A method, system, and medium for non-line-of-sight target positioning based on millimeter wave radar

Abstract

The application relates to a millimeter wave radar-based non-line-of-sight target positioning method, system and medium, which comprises the following steps: establishing a multipath propagation model containing LOS and NLOS target echoes; according to the extracted target distance information, performing spatial filtering on the distance units using a minimum mean square distortion-free response to obtain corresponding angle information; combining the obtained distance information and angle information to separate the LOS and NLOS targets, and then cleaning up the redundant NLOS paths based on a path matching strategy; and calculating the position of the real target according to the geometric symmetry between the real target and its false points. The application fully utilizes the difference between the LOS and NLOS echoes in the geometric propagation path based on the multipath propagation characteristics of electromagnetic waves, and realizes the discrimination and high-precision positioning of the two types of targets.

Description

Technical Field

[0001] This application relates to the field of target localization, and in particular to a non-line-of-sight target localization method, system and medium based on millimeter-wave radar. Background Technology

[0002] The development of non-line-of-sight (NLOS) target detection technology has greatly enhanced the detection range of radar systems, enabling them to monitor areas outside the direct line-of-sight range. However, current NLOS target detection technologies mainly focus on environments involving only NLOS targets, without addressing more complex situations where both line-of-sight (LOS) and NLOS targets are present. Summary of the Invention

[0003] The purpose of this application is to provide a non-line-of-sight target localization method, system, and medium based on millimeter-wave radar. By fully utilizing the differences in distance and angle information between LOS target and NLOS target echoes, the LOS echo and NLOS echo are separated, and the LOS and NLOS targets are located.

[0004] To achieve the above objectives, this application provides the following technical solution:

[0005] In a first aspect, embodiments of this application provide a non-line-of-sight target localization method based on millimeter-wave radar, the method comprising the following steps:

[0006] Step 1: Analyze the multipath signal and establish a multipath propagation model that includes LOS and NLOS target echoes;

[0007] Step 2: Obtain the range profiles of the multi-channel target using Fast Fourier Transform (FFT), construct the accumulated range profiles using an incoherent accumulation method, and build the cumulative range profiles using an incoherent addition accumulation method based on these accumulated range profiles. Based on the cumulative range profiles, extract the target range information using one-dimensional cell mean constant false alarm rate (CFAR). Based on the extracted target range information, perform spatial filtering on these range cells using minimum mean square distortion-free response (MINSD) to obtain the corresponding angle information.

[0008] Step 3: The obtained distance and angle information are combined to separate LOS and NLOS targets, and then redundant NLOS paths are cleaned up based on the path matching strategy.

[0009] Step 4: Calculate the position of the real target based on the geometric symmetry between the real target and its false points.

[0010] The establishment of a multipath propagation model that includes LOS and NLOS target echoes includes constructing a hybrid scenario of LOS and NLOS targets consisting of two buildings.

[0011] The scene structure mainly consists of wall-1, wall-2, and wall-3. The radar is located at the origin O of the coordinate axis, and the corners of the two buildings are C and D respectively. The distance from wall-2 to... The distance between the axes is Wall-3 to The distance between the axes is The extension of line OC intersects wall-3 at point E, and OE is... The included angle of the axis is OD and The included angle of the axis is ,

[0012] Assume there are two types of targets in the scene: LOS and NLOS. The LOS target is located at point P. NLOS target Q Hidden at corner C ( )back.

[0013] The extraction of target distance information in step 2 specifically includes the following steps:

[0014] Using a MIMO radar with M receiving antennas, the echo signal received by the m-th receiving antenna is:

[0015] ,

[0016] in, For the first Chirp cycles The echo signal;

[0017] To reduce environmental noise interference while preserving micro-Doppler information, the phasor mean cancellation method is used to suppress environmental noise. The processed echo signal is then subjected to FFT to obtain the range profile.

