Maritime variable acceleration target motion parameter estimation and segmentation method

Abstract

The application provides a sea variable acceleration target motion parameter estimation and segmentation method, which comprises the following steps: (1) dividing the received radar echo into several CPI segments in the slow time dimension; (2) performing initial velocity estimation on each CPI segment based on the instantaneous autocorrelation function spectrum, and estimating the acceleration according to the estimated initial velocities between adjacent CPIs; (3) combining the CPIs with similar acceleration according to the estimated acceleration, taking the time points with large acceleration change as the segmentation points of the motion, and segmenting the target motion process. Simulation experiments show that the method can realize the motion segmentation of the target and accurately estimate the parameters of each motion segment, and provides prior information for long-time coherent accumulation of the sea target.

Description

Technical fields:

[0001] This invention belongs to the field of radar target detection, specifically relating to a method for motion segmentation and motion parameter estimation of complex moving targets at sea. Background technology:

[0002] The ocean has become a major battlefield for political, economic, and military struggles among nations. The detection of small maritime targets is crucial for building a maritime power and safeguarding maritime rights. When radar observes the sea, sea clutter constitutes a major component of the received echo. The low observability of small maritime targets results in a very low signal-to-clutter ratio (SCR) in the received echo, leading to poor radar detection performance. To improve the performance of detecting weak targets against sea clutter backgrounds, research on long-term accumulation techniques has been conducted both domestically and internationally. Compared to long-term non-coherent accumulation, long-term coherent accumulation utilizes the phase information of the echo signal, effectively compensating for phase fluctuations and ensuring in-phase superposition of signal energy. This is a crucial technique for effectively improving the SCR and thus increasing the detection probability. However, during long-term radar observation of targets, the target's motion may span multiple range cells and involve complex variable acceleration motions. This motion pattern mismatches with typical motion models, leading to a deterioration in the performance of traditional long-term coherent accumulation algorithms based on MTD (Mean Transmission Delay) technology. Therefore, accurate estimation of target motion parameters and fine segmentation of motion are the focus of research on long-time coherent accumulation technology, which can provide prior information about the target motion state for long-time coherent accumulation algorithms. Summary of the Invention:

[0003] This invention addresses the problem of poor long-term coherent accumulation results due to the mismatch between the motion of maritime targets and general motion models. It provides a fine segmentation method for the motion of maritime targets and estimates the motion parameters of each segment. This invention can provide target motion state estimation for research on long-term coherent accumulation technology for maritime targets.

[0004] This invention follows the following technical solution: a method for estimating motion parameters and segmenting motion of a variable-acceleration target at sea, comprising the following steps:

[0005] (1) The radar echo of a target moving at sea with variable acceleration is divided into several CPIs in the slow time dimension.

[0006] (2) Estimate motion parameters for CPI segments.

[0007] (3) Combining the results of parameter estimation for each CPI segment, CPI segments with similar accelerations are merged into a single motion segment, and time points with large acceleration changes are used as segmentation points of the motion, thereby segmenting the complex motion of the target.

[0008] The advantages of this invention compared to the prior art are:

[0009] (1) The present invention divides the target echo observed over a long period of time into multiple CPIs and uses the instantaneous autocorrelation function spectrum to estimate the target motion parameters for the data in each short time period. Compared with directly estimating the motion parameters of the entire echo, the amount of computation is reduced.

[0010] (2) The present invention determines whether the target motion state has changed by measuring the magnitude of the change in acceleration within adjacent CPIs, and uses the time point with the larger change in acceleration as the segment point of motion, thus making the location of the time point when the target motion state changes more accurate.

[0011] (3) The present invention estimates the parameters of the target’s motion at each stage and simplifies the description of complex motion targets by segmenting the motion process. Attached image description:

[0012] Figure 1 This is a flowchart illustrating the segmentation of the movement of a maritime target according to the present invention.

[0013] Figure 2 The results are the initial velocity estimation and the true value.

[0014] Figure 3 The acceleration estimation results are compared with the true values.

[0015] Figure 4 This represents the segmented result of the target motion. Detailed implementation method:

[0016] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0017] The flowchart of the processing of sea surface echoes received by radar in this invention is as follows: Figure 1 As shown, it specifically includes the following three steps.

[0018] 1) Divide the echo signal into multiple CPIs for processing.

[0019] Let r(t) be the radial distance between the moving target at sea and the radar. m ), where t m =mT r m = 0, 1, ..., M-1 represents slow time, T r Given the pulse repetition interval, the baseband target echo after pulse compression can be expressed as (1).

[0020]

[0021] Where A r The amplitude of the signal after pulse compression is given. If the target undergoes uniform acceleration during the observation time, i.e. The echo in the slow time domain then manifests as an LFM signal whose signal parameters are related to the target's initial velocity and acceleration.

[0022] In the slow time dimension, the target echo is divided into multiple coherent processing intervals. Assuming each CPI contains N pulses, the total M pulse echoes can be divided into... Let there be _CPI_, where _round_ is the integer rounding. Then the echo signal within the i-th CPI can be simplified as:

[0023]

[0024] Where A0 is a constant, Let be the Doppler center frequency of the signal within the i-th CPI. For Doppler frequency modulation, v i-1 a i These are the initial velocity and acceleration of the motion, respectively. The duration of the first i-1 CPI.

[0025] 2) Target motion parameter estimation based on the spectrum of instantaneous autocorrelation function

[0026] For each segment with a duration of NT r The echo signal lasting one second is used to estimate the target's velocity and acceleration during this time using the instantaneous autocorrelation function spectrum. The spectrum of the instantaneous autocorrelation function is a simplified representation based on the Wigner-Ville distribution. By transforming the LFM signal with specific parameters from the time domain to the Doppler frequency domain, the energy accumulation of the LFM signal in the Doppler frequency domain is realized.

