, when calculating the accurate coordinate position of the radar platform, the motion equation of the radar platform can be used for calculation. The basic formula of the motion equation of the radar platform in the x motion axis is:

[0084] ;

[0085] wherein, is the position parameter at the initial time of recursion, is the initial speed, is the initial acceleration. After time, the x-axis coordinate position of the radar platform is , that is, is the initial position of the target in the x-axis at the current time. The formula can be extended to the position recursion of other axes (y-axis and z-axis) in three-dimensional space, and the principle is the same:

[0086] ;

[0087] ;

[0088] in, This represents the initial position of the target on the y-axis at the starting recursion time. for y Initial velocity of the shaft, for y Initial acceleration of the shaft. After... The y-axis coordinate position of the radar platform after the time is ,Right now This represents the starting position of the target on the y-axis at the current moment. This represents the initial position of the target on the z-axis at the initial recursive time. for z Initial velocity of the shaft, for z Initial acceleration of the shaft. After... Radar platform after time z The position of the axis coordinates is ,Right now This represents the starting position of the target on the z-axis at the current moment.

[0089] Since FPGAs require fixed-point quantization of data, all data generated by input impulse functions for the FPGA needs to be quantized to a specific bit width before being used as input.

[0090] The method for quantifying speed is as follows: , The quantization precision of the decimal places representing the speed increment. Indicates the time interval of the FPGA clock cycle. This represents the quantized velocity vector. This represents the radar platform velocity at the moment the simulation cycle is triggered, and round represents the rounding function.

[0091] The method for quantifying acceleration is as follows: , The decimal quantization precision representing acceleration. This represents the time interval of the FPGA clock cycle. This represents the quantized acceleration vector. The vector represents the acceleration vector of the radar platform at the moment the clock is triggered, and round represents the rounding function.

[0092] Based on the quantized data above, the FPGA single-cycle position discrete integral recursive method is as follows:

[0093] ;

[0094] ;

[0095] ;

[0096] ;

[0097] wherein, denotes the velocity vector of the last recursion, denotes the velocity vector of the current recursion, denotes the acceleration vector of the last recursion, denotes the acceleration vector of the current recursion, the quantization bit width of and the quantization bit width of are converted to be consistent, to obtain the quantized acceleration vector of the current recursion , facilitating the position recursion. is the position parameter of the current time in the recursion process, and the decimal precision of Q1 is finally reserved.

[0098] Suppose the initial position is 180Km, the initial velocity is 6000m / s, the acceleration is 200m / s^2, and the recursion is 1ms, then the calculation error according to the model is: 1.0e-04m. Meet the error acceptance range under the scene limit condition. Similarly, the position recursion of other axes also follows the same principle.

[0099] Further, as shown in Figure 3 , after starting the test, the RCS point set data of the measured multi-target is pushed according to the simulation beat of 1ms, that is, the RCS point relative centroid local coordinates and RCS gain value are pushed according to the simulation beat. In the impact function generation FPGA, according to the RCS point set data, the RCS point relative centroid local coordinates of the multiple targets are attitude converted in the world coordinate system (i.e. the earth-centered system), to obtain the attitude and world coordinates of the multiple targets; wherein, the attitude contains the azimuth of the target and the attitude angle relative to the world coordinate system; the attitude angle contains the pitch angle , the yaw angle and the roll angle .

[0100] Further, the present application detects the radar signal through the signal acquisition FPGA, and when the radar signal is detected, the signal PRI synchronization pulse is triggered to the impact function generation FPGA. After the impact function generation FPGA receives the PRI synchronization pulse, the position and motion state of the radar platform, and the attitude and world coordinates of the multiple targets are latched, to form a multi-target data set.

[0101] Among them, the impact function generation FPGA of the present application has the following calculation characteristics:

[0102] 1. Multi-instance parallel computing: split multi-target large-scale data into different parallel instances to reduce computing latency.

[0103] 2. Reduction calculation: eliminate read-write access conflicts of different parallel instances on complete data to avoid incorrect calculation results.

[0104] 3. Pipeline algorithm design, which divides the algorithm into multiple stages of parallel execution to further reduce computing latency.

[0105] Further, in the embodiment of the present application, the impact function generation FPGA splits the RCS point set data, attitude and world coordinates of each target in the multi-target data set into multiple computing instances, parallel computes the sub-range image data of each computing instance, and merges and superimposes the sub-range image data of multiple computing instances to form single-target range image data, and then aggregates all single-target range image data to obtain multi-target range image data, and executes the following steps:

[0106] After the scene state is latched, the impact function generation FPGA splits the RCS point set data, attitude and world coordinates of each target in the multi-target data set into multiple computing instances, and distributes the current position of the radar platform according to the motion equation recursion to each computing instance to perform one-dimensional range image calculation in parallel, to obtain sub-range image data corresponding to the computing instance.

[0107] The sub-range image data of each computing instance is merged and superimposed into the complete range image data of the corresponding target to obtain single-target range image data, and all target range image data is aggregated to obtain multi-target range image data.

[0108] The present application performs multi-instance parallel pipeline calculation of sub-range image data by impact function generation FPGA, as shown in Figure 4 The present application splits the RCS point set data of each target in the multi-target data set into multiple data points, each data point corresponding to a computing instance, and multiple data points exist in different stages of the pipeline algorithm at the same time, which can improve the throughput and computing latency of the algorithm.