[0018] ,

[0019] in, For the first The first receiving antenna received the first The first group of echo signals A distance image;

[0020] To ensure the accuracy of target ranging, a non-coherent accumulation method is used to accumulate the range images between channels. The accumulated range image is represented as follows:

[0021] ,

[0022] in, The th after accumulation The first receiver The first cycle A distance image, which can be represented as

[0023] ,

[0024] For the accumulated range image, the accurate target range is obtained using the one-dimensional cell mean constant false alarm rate 1-D CA-CFAR method. The detection threshold of the one-dimensional unit mean constant false alarm rate method. for

[0025] ,

[0026] in, This represents the probability of a false alarm. For the unit being detected The number of surrounding reference cells, for the processed distance cells, can be expressed as:

[0027] ,

[0028] Similarly, repeat the above detection process for all distance cells to obtain the target distance peak value of the accumulated distance image.

[0029] Step 3 is implemented in the following steps:

[0030] The NLOS target echo originates from the reflection of the EM in the ED segment of wall-3; therefore, the angle of the false target formed by the NLOS target echo is... Concentrated at the angle between OD and OE, i.e.

[0031] ,

[0032] Initially, LOS and NLOS targets are screened in the angle domain. The least mean square distortion-free response (MVDR) algorithm is used to estimate the target angles. The weight vector can be expressed as

[0033] ,in, The autocorrelation matrix is... For angle The steering vector, therefore, the power at different angles can be expressed as...

[0034] ,

[0035] Similar calculations are performed on each detected distance peak to obtain the distance and angle information of the targets corresponding to all distance peaks. This yields the coordinates of all false targets and LOS targets, allowing for the filtering of targets with angles in between. The target locations between these points, namely all false targets and some LOS target locations, and Since the targets are all LOS targets, the exact location of the LOS targets can be obtained directly;

[0036] Next, for the mixed range of LOS and NLOS echo angles, the path distance range of the NLOS target echo received by the radar receiver has a minimum and a maximum value. Let the minimum distance of the two-way first-order path be denoted as... The maximum value is Its distance range can be given by the following formula.

[0037] ,

[0038] use As the boundary line separating LOS and NLOS targets, the set of all points within the mixed region can be represented as:

[0039] ,

[0040] in, Points representing the mixed intervals, Indicates its path length. Indicates its relationship with The included angle of the axis is checked for all points within the set. and Compare, if there is If it is not, it is considered a LOS target; otherwise, it is considered an NLOS target.

[0041] The implementation of step 4 can be divided into the following steps:

[0042] The separated NLOS echo targets are sorted in ascending order of their path lengths and represented as a set of paths. Based on geometric relationships, it can be found that The path is the shortest path, therefore the set of paths... In terms of, among them Must be a certain NLOS target Path echo, thus combining This allows us to determine the target's exact location within the NLOS region; further analysis is then performed based on the geometric relationships of the path. Thus obtain The specific length of the path, once the exact location of the target is determined, and the false points formed by its second-order two-way path. The location is thus determined, and the path length is also determined. For the mixed path of this target, there are

[0043] For each element in the path set, the paths are cleaned up using a matching method based on the calculated path length, with a threshold specified here. Its value is the distance resolution, used to measure the allowable error between the calculated path length and the actual path length in the set. Specifically, it can be expressed as...

[0044] ,

[0045] The path that is successfully matched according to the above formula will be regarded as a derivative path of the first-order two-way path to locate the target, and will be removed from the path set along with the main path. This process will then be repeated until the number of remaining paths in the path set is insufficient or the energy of the remaining paths is negligible.

[0046] Secondly, embodiments of this application provide a non-line-of-sight target localization system based on millimeter-wave radar, including a memory and a processor. The memory includes a program for a non-line-of-sight target localization method based on millimeter-wave radar. When the program for a non-line-of-sight target localization method based on millimeter-wave radar is executed by the processor, it implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described above.

[0047] Thirdly, embodiments of this application provide a computer-readable storage medium storing program code, which, when executed by a processor, implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described above.

