A joint amplitude-phase clutter suppression method for three-dimensional micro-motion measurement

Abstract

The invention discloses an amplitude-phase combined clutter suppression method suitable for three-dimensional micro-motion measurement. The method comprises the following steps: performing pulse compression processing on input radar echo data to obtain one-dimensional range image data; using the one-dimensional range image data to obtain a rough clutter estimation value, and obtaining the one-dimensional range image data after clutter removal; based on the rough clutter estimation value, taking the root mean square deviation of the phase of two adjacent range units in the one-dimensional range image data after clutter removal as a cost function for traversal optimization to obtain a precise clutter estimation value; using the precise clutter estimation value to perform clutter removal processing on the one-dimensional range image data in step one to obtain a range image result after clutter removal; and using a method based on phase inferred range and phase inferred angle to obtain a three-dimensional micro-motion measurement result of a target for the range image result after clutter removal obtained in step four; the method can realize high-precision measurement of three-dimensional micro-motion targets in a clutter environment.

Description

Technical Field

[0001] The invention belongs to the technical field of signal processing, and in particular relates to an amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement. Background Art

[0002] Faced with an increasingly complex electromagnetic environment, it is urgent to improve the detection and identification capabilities of modern radars for diverse targets. Micromotion refers to the tiny movements of a target or target component other than the translational movement of the target center, such as vibration, precession, rotation, and tumbling. The radar echo of a micromotion target contains unique motion characteristics, scattering characteristics, and structural characteristics, reflecting the fine structure and motion details of the target. By analyzing the characteristic information contained in the radar echo of a micromotion target through modern signal processing technology, it is possible to better distinguish the target attribute type and motion intention, thereby providing an important feature basis for accurate detection and precise identification of radar targets that is independent of prior information, highly reliable, and well separable.

[0003] At present, the related prior art 1 (Fan Huayu, Li Jiandong, Mao Erke, Ren Lixiang. A wideband radar target positioning method based on relative estimated range and relative estimated angle [P]. General invention patent, application number: 201910647128.X) is a three-dimensional micro-motion measurement method based on relative estimated range and relative estimated angle. However, this method does not consider the impact of clutter in the environment on micro-motion measurement. When there is clutter in the environment, the echo phase information will be seriously affected, the phase-push measurement accuracy will deteriorate sharply, and the difficulty of extracting micro-motion feature signals will increase significantly. Although broadband radar can use the advantage of high distance resolution to distinguish between clutter and targets of interest, when the clutter position is close to the target and the energy is very strong, it will not be possible to distinguish between clutter and target from the distance dimension. In addition, since the echo spectrum of the micro-motion target is mixed with the spectrum components of the stationary clutter, it cannot be distinguished from the Doppler dimension. At present, the existing technology 2 (Dei D, Grazzini G, Luzi G, et al. Non-Contact Detection of Breathing Using a Microwave Sensor [J]. Sensors, 2009, 9 (4): 2574-2585) is a clutter suppression method based on circular fitting. However, this method is mostly used for continuous wave radar. For pulse radar, since the target movement crosses the range unit, the clutter estimation accuracy of this method cannot meet the accuracy required for three-dimensional micro-motion measurement. Summary of the invention

[0004] In view of this, the present invention provides an amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement, which can overcome the defects of the existing technology and achieve high-precision measurement of three-dimensional micro-motion targets in a clutter environment.

[0005] The technical solution for implementing the present invention is as follows:

[0006] An amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement includes the following steps:

[0007] Step 1: Perform pulse compression processing on the input radar echo data to obtain one-dimensional range image data;

[0008] Step 2: using the one-dimensional range image data to obtain a rough estimate of clutter, and obtaining the one-dimensional range image data after clutter removal;

[0009] Step 3: Based on the rough clutter estimate, the root mean square deviation of the phases of two adjacent range units in the one-dimensional range image data after clutter removal is used as a cost function for traversal optimization to obtain a precise clutter estimate;

[0010] Step 4: using the clutter accurate estimation value to perform clutter removal processing on the one-dimensional range image data in step 1 to obtain a range image result after clutter removal;

[0011] Step 5: The three-dimensional micro-motion measurement result of the target is obtained by using the method based on the phase inferred range and phase inferred angle for the range image result after clutter removal obtained in step 4.