[0109] Specifically, the present application distributes the current position of the radar platform according to the motion equation recursion to each computing instance to perform one-dimensional range image calculation in parallel, to obtain sub-range image data corresponding to the computing instance, and executes the following steps:

[0110] Obtain the attitude and world coordinates of each target, read the RCS point local coordinates relative to the centroid and RCS gain value from the RCS point set data of each target, and rotate the RCS point local coordinates relative to the centroid according to the attitude angle of the target relative to the world coordinate system to obtain the relative coordinates of the RCS point.

[0111] Add the relative coordinates of each RCS point to the world coordinates of the target to obtain the world coordinates P1 of the corresponding RCS point;

[0112] Based on the world coordinates P1 of each RCS point and the position of the radar platform, the slant range between the RCS point and the radar platform is calculated;

[0113] Subtract the slant range from the reference slant range, and divide the difference obtained by the distance resolution of signal processing to obtain the distance image slot position;

[0114] According to the signal radio frequency RF obtained by frequency measurement, calculate the wavelength one to obtain 1 / λ, and multiply 1 / λ by 2 times the slant range to obtain the wave number;

[0115] Take the quantization value of 2π as the modulus of the wave number to obtain the K value;

[0116] Take the K value as the index to perform table lookup in the triangular function stored in the impact function generation FPGA to obtain the complex phase gain M of the RCS point;

[0117] Multiply the complex phase gain M by the RCS gain value of the RCS point, and accumulate the multiplication result to the distance image slot position corresponding to the sub-distance image to obtain the sub-distance image data;

[0118] Until all calculation instances complete one-dimensional sub-distance image calculation, complete distance image data corresponding to the target calculation instance is obtained.

[0119] Specifically, after the multi-target data set is divided into multiple calculation instances, pipeline parallel calculation is performed on each calculation instance. As shown in Figure 5 In the impact function generation FPGA, each calculation instance is optimized by pipeline design, and the pipeline process is as follows:

[0120] Initialize all slot positions of the distance image of the calculation instance to 0, RANGE[SLOT_COUNT]=0, SLOT represents the slot position,

[0121] Get the target position (world coordinates), perform attitude transformation on the RCS point, and calculate the slant range according to the platform position of the transformed RCS point.

[0122] According to the slant range, the distance image slot position is calculated, SLOT=(slant range-reference slant range) / distance resolution.

[0123] Then, according to the measured signal frequency RF, calculate the wavelength one to obtain 1 / λ, and multiply 1 / λ by 2 times the slant range to obtain the wave number, i.e. wave number=2×slant range / λ.

[0124] Quantize 2π into N bits, and take the modulus of the wave number with the quantized value of 2π, that is, calculate K according to the slot position of the wave number in the trigonometric function lookup table, K = Mod(wave number, N), N represents that 0~2π is divided into N points to form an N-point trigonometric function complex number lookup table.

[0125] The signal phase gain is calculated, the K value is taken as an index, and the table lookup is performed in the trigonometric function generated by the RAM storage of the FPGA to obtain the complex phase gain M of the RCS point, M = exp(-i(K / N)), exp represents an exponential function, and i represents an imaginary unit.

[0126] The complex phase gain M is multiplied by the RCS gain value of the RCS point, that is, Then, the multiplication result is accumulated into the range image slot corresponding to the sub-range image to obtain the sub-range image data, and the corresponding formula is , wherein RANGE represents a range, and SLOT represents a slot.

[0127] When the RCS point of all calculation instances completes the one-dimensional range image calculation, the sub-range image generation is ended.

[0128] In the present application, if each target only has one complete range image buffer for superposition operation, a large probability will cause different calculation instances to operate the same slot of the complete range image at the same time, so that multiple calculation instances drive a register for calculation at the same time, which is unacceptable.

[0129] However, in the case that there are a large number of points in a single target, parallel calculation is needed to reduce the delay, and the single RCS point is split into multiple sub-point sets, and each sub-point set calculates the sub-range image data of multiple calculation instances of the target in parallel. Finally, the complete range image of the target is reduced and superimposed, which can avoid the above catastrophic error.

[0130] Referring to FIG. Figure 6 , the steps of splitting and calculating the sub-range image data of the multi-target data set are as follows:

[0131] Step 1: split the target into sub-point sets, for example, 65536 points of the target are divided into 32 calculation instances for parallel calculation, and each calculation instance needs to calculate 4096 points in a pipeline;

[0132] Step 2: complete the calculation of the sub-range image according to the description in the algorithm pipeline design chapter within a single calculation instance;

[0133] Step 3: after the calculation of the sub-range image is completed, the sub-range image data is merged and superimposed into the complete range image of the corresponding target.

[0134] Wherein, when the sub-range image data is merged and superimposed, the merging range image pseudo algorithm is described as follows:

[0135] Full range image [range image slot number];

[0136] Sub-range image [sub-range image number] [range image slot number];

[0137] For i=1: range image slot number

[0138] FOR j=1: sub-range image number

[0139] Full range image[i]=Full range image[i]+sub-range image[j][i];

[0140] END

[0141] END

[0142] Specifically, the FPGA has natural advantages in real-time signal processing and parallel computing, and after algorithm pipeline optimization, the data throughput rate of the algorithm can be greatly improved, and multiple instances of parallel computing of multiple target one-dimensional range images can also improve the computing rate of target signals.