[0048] Compared with existing technologies, the advantages of this invention are: based on the multipath propagation characteristics of electromagnetic waves, it fully utilizes the differences in geometric propagation paths between LOS and NLOS echoes to achieve the discrimination and high-precision positioning of two types of targets. This application first performs static clutter suppression on the radar echoes to separate dynamic and static targets; then, it constructs a range image vector through one-dimensional FFT and non-coherent accumulation, and extracts target range information using CA-CFAR detection technology; further, it uses the MVDR method to estimate the azimuth angle of each target; based on this, it achieves target category separation according to the differences in range and angle characteristics between LOS and NLOS echoes; finally, it clears multipath interference through a path matching strategy and achieves precise target positioning based on the first-order image reflection principle. Experimental verification under different target configurations shows that this method exhibits good performance in the separation and positioning of LOS and NLOS targets. Attached Figure Description

[0049] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 This is a scene diagram of LOS and NLOS targets provided in an embodiment of the present invention.

[0051] Figure 2 This is a flowchart of the method provided in an embodiment of the present invention.

[0052] Figure 3 This is a diagram of the radar experimental platform and actual testing environment provided in the embodiments of the present invention.

[0053] Figure 4 This is a comparison chart of the results of Experiment 1 provided in the embodiment of the present invention.

[0054] Figure 5 This is a comparative analysis chart of the results of Experiment 1 provided in the embodiments of the present invention.

[0055] Figure 6 This is a comparison chart of the results of Experiment 2 provided in the embodiment of the present invention.

[0056] Figure 7 This is a comparison chart of the results of Experiment 3 provided in the embodiment of the present invention.

[0057] Figure 8 This is a comparison chart of the results of Experiment 4 provided in the embodiment of the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of this application will now be described with reference to the accompanying drawings. It should be noted that similar reference numerals and letters in the following drawings indicate similar items; therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0059] The terms “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0060] The terms “first,” “second,” etc., are used only to distinguish one entity or operation from another, and should not be construed as indicating or implying relative importance, nor as requiring or implying any such actual relationship or order between these entities or operations.

[0061] This embodiment is achieved through the following technical solution: a non-line-of-sight target localization method based on millimeter-wave radar, comprising the following steps:

[0062] Consider a mixed scenario consisting of two buildings with LOS and NLOS targets, such as... Figure 1 As shown. The scene structure mainly consists of wall-1, wall-2, and wall-3. The radar is placed at the origin O of the coordinate axis, and the corners of the two buildings are C and D respectively. From wall-2 to... The distance between the axes is Wall-3 to The distance between the axes is The extension of line OC intersects wall-3 at point E, and OE is... The included angle of the axis is OD and The included angle of the axis is .

[0063] Assume there are two types of targets in the scene: LOS and NLOS. The LOS target is located at point P. NLOS target Q Hidden at corner C ( After that, LOS targets can be located using a direct path between them and the radar, while NLOS targets do not have a direct path between them and the radar, causing traditional location methods to fail. To detect this type of NLOS target, it is necessary to utilize the multipath propagation characteristics of electromagnetic waves. Since the power of the radar EM will be greatly attenuated after diffraction at the corner and multiple reflections from the wall, this application ignores the diffraction path and higher-order reflection path, and mainly considers the first-order reflection path, second-order reflection path, and their combination path of electromagnetic waves.

[0064] According to the ray tracing model, there are two main unidirectional propagation paths for electromagnetic waves from the radar transmitter to the NLOS target: 1) First-order path: the signal undergoes one reflection at wall-3 to reach target Q; 2) Second-order path: the signal undergoes two reflections between wall-2 and wall-3 to reach target Q. This application defines the transmission path as... ,in These represent the first-order path and the second-order path, respectively. Due to the reflection and backscattering characteristics of an EM, the receiving path also originates from the first-order and second-order paths, which are defined as follows: ,in Furthermore, this application defines the NLOS target echo received by the receiver as... Its path length is expressed as Therefore, based on the geometric relationships between multiple paths, we can conclude... ,and .

[0065] Without loss of generality, assume the NLOS target point The virtual target positions propagated via two-way first-order path and two-way second-order path are respectively and According to the law of specular reflection, it can be calculated that

[0066]

[0067] As can be seen from the above analysis, the echo signal received by the radar includes LOS signal, NLOS signal, and clutter. This application models the echo signal as follows:

[0068]

[0069] in, These are LOS echo and NLOS echo, respectively. This is clutter. This application uses frequency modulated continuous wave (FMCW) millimeter-wave radar to detect targets and transmit signals. It can be represented as

[0070]

[0071] in, Where μ is the carrier frequency and μ is the frequency modulation slope. ; For a rectangular function, it can be represented as

[0072]

[0073] Where T is the chirp duration.

