CN-122017788-A - Three-dimensional constant false alarm multi-target detection and estimation method and device for low-precision quantized pulse phased array radar

CN122017788ACN 122017788 ACN122017788 ACN 122017788ACN-122017788-A

Abstract

The invention discloses a three-dimensional constant false alarm multi-target detection and estimation method for a low-precision quantized pulse phased array radar, which realizes high-precision estimation of target parameters by carrying out Newton correction on azimuth frequency, distance, doppler frequency and amplitude values in a received signal by introducing Newton method and circularly correcting other detection results by fixing part of detected results, eliminates the influence of the detected targets on the current target detection by recalculating detection statistics by taking the results of the detected targets into consideration, realizes joint detection and estimation on target azimuth angle, distance and speed by introducing a Rao detector, and ensures that the system can still maintain constant false alarm characteristics in a multi-target detection scene. The invention also provides a three-dimensional constant false alarm multi-target detection and estimation device. The method provided by the invention can realize the joint detection and estimation of the target azimuth angle, the distance and the speed with high accuracy, and maintain the constant false alarm characteristic in a multi-target detection scene.

Inventors

ZHU JIANG
Di Shuyi
GUO RUOHAI
ZHANG NING
ZHANG HANSHENG
WANG ZHIGANG
QU FENGZHONG

Assignees

浙江大学
中国船舶集团有限公司第七二四研究所

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (9)

1. The three-dimensional constant false alarm multi-target detection and estimation method for the low-precision quantized pulse phased array radar is characterized by comprising the following steps of: Step 1, acquiring quantized echo signals, screening targets in the echo signals through a pre-constructed binary hypothesis testing model, performing discrete grid point division in a distance dimension and a speed dimension, and screening out effective targets serving as first effective targets by using a Rao detector; Step 2, estimating signal parameters of the first effective target; Correcting signal parameters of a first effective target by adopting a Newton orthogonal matching algorithm, and adding a corrected first result into an effective target set; step 3, inputting the corrected first result in the step 2 into a binary hypothesis testing model to eliminate the influence of the first result on the residual undetected targets, recalculating detection statistics on the residual undetected targets to screen out effective targets as second effective targets, and estimating signal parameters of the second effective targets; Correcting signal parameters of a second effective target by adopting a Newton orthogonal matching algorithm, and adding a corrected second result into an effective target set; Fixing the second result at the same time, and circularly correcting the first result in the effective target set; and step 4, repeating the steps 2-3 until the termination condition is met, so as to obtain a final effective target set.
2. The three-dimensional constant false alarm multi-target detection and estimation method for low-precision quantized pulse phased array radar according to claim 1, wherein the expression of low-precision quantization is as follows: ; ; ; ; ; wherein, assuming that K targets exist in the echo signal, The amplitude value is represented by a value of the amplitude, The distance is indicated by the distance between the two points, The azimuth angular frequency is represented by the number of the azimuth angles, Representing Doppler frequency, definition Order-making Representing the amplitude by complex Two-dimensional real vector formed by splicing real part and imaginary part of the vector, and ; Is additive white Gaussian noise, meets the following requirements Wherein As the variance of the noise is the value of the variance of the noise, , , , 、 And Representing the indexes of the fast time domain, the slow time domain and the spatial domain respectively, In order to provide a sampling interval, Representing information about distance Is used to determine the manifold vector of the (c), Is a bandwidth of Is used for the transmission of the baseband waveform of (a), For the chirp rate of the chirped pulse, For the duration of the chirped pulse, Represents the carrier frequency and, Representing the propagation velocity of the electromagnetic wave, Representation of angular frequency with respect to azimuth Is a manifold vector of an array; representation of Doppler frequency Is provided with an array manifold vector of (a), Is a uniform quantizer, tensor Vectorization to obtain 。
3. The three-dimensional constant false alarm multi-target detection and estimation method for low-precision quantized pulse phased array radar according to claim 2, wherein the expression of the uniform quantizer is as follows: ; Wherein the method comprises the steps of Representing uniform quantizer Bit quantizer with output number of , Is the maximum full scale range.
4. The three-dimensional constant false alarm multi-target detection and estimation method for low-precision quantized pulse phased array radar according to claim 1, wherein the echo signals need to be subjected to linear processing after screening, and the process of the linear processing is as follows: Performing pulse compression on the signal by using a matched filtering method to obtain pulse compressed data; And carrying out Doppler oversampling FFT on the data after pulse compression.
5. The method for three-dimensional constant false alarm multi-target detection and estimation for low-precision quantized pulse phased array radar according to claim 1, wherein the Rao detector calculates detection statistics by calculating all detection units and compares the detection statistics with a threshold of a current effective target to judge whether the target exists in echo signals.
6. The three-dimensional constant false alarm multi-target detection and estimation method for low-precision quantized pulse phased array radar according to claim 5, wherein the expression of the threshold is as follows: ; Wherein, the And G represents the total number of received samples.
7. The method for three-dimensional constant false alarm multi-target detection and estimation for low-precision quantized pulse phased array radar according to claim 1, wherein the signal parameters include azimuth frequency, distance and doppler frequency of the target, and the correction process is as follows: constructing a corresponding gradient and hansen matrix based on the azimuth frequency, the distance and the doppler frequency of the target; Updating the roughly estimated amplitude, azimuth frequency, range and Doppler result based on the gradient and Hansen matrix; And adding the corrected azimuth angle frequency, distance, doppler frequency and amplitude value results into the effective target set.
8. The three-dimensional constant false alarm multi-target detection and estimation method for low-precision quantized pulse phased array radar according to claim 1, wherein the termination condition comprises: a. The repetition number reaches the preset maximum iteration number of the Rao detector; b. The Rao detector judges that the detection statistic corresponding to the maximum value point which is the target does not exist does not exceed the corresponding threshold value.
9. A three-dimensional constant false alarm multi-target detection and estimation device, characterized by being used for executing the steps of the three-dimensional constant false alarm multi-target detection and estimation method for the low-precision quantized pulse phased array radar according to any one of claims 1-8.