[0143] Therefore, according to the scale of the radar RCS point in the three-dimensional scene, the number of impact function generation FPGAs can be expanded, a single FPGA configures N1 parallel instances to process M RCS points, the calculation pipeline depth is D, and the clock time interval is C, so the calculation time consumption of a single FPGA is is (unit: us):

[0144] ;

[0145] If a single FPGA is configured to process 65536 points, 64 parallel instances are used, the calculation algorithm pipeline depth is 514, and if a clock interval of 6.667ns is used, the single FPGA calculation output delay is: 7.3us, which meets the real-time requirements of most imaging radars.

[0146] The application utilizes the high-precision calculation rhythm of the impact function generation FPGA, recursively updates the accurate position of the platform within the difference range of the scene rhythm update time and the signal arrival time according to the current platform motion state, so as to compensate the distance difference value into the signal phase gain calculation, thereby improving the accuracy of the generation result of the echo signal.

[0147] In the application, the one-dimensional sub-range image data obtained by calculation also represents an impact function. The impact function generation FPGA receives a plurality of target RCS point set data and generates an impact function of each scattering point, and the formula of the impact function is specifically:

[0148] ;

[0149] wherein, is the RCS of each scattering point, represents the echo delay, is the impulse response of each scattering point, is the signal delay time of target i, is the starting reference time, is the slant range of target i from the radar at the current time, is the signal wavelength of the echo signal, is the Dirac function, used to represent the instantaneous response characteristics of the signal.

[0150] The accuracy of the slant range calculation of the RCS point directly affects the accuracy and accuracy of the signal generation. The impulse function generation FPGA starts the multi-target parallel computing unit, and calculates the slant range of each target to the radar platform in parallel according to the radar platform motion equation and the slant range calculation formula. Among them, the motion equation of the radar platform in the three-dimensional space is:

[0151] ;

[0152] The slant range calculation formula is:

[0153]

[0154] wherein, is the slant range between one of the scattering points (i.e. the RCS point) and the radar platform, which belongs to the i-th range resolution unit, represents the slant range of the target to the radar platform, is the position of the target sub-scattering point in the three-dimensional space at the current time.

[0155] When the signal front edge is detected, the impulse function generation FPGA latches the current radar platform position and the RCS point set of multiple targets. According to the position and motion state of the radar platform and the RCS point set data of multiple targets, the slant range between each target and the radar platform can be calculated, and the distance resolution unit where the slant range is located can be calculated according to the sampling rate.

[0156] Further, the signal acquisition and output FPGA performs signal delay and multi-branch amplitude-phase modulation on the convolution signals of multiple targets in parallel, and simultaneously performs signal superposition on the convolution signals subjected to amplitude-phase modulation in the same branch, and outputs to the radar antenna array, to perform the following steps:

[0157] According to the slant range of each target, the signal acquisition and output FPGA performs signal delay on the convolution signal of the corresponding target;

[0158] The convolution signal after delay is modulated by spatial attenuation formula, and the convolution signal after spatial attenuation modulation is input into N branches, and the pattern gain information and the phase difference information of the physical array are modulated in each branch modulation calculation instance in parallel.

[0159] In the same branch, the multiple target amplitude and phase modulated convolution signals are superimposed and output to the output branch corresponding to the radar antenna array.

[0160] In the application, after the signal acquisition and output FPGA obtains the multi-target convolution result from the DDR, independent time delay and amplitude and phase modulation are performed, and the superposition output is performed in each branch. The process is as follows: based on the branch amplitude correlation of the target, the echo signal after signal delay is modulated in amplitude and phase in N branches, and after amplitude and phase modulation, the echo signals in each branch are superimposed and output.

[0161] Referring to Figure 7 The process of signal delay and amplitude and phase modulation of the convolution signal (i.e. echo signal) is as follows:

[0162] Step 1: obtain the coordinates of each subarray of the radar antenna array ;

[0163] Step 2: perform parallel convolution of each target range image information (i.e. impact function) and radar signal;

[0164] Step 3: perform corresponding signal delay on the convolution signal according to the different slant ranges of each target;

[0165] Step 4: perform spatial attenuation modulation on the signal after signal delay according to the electromagnetic distance attenuation formula;

[0166] Step 5: input the signal after spatial attenuation modulation into multiple branches, and modulate the pattern gain information and the phase difference information of the physical array in each branch modulation calculation instance;

[0167] Step 6: superimpose the signals of each target in the same branch, and output to the output branch corresponding to the radar antenna array.

[0168] The electromagnetic distance attenuation formula is:

[0169] ;

[0170] is the slant range between one scattering point (i.e. RCS point) and the radar platform, which belongs to the i-th range resolution unit, represents the signal echo power, represents the radar transmitting power, represents the radar transmitting antenna gain, Indicates the radar receiving antenna gain. Indicates signal frequency, parameter This represents the speed of light in a vacuum. Indicates the signal propagation radius, A constant factor that is directly related to the properties of geometric space.