[0012] Furthermore, step 2 specifically includes:

[0013] Step 2.1: Find the peak point position of the one-dimensional range image data obtained in step 1. Assuming that the peak point of the one-dimensional range image data is located on K range units, the range unit index value corresponding to the peak point position is recorded as {a1, a2, …, a K};

[0014] Step 2.2: For each range unit index value in step 2.1, extract the complex data corresponding to the peak point in the range unit; estimate the clutter of the range unit, and perform clutter removal processing;

[0015] Step 2.3: All the range unit index values ​​in step 2.1 are processed in step 2.2, and finally the range image data after clutter removal is obtained;

[0016] Step 2.4: Find the peak point position again for the range image data after clutter removal obtained in step 2.3. Assume that the peak point of the one-dimensional range image data is located on K' range cells, and record the range cell index value corresponding to the peak point position as {b1, b2, ..., b K'};

[0017] Step 2.5: Determine the set {a1, a2, …, a K} and the set {b1,b2,…,b K'} are the same, if not, continue to repeat the above steps 2.1 to step 2.4; if the same, continue to step 3, and output the rough estimation value of clutter and the one-dimensional range image data after clutter removal.

[0018] Furthermore, in step 2.2, a clutter estimation method based on circular fitting is used to estimate the clutter of the range unit.

[0019] Furthermore, step three specifically includes:

[0020] Step 3.1: Input the complex data of two adjacent distance units in the one-dimensional range image data obtained in step 2.5, recorded as Data1(m) and Data2(m), m = 1, 2, ..., M, M means that a total of M frames of data need to be processed, and input the set traversal search clutter amplitude step Step amp 、Clutter phase step Step angle and the traversal number N, input the rough clutter estimate corresponding to the two distance units and Input the RMS deviation threshold Γ of the phases of two range units;

[0021] Step 3.2: Initialize the clutter amplitude space and clutter phase space, which are calculated by List amp ={0,±Step amp ,±2Step amp ,…,±NStep amp}、List angle ={0,±Step angle ,±2Step angle ,…,±NStep angle}; Based on the rough estimate of clutter and Perform noise removal on the data Data1(m) and Data2(m), and calculate the initial value Cost0 of the cost function according to formula (1);

[0022] The cost function is expressed as:

[0023]

[0024] Wherein, angle(·) represents the phase operation;

[0025] Step 3.3: Traverse the clutter amplitude space List amp and clutter phase space List angle All elements of Follow these steps:

[0026] (a) Initialize the clutter value, that is,

[0027] (b) performing noise removal processing on the data Data1(m) and Data2(m) according to the noise values ​​C1 and C2;

[0028] (c) Calculate the cost function Cost according to formula (1);

[0029] (d) Compare Cost and Cost0. If Cost < Cost0, update Cost0 and record the clutter estimation value at this time, i.e. Cost0 = Cost, C1' = C1, C2' = C2; otherwise, no update is required.

[0030] Cost0 value;

[0031] Step 3.4: If you have traversed the clutter amplitude space List amp and clutter phase space List angle If all elements of , or the cost function is less than the threshold value Cost0<Γ, then step (3.3) traversal ends; otherwise, continue with step 3.3 traversal;

[0032] Step 3.5: Output the clutter accurate estimation values ​​C1' and C2' of the two range cells.

[0033] Furthermore, the RMS deviation threshold Γ is set according to the actual required estimation accuracy, and is generally set to 10 - 4 rad.

[0034] Beneficial effects:

[0035] (1) The present invention estimates the clutter in the corresponding range units for each of the multiple range units and continuously corrects the peak point position of the one-dimensional range image through iteration, thereby effectively suppressing the influence of clutter on the peak point position of the range image and restoring the peak point position of the range image in the absence of clutter.

[0036] (2) The present invention can obtain a high-precision clutter estimation value by using the root mean square deviation of the phases of two adjacent range units as a cost function for ergodic optimization.

[0037] (3) The present invention can obtain high-precision clutter estimation values ​​through the clutter coarse estimation algorithm and the clutter fine estimation algorithm, thereby meeting the accuracy required for three-dimensional micro-motion measurement.

[0038] (4) The present invention can obtain high-precision micro-motion measurement results of three-dimensional micro-motion targets in a clutter environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The figure is a flow chart of the method of the present invention.