Description

Three-dimensional constant false alarm multi-target detection and estimation method and device for low-precision quantized pulse phased array radar Technical Field The invention belongs to the technical field of radar signal processing, and particularly relates to a three-dimensional constant false alarm multi-target detection and estimation method and device for a low-precision quantized pulse phased array radar. Background The pulse phased array radar has high distance resolution, long distance detection capability and highly flexible beam forming capability, and is widely applied to detection and estimation of targets. However, this type of radar system is limited by the wide bandwidth and large-scale antenna deployment that it requires, and power consumption and cost are extremely high. One promising solution is to introduce low resolution quantization (low-resolution quantization) techniques. However, the signal after nonlinear transformation will generate abundant harmonic components, and the weak target is easily covered by the strong target in the scene of coexistence of the strong and weak targets, so that the dynamic range (DYNAMIC RANGE, DR) is reduced, and the detection performance is seriously affected. While conventional matched filtering (MATCHED FILTER, MF) methods typically do so by performing pulse compression (pulse compression, PC) in the "fast-time" dimension and discrete fourier transforms (1D Discrete Fourier Transform,DFT) in the "slow-time" dimension between pulses. This approach suffers from the "discrete lattice" effect, with the risk of energy leakage in addition to targets located exactly on the distance and DFT lattice. The patent document with publication number CN111896927A discloses a communication-assisted radar target detection method based on 1-bit quantized sampling, which comprises the steps of carrying out 1-bit quantized sampling processing on a received radar detection signal and a base station communication signal, and comparing and judging based on a constructed detection operator and a detection threshold so as to confirm whether a target exists in a unit to be detected. The method utilizes the coverage advantage of the communication signal and the low power consumption characteristic of 1 bit quantization, and has a certain innovation on the signal cooperation and detection framework. However, the invention only focuses on judging whether the target exists or not, and does not relate to the subsequent estimation of specific parameters of the target for tracking, positioning and decision, and has limited applicability in practical application. In addition, the invention lacks systematic analysis of key problems such as limited accuracy, dynamic range compression, etc. caused by 1-bit quantization. The patent document with the publication number of CN109828253B discloses a multi-station radar quantitative fusion target detection method, which comprises the steps of firstly setting a multi-station radar system to cover a plurality of radar stations and a signal fusion center. Each radar station receives the echo signals and calculates detection statistics. And then, mapping and transforming each detection statistic to generate mapped statistic, and calculating the quantization threshold of each statistic through the set false alarm probability. Then, each transformed statistic is quantized to obtain a quantized flag value. These tag values are transmitted to a signal fusion center where a final fusion decision is made as to whether the target is present. The method improves the efficiency of the quantization fusion target detection to a certain extent through mapping transformation and quantization threshold setting. However, this invention lacks correction and feedback mechanisms and may present a high risk of missed detection or false alarms in the face of complex targets or uncertain environments. Disclosure of Invention The invention aims to provide a three-dimensional constant false alarm multi-target detection and estimation method for a low-precision quantized pulse phased array radar, which aims to solve the problems of limited dynamic range and harmonic generation caused by high nonlinearity of low-precision quantization and the problem of mutual covering or interference among targets during multi-target detection in a complex scene, realizes the joint detection and estimation of a target azimuth angle, a distance and a speed with high accuracy, and maintains the constant false alarm characteristic in the multi-target detection scene. In order to achieve the first purpose of the invention, the invention provides a three-dimensional constant false alarm multi-target detection and estimation method and device for a low-precision quantized pulse phased array radar, which comprises the following steps: Step 1, acquiring quantized echo signals, screening targets in the echo signals through a pre-constructed binary hypothesis testing model, performing discrete gr