[0171] The above methods can achieve the amplitude and phase relationship of multiple targets in multiple branches of the radar aperture at the radio frequency signal level, so as to meet the function of multi-target angle measurement guidance.

[0172] This invention utilizes an impulse function to generate a high-precision calculation cycle for an FPGA. Based on the current platform motion state, it recursively calculates the platform motion distance difference within the range of the difference between the scene cycle update time and the signal arrival time. This distance difference is then used to compensate for the signal phase gain calculation, thereby improving the accuracy of the echo signal generation results.

[0173] Here, we assume the radar transmitted signal is: S(t) Then there is an echo signal model. for:

[0174] ;

[0175] in, This represents the mathematical symbol for convolution operation. The impact response at each scattering point, Let be the signal delay time for target i.

[0176] Furthermore, in this invention, the complex phase gain M is multiplied with the RCS gain value of the target point, and the multiplication result is accumulated into the range image slot corresponding to the sub-range image to obtain sub-range image data. This includes: first, multiplying the phase gain of the range difference with the phase gain of the range resolution unit in a complex manner, and then multiplying the complex multiplication result with the RCS gain value of the scattering point of the radar target to obtain the single-point RCS, i.e., sub-range image data.

[0177] Finally, this invention superimposes the single-point RCS of the radar target into the data slots within the range resolution unit index, until all single-point RCSs are mapped into the one-dimensional range resolution unit to form a one-dimensional range image of the radar target under the current viewpoint, i.e., single-target range image data.

[0178] In this invention, to achieve rapid calculation of echo signals for a large number of points, it is necessary to normalize the RCS points in the scene into different range resolution cells, and then superimpose the amplitude-phase product values ​​of all reflection points within each range resolution cell to form a one-dimensional range image that retains phase information within the range cell. The basic formula for forming a single point is:

[0179] ;

[0180] wherein, is the slant range between one scattering point (i.e. RCS point) and radar platform, which belongs to the i th distance resolution unit; is the RCS value of the corresponding target sub-scattering point; is the corresponding total RCS value of the i th distance resolution unit (i th distance gate) after normalization superposition; represents the process of calculating the RCS point phase gain of the sub-scattering point according to the slant range; is the Dirac function, which is used to represent the instantaneous response characteristics of the signal.

[0181] In the present application, when the echo signal is input to each subarray of the radar array, the corresponding amplitude relationship needs to be met, and the amplitude relationship between each branch needs to be calculated according to the radar pattern. The amplitude relationship calculation process between each branch is as follows:

[0182] (1) For the pattern query mode of multiple antennas:

[0183] Array radar direction Figure 1 Generally represented by sinusoidal space, this representation can overcome the disadvantage of beam width deformation caused by the scanning angle deviating from the normal, and at the same time, the scanning loss caused by the aperture becoming smaller after the scanning angle deviates from the normal needs to be considered.

[0184] In the array spherical coordinate system, respectively represent the slant range, the pitch angle, and the azimuth angle. The conversion relationship between the array rectangular coordinate system and the array spherical coordinate system is:

[0185] ;

[0186] ;

[0187] ;

[0188] Mapping the antenna pattern to the sinusoidal space , :

[0189] ;

[0190] ;

[0191] From the shape invariance of the half-power point equal-intensity line of the antenna pattern mapped to the sinusoidal space in the array spherical coordinate, the pattern of the sinusoidal space satisfies:

[0192] ;

[0193] wherein, represents the beam pattern query by the sinusoidal space ); represents the data amount deviating from the 0 point of the x axis of the sinusoidal space; represents the data amount deviating from the 0 point of the y axis of the sinusoidal space.

[0194] Then, when the beam pattern deviating from the normal direction is queried, the following formula can be used:

[0195] Suppose that the target pointing azimuth is: , , and the beam pointing azimuth is: , .

[0196]

[0197]

[0198] and satisfies . Wherein, represents the coordinate position of the target in the original direction pattern sinusoidal space x axis query pattern; represents the coordinate position of the target in the original direction pattern sinusoidal space y axis query pattern.

[0199] The present application can use the values of and to query the gain value of the antenna pattern in the normal direction.

[0200] (2) Array antenna scanning surface loss calculation:

[0201] ;

[0202] is the scanning loss corresponding to the scanning angle, is the lossless antenna aperture, is the antenna aperture with the scanning angle , according to the geometric mapping relationship between the scanning angle and the effective aperture , the scanning loss can be further derived as:

[0203] ;

[0204] Wherein, is a scalable coefficient, is the scanning loss corresponding to the scanning angle.

[0205] Then the amplitude relationship between each branch can be represented as follows:

[0206] ;

[0207] wherein, represents the radar signal echo power calculated by the basic radar equation; represents the signal output power corresponding to the first branch; represents the gain of the directional diagram of the target in the antenna subarray obtained by querying the directional diagram of the subarray through the sine spatial coordinates; represents the signal output power corresponding to the nth branch; represents the gain of the directional diagram of the target in the antenna subarray obtained by querying the directional diagram of the subarray through the sine spatial coordinates.