[0074] Suppose there are p LOS targets, denoted as p. There are q NLOS objectives, denoted as . Since there is a direct path between the LOS target and the radar, the LOS echo can be described as follows:

[0075]

[0076] in, The time delay of the p-th LOS target echo; while the NLOS target signal travels through the combined path described above. Therefore, the q-th NLOS target echo can be expressed as...

[0077]

[0078] in, For path The corresponding reflection loss, This refers to the NLOS target echo delay. As analyzed above, and It can be represented as

[0079]

[0080]

[0081] in, Represents the speed of light. For the first The corresponding NLOS target The length of the path.

[0082] This application proposes to use the SLN-ML algorithm to separate LOS targets and NLOS targets from the echo signal model (1), thereby achieving multi-target localization that combines LOS and NLOS. The overall idea and block diagram of the algorithm proposed in this application are as follows: Figure 2 As shown, the main steps of the algorithm can be divided into three parts: echo preprocessing, separation of LOS and NLOS targets, and localization of NLOS targets.

[0083] Signal preprocessing

[0084] In the preprocessing stage, this application employs a MIMO radar with M receiving antennas, where the echo signal received by the m-th receiving antenna is...

[0085]

[0086] in, For the first Chirp cycles The echo signal.

[0087] To reduce environmental noise interference while preserving micro-Doppler information, this application first employs the phasor mean cancellation method to suppress environmental noise. Then, the processed echo signal is subjected to FFT to obtain the range profile, i.e.

[0088]

[0089] in, For the first The first receiving antenna received the first The first group of echo signals A distance image.

[0090] Next, to ensure the accuracy of target ranging, this application employs a non-coherent accumulation method to accumulate the range images between channels. The accumulated range image can be represented as follows:

[0091]

[0092] in, The th after accumulation The first receiver The first cycle A distance image, which can be represented as

[0093]

[0094] Finally, for the accumulated range image, this application uses the 1-D CA-CFAR method to obtain the accurate range of the target. For the first... The detection threshold of the CA-CFAR method for each unit. for

[0095]

[0096] in, This represents the probability of a false alarm. For the unit being detected The number of surrounding reference cells, for the processed distance cells, can be expressed as:

[0097]

[0098] Similarly, by repeating the above detection process for all distance cells, the peak value of the target distance in the accumulated distance image can be obtained.

[0099] LOS and NLOS target separation

[0100] The echo signal of a LOS target propagates via a direct path and can originate from any direction in the angular domain, exhibiting a clear geometric perspective relationship with the target's true location. However, the signal propagation of an NLOS target relies on an indirect path formed by wall reflection or diffraction, causing its angular domain characteristics to deviate from its true spatial orientation and originate from a specific direction. In this model, the NLOS target echo originates from the reflection of the EM at the ED segment of wall-3; therefore, the angle of the false target formed by the NLOS target echo varies. Concentrated at the angle between OD and OE, i.e.

[0101]

[0102] For the reasons mentioned above, this application first considers preliminary screening of LOS and NLOS targets from the angle domain. This application uses the MVDR algorithm to estimate the target angle. The weight vector can be expressed as

[0103]

[0104] in, The autocorrelation matrix is... For angle The steering vector. Therefore, the power at different angles can be expressed as...

[0105]

[0106] Similarly, a similar calculation is performed on the 1D-FFT slice of each detected distance peak to obtain the distance and angle information of the targets corresponding to all distance peaks, thus obtaining the coordinate positions of all false targets and LOS targets. Then, targets with angles in the range of... The target locations between these points, namely all false targets and some LOS target locations, and Since the target is always a LOS target, the exact location of the LOS target can be obtained directly.