[0027] For a center frequency of f di The frequency of the modulation is k i The echo signal, at a certain time t 0i The instantaneous autocorrelation function is:

[0028]

[0029] For a given t0, F i (t0,τ) can be viewed as a single-variable function F i (τ), for F i (τ) Fourier transform yields:

[0030]

[0031] Because the duration of each CPI obtained from the segmentation is short, and the acceleration of the maritime target is small, then The effect of this term on the initial velocity estimation is negligible. Therefore, the Doppler frequency corresponding to the peak value of the transformed signal is... The estimated value of the Doppler frequency can be calculated according to (5).

[0032]

[0033] Where Δf is the Doppler window length, and based on the estimated Doppler frequency and radar wavelength, the estimated initial velocity of the target within the i-th CPI can be obtained as follows:

[0034] Based on the Doppler spectrum estimation, the acceleration can be estimated by the Doppler bandwidth.

[0035] 3) Divide the motion into segments based on the estimated motion parameters.

[0036] If the estimated acceleration change between two adjacent CPIs is significant, it can be considered that the target's motion state has changed at this point in time, and this point can be used as a motion segmentation point. Based on the estimated velocity and acceleration corresponding to each CPI in step 2), since the estimation of individual CPI sequence parameters is greatly affected by clutter, velocity and acceleration are used together as the basis for motion segmentation. Specifically, if the estimated acceleration values ​​of the i-th and (i+1)-th CPIs are... and The initial velocities of the three adjacent CPIs are v i ,v i+1 ,v i+2 When satisfied

[0037]

[0038] The i-th and (i+1)-th CPIs can be combined into a single motion segment. If the above formula is not satisfied, then the end time of the i-th CPI is determined as the segmentation point of the target motion. In the above formula, N is the number of pulses contained in a CPI segment, and T... r The time corresponding to each pulse, a T This is the threshold for the judgment.

[0039] After traversing all accelerations, the complex motion of the target over a relatively long period of time is decomposed into a combination of multiple uniformly accelerated motions.

[0040] To verify the effectiveness of the proposed method for estimating the motion parameters and segmenting the motion of a variable-acceleration target at sea, a simulation experiment was conducted. The sea clutter data used was collected in 2006 along the southwest coast of South Africa, designated CFC17-001. The target was a target undergoing uniform acceleration in two segments, with a signal-to-clutter ratio of -10 dB. The initial velocity of the first segment was 6 m / s, and the acceleration was -2 m / s². 2 The duration of the first phase is 0.5s; the acceleration of the second phase is 1m / s². 2 The duration is 0.5s.

[0041] The simulated target echo signal was divided into 10 CPIs, each with a duration of 0.1 seconds. Motion parameters for each CPI were estimated using the instantaneous autocorrelation function spectrum. The estimated initial velocities for each CPI were obtained as follows: Figure 2 As shown, the acceleration estimation results are as follows: Figure 3 As shown.

[0042] Next, using the estimated acceleration, the changes in acceleration and velocity estimates for adjacent CPIs are calculated, and a threshold 'a' is set. T 1.5m / s 2 After traversing all acceleration estimates, the segmented result of the target motion is obtained as follows: Figure 4 As shown. By Figure 4 It can be seen that the target's motion can be divided into two segments: 0-0.5 seconds and 0.5-1 second. The first segment has an initial velocity of 5.9 m / s and an acceleration of -2.05 m / s². 2 The motion is uniformly decelerated; the second segment has an initial velocity of 5.08 m / s and an acceleration of 0.7 m / s². 2 The motion is uniformly accelerated. Comparing the segmented motion results with the set simulation parameters, it can be seen that the initial velocity estimation deviation for the first segment is 0.1 m / s², and the acceleration estimation deviation is 0.05 m / s². 2 The initial velocity estimation error for the second segment of motion is 0.08 m / s, and the acceleration estimation error is 0.3 m / s². 2 The proposed method for estimating motion parameters and segmenting motion is relatively accurate. Therefore, it can effectively segment the complex motion of the target into multi-stage uniformly accelerated motion, thus simplifying the description of the target motion.

Claims

1. A method for estimating and segmenting motion parameters of a variable-acceleration target at sea, characterized in that... The steps are as follows: Step (1): Divide the target echo into multiple CPI segments in the slow time dimension; Step (2): Estimate the motion parameters for each CPI segment; In practical implementation, the peak value in the instantaneous autocorrelation function spectrum is used to estimate the initial velocity of each CPI segment. Taking advantage of the premise that the target motion is relatively stable over a short period, and ignoring minor interference factors, the instantaneous autocorrelation function spectrum is simplified to... The form of the function transforms the target velocity estimation problem into a peak search problem of the instantaneous autocorrelation function spectrum; Step (3): Segment the target motion based on the obtained estimation results of the motion parameters; In specific implementation, based on the velocity and acceleration corresponding to each CPI estimated in step (2), since the estimation of parameters of a single CPI sequence is greatly affected by clutter, velocity and acceleration are used together as the basis for motion segmentation, with the first CPI as the basis for the second CPI sequence. The and the first Taking a time point between +1 CPI points as an example, the estimated initial velocity and acceleration are as follows: , , , , No. The initial velocity of each CPI is When satisfied （1） Then the first The and the first +1 CPI is combined into a motion segment. If the above formula is not satisfied, this point is the motion segmentation point. In the above formula... This represents the number of pulses contained in a CPI segment. The time corresponding to each pulse, This is the threshold for the judgment.

Images