[0208] (3) Array wave path difference and phase solution:

[0209] When the echo signal is output to the radar antenna array multi-subarray, the phase relationship also exists in the three-dimensional space as the amplitude relationship. The phase relationship of the target with each subarray corresponding radio frequency branch can be solved by calculating the wave path difference of the target to each subarray.

[0210] In the array spherical coordinate system, assuming that the target is at the , angle of the antenna,

[0211] ;

[0212] ;

[0213] ;

[0214] ;

[0215] wherein, represents the normalized unit vector obtained by inputting the rotation angle and the deviation angle of the target in the antenna coordinate system, represents the coordinate point of the target in the x-axis of the array spherical coordinate, represents the coordinate point of the target in the y-axis of the array spherical coordinate, represents the coordinate point of the target in the z-axis of the array spherical coordinate; represents the vector coordinates of the spherical coordinates.

[0216] Assuming that the coordinates of each subarray of the array are , the wave path difference and the phase corresponding to the antenna coordinate origin are:

[0217] ;

[0218] ;

[0219] wherein, represents the wave path difference of the target relative to the 0 coordinate point of the relative array of the i th subarray; represents the phase difference of the target relative to the 0 coordinate point of the relative array of the i th subarray, which can be directly modulated on the corresponding branch. represents the signal wavelength of the echo signal.

[0220] Embodiment 2: Reference Figure 2 As shown in the figure, the embodiment provides a radar multi-target tracking and imaging echo simulation system, which comprises an external scene real-time deduction subsystem, a baseband signal processing subsystem, a radio frequency injection and transmission subsystem and a global high-precision clock subsystem.

[0221] The external scene real-time deduction subsystem is configured to set the target simulation number, initialize the RCS point set data of a plurality of targets according to the target simulation number scene, perform attitude conversion on the RCS points of the plurality of targets relative to the local coordinates of the center of mass in the world coordinate system, obtain the attitude and world coordinates of the plurality of targets, update the scene data at the accurate time of 1 ms after transmitting the radar signal, and calculate the initial position and motion state of the radar platform at the simulation beat trigger time in the three-dimensional scene according to the updated scene data, and simultaneously synchronize the initial position and motion state of the radar platform and the RCS point set data of the plurality of targets to the baseband signal processing subsystem.

[0222] Figure 2 The baseband signal processing subsystem comprises a signal acquisition and output FPGA, an impact function generation FPGA, a multi-target signal convolution FPGA and a platform position recursion module. The signal acquisition and output FPGA is configured to receive the initial position and motion state of the radar platform and the RCS point set data, attitude and world coordinates of the plurality of targets, and output the initial position and motion state of the radar platform and the RCS point set data, attitude and world coordinates of the plurality of targets to the impact function generation FPGA.

[0223] The platform position recursion module is configured to push the initial position and motion state of the radar platform and the RCS point set data, attitude and world coordinates of the plurality of targets to the impact function generation FPGA according to the simulation beat of 1 ms.

[0224] The impact function generation FPGA is configured to, when receiving the PRI synchronization pulse, synchronize and latch the initial position and motion state of the radar platform and the RCS point set data, attitude and world coordinates of the plurality of targets, form a multi-target data set, simultaneously split the RCS point set data, attitude and world coordinates of each target in the multi-target data set into a plurality of calculation instances, calculate the sub-range image data of each calculation instance in parallel, merge and superimpose the sub-range image data of the plurality of calculation instances to form single-target range image data, and collect all single-target range image data.

[0225] The multi-target signal convolution FPGA is configured to convolve the multi-target range image data with the radar signal to form the convolution signals of the plurality of targets.

[0226] The signal acquisition and output FPGA is used for detecting radar signals and generating a PRI synchronization pulse, transmitting the PRI synchronization pulse to the impact function generation FPGA, and performing signal delay and multi-branch amplitude-phase modulation on the convolution signals of multiple targets in parallel, and performing signal superposition on the echo signals obtained through amplitude-phase modulation in the same branch in parallel and outputting the signals to the radar antenna array.

[0227] The radio frequency injection and transmission subsystem is used for controlling the injection of echo signals into the radar antenna array.

[0228] The global high-precision clock subsystem is used for assisting the signal acquisition and output FPGA in echo signal delay, providing a synchronous clock signal for the whole system, ensuring strict signal phase continuity among multiple systems, reducing the local oscillator phase noise of the system, and avoiding signal spectrum broadening among multiple systems.

[0229] In the embodiment of the present application, Figure 2 In the embodiment of the present application, the external scene real-time deduction subsystem includes an external real-time scene module, a reflection memory module and a real-time calculation module. The external real-time scene module is used for setting the number of radar target simulations, and initializing the RCS information of multiple radar target scattering points in a three-dimensional scene. The real-time calculation module is used for calculating the position coordinates, velocity vector and acceleration vector of the radar platform in the three-dimensional scene, and the position coordinates and attitude, translation and branch amplitude correlation of multiple radar targets in the world system according to scene update data in real time.

[0230] Specifically, in the working process of the external scene real-time deduction subsystem, the real-time scene is updated at a 1ms accurate time in the reflection memory, the real-time calculation module receives the scene update interrupt to quickly calculate the current motion state (position coordinates, velocity vector, acceleration vector information, etc.) of the radar platform in the three-dimensional space, the coordinates and attitude of the target in the world system, the branch amplitude correlation of each target, etc., and synchronously transmits the data to the impact function generation FPGA and the signal acquisition and output FPGA through a high-speed interface.