[0040] Figure 2Flowchart for implementing the rough clutter estimation process.

[0041] Figure 3 Flowchart for clutter accurate estimation processing implementation.

[0042] Figure 4 is the micro-motion measurement result including clutter.

[0043] Figure 5 This is the micro-motion measurement result obtained by using the prior art 2.

[0044] Figure 6 This is the micro-motion measurement result obtained by the method of the present invention. DETAILED DESCRIPTION

[0045] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0046] like Figure 1 As shown, the present invention provides an amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement.

[0047] Firstly, the echo data received by the radar is pulse compressed to obtain a one-dimensional range image. Then, the clutter in the corresponding range unit is estimated for multiple range units respectively, and the peak point position of the one-dimensional range image is continuously corrected through iteration to obtain a rough estimate of the clutter. Then, based on the rough estimate of the clutter, the root mean square deviation of the phase of two adjacent range units is used as the cost function for traversal optimization to obtain a precise estimate of the clutter. Finally, the three-dimensional micro-motion measurement result of the target is obtained by using the method based on phase inference range and phase inference angle for the range image result after clutter removal.

[0048] It should be pointed out that the main innovation of the method proposed in the present invention is the amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement. Therefore, the implementation steps of the clutter estimation algorithm are introduced in detail (i.e., step (2) and step (3)). After obtaining the clutter estimation value, the range image result can be de-cluttered, and then the method in the prior art 1 can be used to complete the final three-dimensional micro-motion measurement processing.

[0049] The method proposed by the present invention specifically comprises the following steps:

[0050] Step (1)

[0051] Perform pulse compression processing on the input radar echo data to obtain a one-dimensional range image result;

[0052] This step completes the preprocessing process of the echo data received by the radar and obtains the one-dimensional range image result. Subsequent related processing is carried out on this basis.

[0053] Step (2)

[0054] Combination Figure 2 , process the one-dimensional range image data in step (1) in the following steps to obtain a rough estimate of the clutter:

[0055] Step (2.1): Find the peak point position of the one-dimensional range image data obtained in step (1). Assuming that the peak point of the one-dimensional range image data is located on K range units, the range unit index value corresponding to the peak point position is recorded as {a1, a2, …, a K};

[0056] Step (2.2): For each distance unit index value in step (2.1), extract the complex data corresponding to the peak point in the distance unit. Use a clutter estimation method based on circular fitting to estimate the clutter of the distance unit and perform clutter removal processing, wherein the clutter estimation method based on circular fitting can refer to prior art 2;

[0057] Step (2.3): All the range unit index values ​​in step (2.1) are processed in step (2.2), and finally the range image data after clutter removal is obtained;

[0058] Step (2.4): Find the peak point position again for the range image data after clutter removal obtained in step (2.3). Assume that the peak point of the one-dimensional range image data is located on K' range cells, and record the range cell index value corresponding to the peak point position as {b1, b2, ..., b K'};

[0059] Step (2.5): Determine the set {a1, a2, …, a K} and the set {b1,b2,…,b K'} are the same. If not, continue to repeat the above steps (2.1) to (2.4); if the same, continue to step (3) and output the rough estimation value of clutter and the one-dimensional range image data after clutter removal.

[0060] This step realizes the rough estimation of clutter. This step provides the rough estimation value of clutter for the subsequent step (3) of accurate clutter estimation, effectively reducing the time and space resource consumption of the traversal search in step (3); this step can also obtain the peak point position of the range image in the absence of clutter, which is the basis for the subsequent three-dimensional micro-motion measurement.

[0061] Step (3)

[0062] Based on the rough clutter estimate obtained in step (2), the root mean square deviation of the phases of two adjacent range units is used as the cost function for traversal optimization to obtain the precise clutter estimate. Figure 3 , follow the steps below to get the accurate estimate of clutter:

[0063] Step (3.1) - Input: Input the complex data of two adjacent range units in the one-dimensional range image data obtained in step (2.5), recorded as Data1(m) and Data2(m) (m = 1, 2, ..., M, M means that a total of M frames of data need to be processed), which represent the complex data of two range units respectively, and input the clutter amplitude step length Step of the traversal search amp 、Clutter phase step Step angle and the traversal number N, input the rough clutter estimate corresponding to the two distance units obtained in step (2.5) and Enter the RMS deviation threshold Γ of the phase of two range units (this threshold value can be set according to the actual estimation accuracy required, generally set to 10 - 4 rad);

