Method, device and storage medium for multipath false alarm suppression for radar

Abstract

The application discloses a method and device for radar multipath false alarm suppression and a storage medium, wherein the method comprises: obtaining echo signals fed back by a moving target to radar detection signals to obtain a corresponding detection point set; clustering the detection point set to obtain a density center point set of the detection point set in a Cartesian coordinate system with the position of the radar as the origin; fitting the density center point set to obtain a density center curve; searching the detection point set based on the density center curve to obtain an upper boundary curve and a lower boundary curve corresponding to the detection point set, and removing multipath target points in the detection point set according to the upper boundary curve and the lower boundary curve. The technical scheme provided by the application can solve the technical problem in the prior art that when a traffic radar tracks and detects a vehicle, a metal object appearing directly in front of the radar will reflect radar signals, resulting in mirror image multipath false alarms.

Description

Technical Field

[0001] This invention relates to the field of technology, and in particular to a method, apparatus and storage medium for suppressing multipath false alarms in radar. Background Technology

[0002] In the field of traffic radar, millimeter-wave radar used for vehicle detection and tracking on highways inevitably encounters highway warning signs directly in front of the radar during deployment. These signs, typically made of metal and with smooth surfaces, reflect signal waves. When the radar tracks a target, the strong reflection of the radar's emitted waves by the road sign can easily generate false alarms due to mirrored multipath propagation, leading to the detection of multipath targets during target tracking. In some radar applications, the detection and tracking scenarios are complex. If metal road signs repeatedly appear in front of the radar, the number of multipath targets generated during radar tracking will be large, interfering with target tracking. To improve target tracking accuracy, it is necessary to distinguish between multipath targets and true targets, as well as the true targets corresponding to multipath targets, from a large number of target detection points.

[0003] In the prior art, some patents provide multipath target identification methods, but most of these patents target multipath targets generated by metal objects parallel to the road on both sides, such as road guardrails. For example, patent CN111699404 provides a driving assistance target acquisition method and device, radar, driving system and vehicle, and patent CN112213725A provides a radar multipath false alarm suppression method, device and terminal equipment. The multipath targets identified in these patents are all generated by guardrails. When a metal road sign appears directly in front of the radar, a moving target appearing behind the radar will generate a multipath detection point, but the radar cannot detect the real detection point generated by the moving target at this time. In the above patents, multipath targets can only be identified when the tracked target generates both multipath targets and real targets simultaneously, therefore, it cannot be used to identify multipath targets generated by warning signs in front of the radar.

[0004] In summary, although existing technologies offer various methods for identifying multipath targets, radar systems cannot quickly, accurately, and in real-time identify multipath targets when metal road signs appear in front of the radar. Therefore, a method for suppressing false alarms in multipath targets is needed to eliminate multipath targets generated by metal objects in front of the radar. Summary of the Invention

[0005] This invention provides a method, apparatus, and storage medium for suppressing multipath false alarms in radar, aiming to effectively solve the technical problem in the prior art where metallic objects appearing directly in front of the radar reflect radar signals when traffic radar tracks and detects vehicles, resulting in mirrored multipath false alarms.

[0006] According to one aspect of the present invention, a method for suppressing multipath false alarms in radar is provided, the method comprising:

[0007] Acquire the echo signal fed back by the moving target to the radar detection signal to obtain the corresponding set of detection points;

[0008] Cluster the set of detection points to obtain the set of density center points of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin;

[0009] The density center point set is fitted to obtain the density center curve;

[0010] The detection point set is searched based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and multipath target points in the detection point set are removed according to the upper boundary curve and the lower boundary curve.

[0011] Furthermore, the step of acquiring the echo signal fed back by the moving target to the radar detection signal to obtain the corresponding set of detection points includes:

[0012] Multiple original detection points are obtained based on the echo signal;

[0013] The distance information of each original detection point is normalized based on the maximum distance value of the plurality of original detection points, and the pitch angle information of each original detection point is normalized based on the maximum pitch angle value of the plurality of original detection points, so as to obtain the detection point set.

[0014] Further, the step of clustering the detection point set to obtain the density center set of the detection point set in a Cartesian coordinate system with the radar location as the origin includes:

[0015] Determine the coordinates of the cluster starting point, which serves as the first density center point; the coordinates include distance coordinates and angular coordinates.

[0016] Based on the coordinates of the clustering starting point and a preset search window, multiple density center points are determined iteratively according to the principle of maximizing the density of detection points, so as to obtain the set of density center points.