[0231] Specifically, the working principle of the radar multi-target tracking and imaging echo simulation system in the embodiment of the present application is as follows:

[0232] Step 1: The external scene real-time deduction subsystem sets the number of target simulations, and initializes the RCS reference image of multiple radar targets (which can be mobile target and terrain RCS information);

[0233] Step 2: Start the test, and the external test scene pushes the radar platform position, velocity and acceleration (which can be used for high-precision position recursion of a high-speed platform) at a 1ms high-precision simulation beat, and the positions and attitudes of multiple radar targets or a single target in the world system (slow target);

[0234] Step 3: The signal acquisition FPGA detects the radar signal, triggering the PRI synchronization pulse into the impact function generation FPGA;

[0235] Step 4: The impact function generation FPGA latches the current scattering point position and the attitude under the coordinate system;

[0236] Step 5: Start the multi-target parallel computing unit to start calculation;

[0237] Step 6: In the impact function generation FPGA, obtain the attitude and world coordinates of each target, read the RCS point relative centroid local coordinates and RCS gain value from the RCS point set data of each target, and rotate the RCS point relative centroid local coordinates according to the attitude angle of the target relative world coordinate system to obtain the relative coordinates of the RCS point; add the relative coordinates of each RCS point to the world coordinates of the target to obtain the world coordinates P1 of the corresponding RCS point; based on the world coordinates P1 of each RCS point and the position of the radar platform, calculate the slant range between the RCS point and the radar platform;

[0238] Step 7: Subtract the slant range from the reference slant range, and divide the difference obtained by the distance resolution of signal processing to obtain the distance image slot position within the distance resolution unit index;

[0239] Step 8: According to the signal radio frequency RF obtained by frequency measurement outside, calculate the wavelength to obtain 1 / λ, multiply 1 / λ by 2 times the slant range to obtain the wave number; take the modulus of the wave number and the quantization value of 2π to obtain the K value; take the K value as an index, and perform table lookup in the trigonometric function stored in the impact function generation FPGA to obtain the complex phase gain M of the RCS point;

[0240] Step 9: Multiply the complex phase gain M of the previous step and the RCS gain value of the RCS point to obtain the RCS of a single point;

[0241] Step 10: Parallelly reduce and superimpose the single-point RCS into the data slot position within the distance resolution unit index to obtain sub-range image data, until all point RCS values are mapped into one-dimensional distance resolution unit to form one-dimensional range image under the current view angle, to obtain single-target range image data, and all target single-target range image data is obtained, that is, multi-target range image data;

[0242] Step 11: After the multi-target range image data is processed in parallel, the multi-target signal convolution FPGA is completed with the radar signal for convolution processing;

[0243] Step 12: After the multi-target convolution result, the signal delay and multi-branch amplitude and phase modulation are performed in the signal acquisition and output FPGA, and the signal is superimposed and output to the subarray back-end input port of the measured radar antenna array.

[0244] Specifically, in the application, the baseband signal processing subsystem is composed of multiple large-scale field programmable gate arrays, which are respectively used for high-precision interpolation recursion of radar platform position, parallel pipeline calculation of multi-radar target or large-scale scene RCS point scattering point slant range, impact function generation (multi-target can be selected to be split or large-scale scene can be combined), multi-target or large-scale signal correlation processing, real-time output of convolution result to verify signal calculation quality, signal acquisition, multi-branch amplitude and phase modulation, and high-precision time delay modulation output.

[0245] The application utilizes the high-precision calculation rhythm of the FPGA to recursively update the platform motion distance difference within the difference range of the scene rhythm and the signal arrival time according to the current platform motion state, so as to compensate the distance difference value into the signal phase gain calculation, thereby improving the accuracy of the generation result of the echo signal. The embodiment of the application can combine a large number of RCS points into a single large-scale scene (output a single branch) or split them into multiple targets for separate simulation (output multiple branches).

[0246] The system improves the data output throughput and reduces the time delay of impact function generation by utilizing the parallelism and pipeline optimization characteristics of the FPGA in calculation. Meanwhile, according to the parallel output of the impact function, the convolution of the signal is processed in parallel in the multi-target signal convolution FPGA, and the convolution result is output to the signal acquisition and output FPGA, and the echo signals of the multi-radar targets after the convolution of the upper stage are respectively subjected to high-precision time delay, multi-branch phase and amplitude modulation, and the echo signals of the multi-radar targets after the delay processing are output to each output branch of the radar subarray in pipeline.

[0247] The whole process fully utilizes the parallelism and large throughput characteristics of the programmable gate array to reduce the delay output of the signal and achieve real-time SAR echo simulation of a large number of RCS points.

[0248] The application mainly solves the echo simulation of radar multi-target tracking and imaging in a complex multi-target and large-scale scene. The application can access an external real-time complex electromagnetic countermeasure scene, quickly form an impact response function (i.e. range image data) according to the complex terrain and target scattering point RCS in a three-dimensional scene, and form an echo signal by quickly convolving the impact response function with the radar signal. After high-precision time delay modulation of the signal, the signal is subjected to amplitude and phase modulation of each radio frequency branch, and finally the signal is closed-loop output to a radar signal processor. Meanwhile, the target angle measurement tracking function is met, and the angle measurement guidance simulation (array signal simulation and single pulse signal simulation) of the radar is completed.