[0064] Step (3.2) - Initialization: Initialize the clutter amplitude space and clutter phase space, which are calculated by List amp ={0,±Step amp ,±2Step amp ,…,±NStep amp}、List angle ={0,±Step angle ,±2Step angle ,…,±NStep angle According to the rough estimate of clutter obtained in step (2.5) and Perform noise removal on the data Data1(m) and Data2(m), and then calculate the initial value Cost0 of the cost function according to formula (1);

[0065] The cost function can be expressed as:

[0066]

[0067] Here, angle(·) represents a phase operation.

[0068] Step (3.3) - Traverse: Traverse the clutter amplitude space List amp and clutter phase space List angle All elements of Follow these steps:

[0069] (a) Initialize the clutter value, that is,

[0070] (b) De-cluttering the data Data1(m) and Data2(m) according to the clutter values ​​C1 and C2.

[0071] (c) Calculate the cost function Cost according to formula (1).

[0072] (d) Compare Cost and Cost0. If Cost < Cost0, update Cost0 and record the clutter estimation value at this time, that is, Cost0 = Cost, C1' = C1, C2' = C2; otherwise, do not update Cost0.

[0073] Step (3.4) - Judgment: If the clutter amplitude space List is traversed amp and clutter phase space List angle If all elements of , or the cost function is less than the threshold value Cost0<Γ, then step (3.3) traversal ends; otherwise, continue step (3.3) traversal;

[0074] Step (3.5) - Output: accurate clutter estimates C1' and C2' of two range cells;

[0075] This step realizes the precise estimation of clutter. Based on the rough estimation of clutter obtained in step (2), this step obtains the precise estimation of clutter by traversal search, thereby realizing the subsequent clutter removal and three-dimensional micro-motion measurement.

[0076] Step (4)

[0077] Using the clutter accurate estimation value obtained in step (3) to perform clutter removal processing on the range image data in step (1) to obtain a range image result after clutter removal;

[0078] Step (5)

[0079] The three-dimensional micro-motion measurement result of the target is obtained by using a method based on phase inference range and phase inference angle for the range image data after clutter removal obtained in step (4). The detailed processing process of this step can refer to the prior art 1.

[0080] Since then, the three-dimensional micro-motion measurement processing in a clutter environment has been completed, and high-precision three-dimensional micro-motion measurement results have been obtained.

[0081] In order to verify the effectiveness of the method provided by the present invention, a microwave darkroom experiment was conducted. The radar is an S-band synthetic broadband radar, and the transmitted signal is a simple frequency step signal. The experiment uses a one-transmit and four-receive antenna array, and the distance between the transmitting and receiving antennas is about 0.5m. The experimental target is a precision three-dimensional displacement platform with a fixed angular reflection. The target moves cyclically along a square trajectory with a side length of 50mm on the horizontal plane, at a speed of 20mm / s, and there is no movement in the vertical direction.

[0082] Figure 4 The three-dimensional micro-motion measurement results without clutter removal are given. Figure 4It can be seen that due to the influence of clutter, the micro-motion trajectory is severely distorted and stretched, and the measurement results deviate from the calibration value of the total station. Figure 5 The three-dimensional micro-motion measurement results after estimating the clutter value and removing the clutter using the existing technology 2 method are given. Since the clutter estimation accuracy of the existing technology 2 method is low, the micro-motion measurement results of this method are still greatly distorted and deviate from the calibration value of the total station. Figure 6 The three-dimensional micro-motion measurement results after the clutter value estimation and clutter removal process using the method proposed in the present invention are given, where the fitting value is the micro-motion trajectory obtained by fitting the calibration values ​​of the four corners of the square measured by the total station. Figure 6 It can be seen that after the high-precision clutter estimation value obtained by the clutter estimation algorithm proposed in the present invention is denoised, the micro-motion trajectory measurement value and the fitting value are basically consistent, which verifies the effectiveness of the method proposed in the present invention.