[0107] Next, for the mixed range of LOS and NLOS echo angles, the path distance range of the NLOS target echo received by the radar receiver has a minimum and a maximum value, which can be calculated. Let the minimum distance of the two-way first-order path be... The maximum value is Its distance range can be given by the following formula.

[0108]

[0109] The line-of-sight distance of LOS targets within this area The range is between [OD, OE], and the two only... Sometimes they have a clear inclusion relationship, so they cannot be simply distinguished by the range of their distance domains. Suppose there are two targets in the scene, one a false point generated by an NLOS target, and the other a LOS target. If their echo angles are the same and both are... At that time, there is

[0110]

[0111] Therefore, it can be adopted The distance serves as the dividing line between LOS and NLOS targets. The set of all points within the mixed region can be represented as...

[0112]

[0113] in, Points representing the mixed intervals, Indicates its path length. Indicates its relationship with The included angle of the axis is checked for all points within the set. and Compare, if there is If it is not, it is considered a LOS target; otherwise, it is considered an NLOS target.

[0114] NLOS target localization

[0115] The separated NLOS echo targets are sorted in ascending order of their path lengths and represented as a set of paths. Based on geometric relationships, it can be found that... The path is the shortest path. Therefore, for the set of paths... In terms of, among them Must be a certain NLOS target Path echo, thus combining This allows us to determine the target's exact location within the NLOS region; further analysis is then performed based on the geometric relationships of the path. Thus obtain The specific length. Once the exact location of the target is determined, the false points formed by its second-order two-way path. The location is thus determined, and the path length is also determined. For the mixed path of this target, there are...

[0116]

[0117] For each element in the path set, paths are cleaned up using a matching method based on the calculated path length. A threshold is specified here. Its value is the distance resolution, used to measure the allowable error between the calculated path length and the actual path length in the set. Specifically, it can be expressed as...

[0118]

[0119] The path that is successfully matched according to the above formula will be regarded as a derived path for locating the target using the first-order two-way path, and will be removed from the path set along with the main path. This process is then repeated until the number of remaining paths in the path set is insufficient or the energy of the remaining paths is negligible.

[0120] Experimental Results and Analysis

[0121] To verify the performance of the proposed algorithm, this section will conduct multiple sets of experimental tests to demonstrate the effectiveness of the proposed target localization algorithm that separates NLOS and LOS. In the experiments, the radar module used in this application is the TI millimeter-wave radar IWR6843-ISK, which has 3 transmit channels and 4 receive channels. The frequency coverage range of the antenna transmission is 60 to 64 GHz, with a maximum continuous bandwidth of 4 GHz. This application adopts a time-division multiplexing approach, that is, the transmit antenna alternately transmits linear frequency modulated signals to detect the target scene. Detailed parameters are shown in Table 1.

[0122] Table 1. Millimeter-wave radar parameters

[0123]

[0124] The radar experimental platform and actual testing environment of this application are as follows: Figure 3 As shown in (a). In the experimental scenario, the radar is located at the origin (0,0)m, and the main reflector wall is wall-3, perpendicular to... Axis and located Location; secondary reflective wall-2, also perpendicular to Axis and located At; parallel to wall-1 Axis and located From. Figure 3 (a) It can be seen that the NLOS target is invisible to the radar. The effectiveness of the proposed algorithm will be verified by five sets of field experiments with different numbers of targets.

[0125] Case 1 (One stationary NLOS target, one stationary LOS target): Consider two human targets, Target-1 and Target-2, located at coordinates (3.17, 3.45) m and (4.58, -0.2) m respectively. The experimental scenario is as follows. Figure 3 As shown in (b), Target-1 is the NLOS target, and Target-2 is the LOS target. Throughout the data acquisition process, it was assumed that the subjects experienced slight body swaying.