[0249] Specifically, refer to Figure 8 、 Figure 9 、 Figure 10 、 Figure 11 andFigure 12 The present application verifies the quality and accuracy of the system output signal, and obtains target imaging results after correlatively compressing in the azimuth direction and pulse compressing in the distance direction of the monitoring digital signal of the multi-target signal convolution FPGA output. Figure 8 is a verification test scene schematic diagram, the starting coordinates of the target in the verification scene are (-150, 0, 5000), the starting speed is (6000, 0, 0), and the acceleration is (0, 0, 0). The radar tracking and echo simulation of imaging are performed from the synthetic starting point, and the synthetic aperture length of 300 m is passed on the x-axis to reach the synthetic end point, and projection imaging is performed according to the target motion to the calibration point. Figure 9 is a test target RCS reference map, which needs to be pixel mapped to the world coordinates, and the pixel gray value is used as the RCS value. Figure 10 is a schematic diagram of real-time solving of the RCS impact function of a target in a frame, Figure 10 The x-axis is the distance resolution unit index. Figure 11 is a schematic diagram of a real-time solved echo signal in a frame, Figure 11 The x-axis is the distance resolution unit index. Figure 12 The radar program result shown in is the imaging result in the distance range of 19300-21300 m, and the results of two targets with different orientations can be clearly distinguished.

[0250] The simulation scene data are parameter injected by an external scene at 1 ms, and then high-precision target slant range recursion is performed on a large number of points by the FPGA, and finally the quality and accuracy verification results of the signal are shown in Table 1.

[0251] Table 1 Quality and accuracy verification results of the signal

[0252]

[0253] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A radar multi-target tracking and imaging echo simulation method, applied to a radar multi-target tracking and imaging echo simulation system, characterized in that, The methods include: Set the number of simulated targets, and initialize the RCS point set data of multiple targets according to the scenario of the number of simulated targets; the RCS point set data includes the local coordinates of the RCS points inside each target relative to the centroid and the RCS gain value; The attitude of multiple targets is obtained by transforming the local coordinates of their RCS points relative to their centroids in the world coordinate system. The radar platform position and motion state, as well as the RCS point set data, attitude and world coordinates of multiple targets are pushed to the impact function to generate the FPGA according to the simulation cycle of 1ms. The radar signal is detected by the FPGA, and when the radar signal is detected, the PRI synchronization pulse is triggered to the impulse function to generate the FPGA. The impulse function generates the position and motion state of the radar platform after the FPGA receives the PRI synchronization pulse, as well as the attitude and world coordinates of multiple targets, forming a multi-target dataset. The FPGA generates the impact function by splitting the RCS point set data, attitude and world coordinates of each target in the multi-target dataset into multiple computational instances, computing the sub-range image data of each computational instance in parallel, merging and superimposing the sub-range image data of multiple computational instances to form single-target range image data, and summing all single-target range image data to obtain multi-target range image data. Multi-target signal convolution FPGA is used to quickly convolve multi-target range profile data with radar signals to form convolution signals of multiple targets; The FPGA performs parallel signal delay and multi-branch amplitude-phase modulation on the convolutional signals of multiple targets through signal acquisition and output. At the same time, the amplitude-phase modulated convolutional signals are superimposed in parallel in the same branch and output to the radar antenna array.

2. The radar multi-target tracking and imaging echo simulation method according to claim 1, characterized in that, The motion state of the radar platform includes velocity and acceleration vectors; the radar platform's position and motion state, along with the RCS point set data, attitude, and world coordinates of multiple targets, are pushed to the impact function to generate the FPGA according to a 1ms simulation cycle, and the following steps are performed: In a 3D scene, a radar platform transmits radar signals and updates scene data at precise 1ms intervals. Based on the updated scene data, the precise coordinates of the radar platform at 1ms are calculated using the initial position, velocity vector, and acceleration vector as initial conditions, according to the radar motion equation. Using the position recursion formula, the precise coordinates, velocity vector, and acceleration vector of the radar platform at 1ms are pushed to the FPGA for generating the impact function at a simulation cycle of 1ms. At the same time, the RCS point set data, attitude, and world coordinates of multiple targets are pushed to the FPGA for generating the impact function.

3. The radar multi-target tracking and imaging echo simulation method according to claim 1, characterized in that, The impact function generation FPGA splits the RCS point set data, attitude, and world coordinates of each target in the multi-target dataset into multiple computational instances, calculates the sub-range image data of each computational instance in parallel, merges and superimposes the sub-range image data of multiple computational instances to form single-target range image data, and summarizes all single-target range image data to obtain multi-target range image data. The following steps are performed: After the scene state is latched, the FPGA generates the impact function and splits the RCS point set data, attitude and world coordinates of each target in the multi-target dataset into multiple computing instances. The radar platform is assigned to each computing instance according to the current position recursively derived from the motion equation and performs one-dimensional range image calculation in parallel to obtain the sub-range image data of the corresponding computing instance. The attitude includes the azimuth of the target and the attitude angle relative to the world coordinate system. The sub-range image data of each calculation instance is merged and superimposed onto the complete range image data of the corresponding target to obtain single-target range image data. At the same time, all single-target range image data are summarized to obtain multi-target range image data.