[0083] In summary, the above are only preferred embodiments of the present invention and are not intended to limit the protection scope of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An amplitude-phase joint clutter suppression method suitable for three-dimensional micro-motion measurement, characterized in that: It includes the following steps: Step 1: Perform pulse compression processing on the input radar echo data to obtain one-dimensional range profile data; Step 2: Obtain a rough clutter estimate value using the one-dimensional range profile data, and obtain the one-dimensional range profile data after clutter suppression processing; Step 2 specifically includes: Step 2.1: Find the peak point position of the one-dimensional range image data obtained in step 1. Assuming that the peak point of the one-dimensional range image data is located on K range units, the range unit index value corresponding to the peak point position is recorded as {a1, a2, …, a K }; Step 2.2: For each range cell index value in Step 2.1, extract the complex data whose peak point is located in the corresponding range cell; estimate the clutter of this range cell, and perform clutter suppression processing; Step 2.3: Perform the processing of Step 2.2 for all range cell index values in Step 2.1, and finally obtain the range profile data after clutter suppression processing; Step 2.4: Find the peak point position again for the range image data after clutter removal obtained in step 2.

3. Assume that the peak point of the one-dimensional range image data is located on K' range cells, and record the range cell index value corresponding to the peak point position as {b1, b2, ..., b K' }; Step 2.5: Determine the set {a1, a2, …, a K } and the set {b1,b2,…,b K' } are the same, if not, continue to repeat the above steps 2.1 to 2.4; if the same, continue to step 3, and output the rough estimation value of clutter and the one-dimensional range image data after clutter removal; Step 3: Based on the rough clutter estimate value, use the root mean square deviation of the phases of two adjacent range cells in the one-dimensional range profile data after clutter suppression processing as a cost function to perform traversal optimization, and obtain a fine clutter estimate value; Step 4: Use the fine clutter estimate value to perform clutter suppression processing on the one-dimensional range profile data in Step 1 to obtain the range profile result after clutter suppression; Step 5: Use the method based on phase speculation ranging and phase speculation angle to obtain the three-dimensional micro-motion measurement result of the target for the range profile result after clutter suppression obtained in Step 4.

2. The amplitude-phase combined clutter suppression method for three-dimensional micro-motion measurement according to claim 1, characterized in that: In Step 2.2, a clutter estimation method based on circular fitting is used to estimate the clutter of the range cell.

3. The amplitude-phase combined clutter suppression method for three-dimensional micro-motion measurement according to claim 1 or 2, characterized in that: Step 3 specifically includes: Step 3.1: Input the complex data of two adjacent distance units in the one-dimensional range image data obtained in step 2.5, recorded as Data1(m) and Data2(m), m = 1, 2, ..., M, M means that a total of M frames of data need to be processed, and input the set traversal search clutter amplitude step Step amp 、Clutter phase step Step angle and the traversal number N, input the rough clutter estimate corresponding to the two distance units and Input the RMS deviation threshold Γ of the phases of two range units; Step 3.2: Initialize the clutter amplitude space and clutter phase space, which are calculated by List amp ={0,±Step amp ,±2Step amp ,…,±NStep amp }、 List angle ={0,±Step angle ,±2Step angle ,…,±NStep angle }; Based on the rough estimate of clutter and Perform noise removal on the data Data1(m) and Data2(m), and calculate the initial value Cost0 of the cost function according to formula (1); The cost function is expressed as: where angle(·) represents the phase calculation operation; Step 3.3: Traverse the clutter amplitude space List amp and clutter phase space List angle All elements of Follow these steps: (a) Initialize the clutter value, that is, (b) Perform clutter suppression processing on the data Data1(m) and Data2(m) according to the clutter values C1 and C2; (c) Calculate the cost function Cost according to Equation (1); (d) Compare the magnitudes of Cost and Cost0. If Cost < Cost0, update the Cost0 value and record the clutter estimate value at this time, that is, Cost0 = Cost, C1' = C1, C2' = C2; otherwise, do not update the Cost0 value; Step 3.4: If you have traversed the clutter amplitude space List amp and clutter phase space List angle If all elements of , or the cost function is less than the threshold value Cost0<Γ, then step (3.3) traversal ends; otherwise, continue with step 3.3 traversal; Step 3.5: Output the fine clutter estimate values C1' and C2' of the two range cells.

4. The amplitude-phase combined clutter suppression method for three-dimensional micro-motion measurement according to claim 3, characterized in that: The RMS deviation threshold Γ is set according to the actual estimation accuracy required, and is generally set to 10 -4 rad.

Images