[0126] The algorithm in this application first performs static clutter suppression on the radar echo to separate dynamic and static target components; then it constructs a range profile through one-dimensional FFT transformation and incoherent accumulation, and uses a CFAR detector to achieve target identification. Figure 4(a) shows the accumulated range profile. Based on this accumulated range profile, the two-dimensional position of the target can be further estimated. The localization results of the first experiment are as follows: Figure 4 As shown in (b), both targets are clearly presented. The average positioning result of NLOS target Target-1 is (3.13 m, 3.42 m), with a deviation of approximately 0.28 m from the set coordinates. Considering the actual size of the human target and the natural shaking during the acquisition process, this accuracy is acceptable in practical applications. It is worth noting that due to the weak NLOS path echo signal, the higher-order paths of the NLOS target are not visible in the range image at certain times, such as... Figure 4 As shown in (c). However, even in this case, a relatively accurate target location can still be obtained by performing localization calculations on the accumulated distance image of that frame separately, such as... Figure 4 As shown in (d), the experimental results show that the proposed algorithm still has good robustness when the echo is weak.

[0127] Figure 5 (a) presents the estimation results of the traditional NLOS localization method based on the first-order path. As shown in 5(a), although the localization results for NLOS targets are relatively close to the actual locations, the estimation for LOS targets has a significant deviation. Not only are false points appearing, but the average location of the localized points also deviates from the actual target by approximately 0.6 m. Furthermore, we found that the greater the distance between the LOS target and Wall-3, the greater the deviation in the estimated location.

[0128] The corresponding distance image profile (e.g.) Figure 5 (b) further verifies this bias. This experiment shows that the proposed algorithm can effectively distinguish static targets under LOS and NLOS conditions and achieve high-precision localization.

[0129] Case 2 (A Moving NLOS Target): This experiment sets up a moving NLOS target, Target-1, whose trajectory is a straight line parallel to the coordinate axis from the initial position (2.35, 2.2)m to the endpoint (2.35, 3.5)m. The experimental scene layout is as follows. Figure 6 As shown in (a). Unlike the previous experiment, this group of experiments did not include LOS targets, aiming to evaluate the algorithm's localization performance for NLOS targets in the absence of LOS target interference.

[0130] Figure 6 (b) shows the positioning results of the target moving along a straight line. (From...) Figure 6 (b) It can be seen that the two-dimensional positioning results clearly indicate its motion trajectory. Figure 6 (c) The distance image of this set of experiments is given, from Figure 6(c) It can be seen that, due to the absence of interference from the LOS target, the two-way first-order path, two-way second-order path, and combined path in the accumulated distance image are all clearer and more significant. Furthermore, Figure 6 (d) The corresponding distance profile for this experiment is given. Since the target moves along a straight line and the displacement range is limited, the distance profile curve is relatively smooth and the variation range is small, which is consistent with the actual situation.

[0131] Case 3 (two NLOS moving targets and one LOS moving target located in the angular mixing range): The third group of experiments set up three moving targets. Figure 3 (d) is the experimental scenario diagram of Case 3 multi-target. In this experiment, the LOS target will move within the angle mixing range. The starting points of the three moving targets, Target-1, Target-2 and Target-3, are located at (1.9,2.2)m, (3.16,3.3)m and (1.55,0.13)m, respectively.

[0132] Figure 7 (a) shows the final target localization result obtained by the method of this application. From Figure 7 As shown in (a), the NLOS location points lie on two different straight lines. For ease of comparison, the actual movement paths of the two NLOS targets are marked. Furthermore, due to clutter and interference, a few location results show errors, but most are relatively accurate. Figure 7 (b) The accumulated range profile is shown. Because the LOS target obstructs the echo path of the NLOS target to some extent, the NLOS target echo is relatively weak, and higher-order paths are not visible, but this does not affect the final localization result. It should be noted that... Figure 7 In (b), a peak exists at approximately 3m in the accumulated distance image, which is caused by the multipath effect induced by Wall-1. This can be applied to detected angles greater than [a certain value]. The goal is to judge its Is it greater than If it is greater than, then the target can be considered as being caused by The algorithm identifies multipath targets reflected from the line-of-sight (LOS) target and eliminates these targets, along with all targets whose length difference is less than the range resolution. This experiment verifies that the proposed algorithm can accurately locate LOS targets even when they are in a mixed angular range.