4. The radar multi-target tracking and imaging echo simulation method according to claim 3, characterized in that, The process involves allocating the current position of the radar platform, recursively derived from the equation of motion, to each computation instance for parallel one-dimensional range image calculation, thereby obtaining sub-range image data for the corresponding computation instance. The following steps are then performed: The attitude and world coordinates of each target are obtained, and the local coordinates of the RCS point relative to the centroid and the RCS gain value are read from the RCS point set data of each target. The local coordinates of the RCS point relative to the centroid are rotated according to the attitude angle of the target relative to the world coordinate system to obtain the relative coordinates of the RCS point. Add the relative coordinates of each RCS point to the world coordinates of the target to obtain the world coordinates P1 of the corresponding RCS point; Based on the world coordinates P1 of each RCS point and the position of the radar platform, calculate the slant distance between the RCS point and the radar platform. Subtract the slant range from the reference slant range, and divide the difference by the signal processing range resolution to obtain the range image slot. Based on the radio frequency (RF) of the signal obtained from external frequency measurement, calculate 1 / λ, which is one-half of the wavelength. Multiply 1 / λ by twice the slant distance to obtain the wave number. The K value is obtained by taking the modulus of the wavenumber with the quantization value of 2π. Using K as an index, the complex phase gain M of the RCS point is obtained by looking up a table in the trigonometric functions stored in the FPGA where the impulse function is generated. The complex phase gain M is multiplied by the RCS gain value of the RCS point, and the multiplication result is accumulated into the range image slot corresponding to the sub-range image to obtain the sub-range image data; This continues until all computation instances have completed one-dimensional distance image calculations, resulting in sub-distance image data for the corresponding target computation instance.

5. The radar multi-target tracking and imaging echo simulation method according to claim 1, characterized in that, The method involves using an FPGA to perform parallel signal delay and multi-branch amplitude-phase modulation on the convolutional signals of multiple targets, and simultaneously superimposing the amplitude-phase modulated convolutional signals in the same branch before outputting them to the radar antenna array. The steps include: The signal is delayed by using the FPGA for signal acquisition and output, according to the slant range of each target and the convolution signal of the corresponding target. The delayed convolutional signal is spatially attenuated and modulated using the electromagnetic distance attenuation formula. The spatially attenuated and modulated convolutional signal is then input into N branches. The modulation calculation instance of each branch is used to parallelly modulate the pattern gain information and the phase difference information of the physical antenna array. In the same branch, the convolutional signals of multiple targets after amplitude and phase modulation are superimposed and output to the corresponding output branch of the radar antenna array.

6. A radar multi-target tracking and imaging echo simulation system, characterized in that, include: The system comprises an external scenario real-time simulation subsystem, a baseband signal processing subsystem, a radio frequency injection and transmission subsystem, and a global high-precision clock subsystem; among which... The external scene real-time simulation subsystem is used to set the number of simulated targets and initialize the RCS point set data of multiple targets according to the number of simulated targets. It then performs attitude transformation on the local coordinates of the RCS points of multiple targets relative to the centroid in the world coordinate system to obtain the attitude and world coordinates of multiple targets. At a precise 1ms after the radar signal is transmitted, the scene data is updated, and the initial position and motion state of the radar platform at the simulated beat trigger time in the three-dimensional scene are calculated based on the updated scene data. At the same time, the initial position and motion state of the radar platform, as well as the RCS point set data of multiple targets, are synchronized to the baseband signal processing subsystem. The baseband signal processing subsystem includes a signal acquisition and output FPGA, an impulse function generation FPGA, a multi-target signal convolution FPGA, and a platform position recursion module; wherein... The platform position recursion module is used to push the radar platform position and motion state, as well as the RCS point set data, attitude and world coordinates of multiple targets to the impact function to generate the FPGA according to the simulation cycle of 1ms. The impulse function generates an FPGA that, upon receiving a PRI synchronization pulse, synchronously latches the position and motion state of the radar platform, as well as the attitude and world coordinates of multiple targets, forming a multi-target dataset. Simultaneously, it splits the RCS point set data, attitude, and world coordinates of each target in the multi-target dataset into multiple computational instances, calculates the sub-range image data of each computational instance in parallel, and merges and superimposes the sub-range image data of multiple computational instances to form a single-target range image data, and summarizes all single-target range image data. The multi-target signal convolution FPGA is used to convolve multi-target range image data with radar signals to form convolution signals of multiple targets; The signal acquisition and output FPGA is used to detect radar signals and generate PRI synchronization pulses. At the same time, the PRI synchronization pulses are transmitted to the impulse function generation FPGA, and the convolution signals of multiple targets are subjected to signal delay and multi-branch amplitude-phase modulation in parallel. Simultaneously, the echo signals obtained by amplitude-phase modulation are superimposed in the same branch and output to the radar antenna array. The radio frequency injection and transmission subsystem is used to control the injection of echo signals into the radar antenna array; The global high-precision clock subsystem is used to assist the signal acquisition and output FPGA in delaying the echo signal.

Images