[0133] Case 4 (Two moving NLOS targets, one moving LOS target): The fourth experiment also used three moving targets: Target-1, Target-2, and Target-3, with initial coordinates of (1.9, 2.3) m, (3.2, 3.6) m, and (4.8, -0.3) m, respectively. Target-1 and Target-2 were located in the NLOS region, while Target-3 was located in the non-angular mixed region of the LOS region. The experimental layout is as follows: Figure 3 As shown in (c).

[0134] During data acquisition, Target-1 and Target-2 are positioned within the NLOS region along a path parallel to... Target-3 reciprocates along the axis direction within the LOS region. Axial movement. Figure 7 Figure (a) shows the final localization results of the algorithm in this application after processing and analyzing the echo data in this multi-target scene. The motion trajectories of each target are marked in the figure, and it can be seen that most of the localization points are distributed along a straight line, which matches the preset path well.

[0135] Figure 8 (b) presents the accumulated distance profile of this set of experiments, from Figure 8 (b) It can be seen that, unlike Experiment 3, in the absence of LOS target occlusion path, the echoes corresponding to different paths of each NLOS target can be clearly distinguished.

[0136] It is worth noting that during the relative movement of Target-1 and Target-2, their trajectories exhibited some discontinuities, and a few positioning points deviated from the actual paths. This phenomenon mainly stemmed from occlusion between NLOS targets and multiple reflections of echoes between NLOS targets, resulting in weaker echo energy or path confusion at certain moments. Nevertheless, considering the volume effect and natural body sway of human targets during actual movement, such deviations remain within an acceptable range in practical applications. Overall, the algorithm in this application still demonstrates good resolution and positioning stability in complex multi-target motion scenarios. This experiment verifies that the proposed algorithm can still achieve effective separation and accurate positioning in mixed (LOS / NLOS) multi-moving target scenarios.

[0137] This application embodiment also provides a non-line-of-sight target localization system based on millimeter-wave radar, including a memory and a processor. The memory includes a program for non-line-of-sight target localization based on millimeter-wave radar. When the program for non-line-of-sight target localization based on millimeter-wave radar is executed by the processor, it implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described above.

[0138] This application also provides a computer-readable storage medium storing program code, which, when executed by a processor, implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described above.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0144] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0145] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined in this application, computer-readable media does not include transient media, such as modulated data signals and carrier waves.

[0146] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A non-line-of-sight target localization method based on millimeter-wave radar, characterized in that, Includes the following steps: Step 1: Analyze the multipath signal and establish a multipath propagation model that includes LOS and NLOS target echoes; Step 2: Obtain the range profiles of the multi-channel target using Fast Fourier Transform (FFT), construct the accumulated range profiles using an incoherent accumulation method, and build the cumulative range profiles using an incoherent addition accumulation method based on these accumulated range profiles. Based on the cumulative range profiles, extract the target range information using one-dimensional cell mean constant false alarm rate (CFAR). Based on the extracted target range information, perform spatial filtering on these range cells using minimum mean square distortion-free response (MINSD) to obtain the corresponding angle information. Step 3: The obtained distance and angle information are combined to separate LOS and NLOS targets, and then redundant NLOS paths are cleaned up based on the path matching strategy. Step 4: Calculate the position of the real target based on the geometric symmetry between the real target and its false points.

2. The non-line-of-sight target localization method based on millimeter-wave radar according to claim 1, characterized in that, The establishment of a multipath propagation model that includes LOS and NLOS target echoes includes constructing a hybrid scenario of LOS and NLOS targets consisting of two buildings. The scene structure consists of wall-1, wall-2, and wall-3. The radar is located at the origin O of the coordinate axis, and the corners of the two buildings are C and D respectively. From wall-2... The distance between the axes is Wall-3 to The distance between the axes is The extension of line OC intersects wall-3 at point E, and OE is... The included angle of the axis is OD and The included angle of the axis is , Assume there are two types of targets in the scene: LOS and NLOS. The LOS target is located at point P. NLOS target Q Hidden at corner C ( )back.

3. The non-line-of-sight target localization method based on millimeter-wave radar according to claim 2, characterized in that, The extraction of target distance information in step 2 specifically includes the following steps: Using a MIMO radar with M receiving antennas, the echo signal received by the m-th receiving antenna is: ， in, For the first Chirp cycles The echo signal; To reduce environmental noise interference while preserving micro-Doppler information, the phasor mean cancellation method is used to suppress environmental noise. The processed echo signal is then subjected to FFT to obtain the range profile. ， in, For the first The first receiving antenna received the first The first group of echo signals A distance image; To ensure the accuracy of target ranging, a non-coherent accumulation method is used to accumulate the range images between channels. The accumulated range image is represented as follows: ， in, The th after accumulation The first receiver The first cycle A distance image, which can be represented as ， For the accumulated range image, the accurate target range is obtained using the one-dimensional cell mean constant false alarm rate 1-D CA-CFAR method. The detection threshold of the one-dimensional unit mean constant false alarm rate method. for ， in, This represents the probability of a false alarm. For the unit being detected The number of surrounding reference cells, for the processed distance cells, is represented as... ， Repeat the above detection process for all distance cells to obtain the target distance peak value of the accumulated distance image.

4. The non-line-of-sight target localization method based on millimeter-wave radar according to claim 3, characterized in that, The specific steps of step 3 are as follows: The NLOS target echo originates from the reflection of the EM in the ED segment of wall-3; therefore, the angle of the false target formed by the NLOS target echo is... Concentrated at the angle between OD and OE, i.e. ， Initially, LOS and NLOS targets are screened in the angle domain. The least mean square distortion-free response (MVDR) algorithm is used to estimate the target angles. The weight vector can be expressed as ,in, The autocorrelation matrix is... For angle The steering vector, therefore, the power at different angles can be expressed as... ， Similar calculations are performed on each detected distance peak to obtain the distance and angle information of the targets corresponding to all distance peaks. This yields the coordinates of all false targets and LOS targets, allowing for the filtering of targets with angles in between. The target locations between these points, namely all false targets and some LOS target locations, and Since the targets are all LOS targets, the exact location of the LOS targets can be obtained directly; Next, for the mixed range of LOS and NLOS echo angles, the path distance range of the NLOS target echo received by the radar receiver has a minimum and a maximum value. Let the minimum distance of the two-way first-order path be denoted as... The maximum value is Its distance range can be given by the following formula. ， use The distance serves as the dividing line between LOS and NLOS targets. The set of all points within the mixed region is represented as follows: ， in, Points representing the mixed intervals, Indicates its path length. Indicates its relationship with The included angle of the axis is checked for all points within the set. and Compare, if there is If it is not, it is considered a LOS target; otherwise, it is considered an NLOS target.

5. A non-line-of-sight target localization method based on millimeter-wave radar according to claim 4, characterized in that, The specific steps of step 4 are as follows: The separated NLOS echo targets are sorted in ascending order of their path lengths and represented as a set of paths. Based on geometric relationships, it can be found that The path is the shortest path, therefore the set of paths... In terms of, among them Must be a certain NLOS target Path echo, thus combining This allows us to determine the target's exact location within the NLOS region; further analysis is then performed based on the geometric relationships of the path. Thus obtain The specific length of the path, once the exact location of the target is determined, and the false points formed by its second-order two-way path. The location is thus determined, and the path length is also determined. For the mixed path of this target, there are For each element in the path set, the paths are cleaned up using a matching method based on the calculated path length, with a threshold specified here. Its value is the distance resolution, used to measure the allowable error between the calculated path length and the actual path length in the set, specifically expressed as... ， The path that is successfully matched according to the above formula will be regarded as a derivative path of the first-order two-way path to locate the target, and will be removed from the path set along with the main path. This process will then be repeated until the number of remaining paths in the path set is insufficient or the energy of the remaining paths is negligible.

6. A non-line-of-sight target localization system based on millimeter-wave radar, characterized in that, The device includes a memory and a processor. The memory includes a program for a non-line-of-sight target localization method based on millimeter-wave radar. When the program for a non-line-of-sight target localization method based on millimeter-wave radar is executed by the processor, it implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described in any one of claims 1 to 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program code, which, when executed by a processor, implements the steps of the non-line-of-sight target localization method based on millimeter-wave radar as described in any one of claims 1 to 5.

Images