[0017] Furthermore, determining the coordinates of the clustering starting point, which serves as the first density center point, includes:

[0018] Obtain the preset detection distance range;

[0019] Within the detection distance range, the region with the highest detection point density is searched in a stepwise manner based on a preset search window size, and the coordinates of the clustering starting point are determined according to the location of the region with the highest detection point density.

[0020] Further, the step of iteratively determining multiple density center points based on the coordinates of the clustering starting point and a preset search window, and sequentially determining multiple density center points according to the principle of maximizing the detection point density, to obtain the set of density center points includes:

[0021] The clustering starting point is used as the first latest density center point, and the following operation is performed iteratively to obtain a density center point in each iteration until the iteration termination condition is met:

[0022] (1) Take the distance coordinate value in the coordinates of the latest normalized density center point as the center distance, take the angle coordinate value in the coordinates of the previous normalized density center point as the center drift direction angle, and determine the fan-shaped search area according to the center distance, the center drift direction angle, the preset search distance range, and the preset search angle range.

[0023] (2) The number of searches is determined based on the preset angle size of the fan-shaped search window and the preset search angle range, and the detection points in the fan-shaped search area are searched in a step-by-step manner based on the fan-shaped search window.

[0024] (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

[0025] Furthermore, the method also includes:

[0026] The maximum lateral search range is determined based on the latest distance coordinate value of the density center point, the preset search distance range, and the current center drift direction angle. When the maximum lateral search range is greater than 1, the iteration termination condition is satisfied, and the iteration ends.

[0027] Further, fitting the set of density center points to obtain a density center curve includes:

[0028] Obtain the distance and angle coordinates of each density center point in the set of density center points to construct a distance vector and an angle vector respectively;

[0029] The distance vector and the angle vector are optimized and fitted based on the LM algorithm to obtain the density center curve.

[0030] Further, the step of searching the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set includes:

[0031] Based on the density center curve, and according to a preset angle search step size and a preset number of search times, a translational search is performed on the detection point set to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set.

[0032] Further, the step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search iterations to obtain the upper and lower boundary curves corresponding to the detection point set includes:

[0033] Perform the following operations to obtain the upper boundary curve:

[0034] The angle search step size is added to the density center curve as a positive value to determine the search curve corresponding to each search;

[0035] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0036] The critical value for the number of upper boundary detection points is determined based on the maximum number of effective detection points and the preset upper boundary decision threshold;

[0037] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of upper boundary detection points is determined as the upper boundary curve.

[0038] Furthermore, the step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search iterations to obtain the upper and lower boundary curves corresponding to the detection point set also includes:

[0039] Perform the following operations to obtain the lower boundary curve:

[0040] The angle search step size is added to the density center curve as a negative value to determine the search curve corresponding to each search;

[0041] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0042] The critical value for the number of lower boundary detection points is determined based on the maximum number of effective detection points and the preset lower boundary decision threshold;

[0043] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of lower boundary detection points is determined as the lower boundary curve.

[0044] Further, the step of removing multipath target points from the detection point set based on the upper boundary curve and the lower boundary curve includes:

[0045] Detection points located outside the upper and lower boundary curves are discarded as multipath target points.

[0046] According to another aspect of the present invention, the present invention also provides a multipath false alarm suppression device for radar, the device comprising:

[0047] The data acquisition unit is used to acquire the echo signal fed back by the moving target to the radar detection signal, so as to obtain the corresponding set of detection points;

[0048] A density center point set determination unit is used to cluster the detection point set to obtain the density center point set of the detection point set in a Cartesian coordinate system with the location of the radar as the origin.

[0049] A density center curve determination unit is used to fit the set of density center points to obtain a density center curve;

[0050] The multipath target point elimination unit is used to search the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and to eliminate multipath target points in the detection point set according to the upper boundary curve and the lower boundary curve.

[0051] According to another aspect of the invention, the invention also provides a storage medium storing a plurality of instructions adapted for loading by a processor to execute any of the multipath false alarm suppression methods for radar as described above.

[0052] Through one or more embodiments of the above embodiments of the present invention, at least the following technical effects can be achieved:

[0053] In the technical solution disclosed in this invention, by estimating the distribution range of the elevation angle of the real detection point, and using the difference information between the false multipath target and the real target in the elevation dimension, the multipath target can be identified and eliminated, which can solve the technical problem of false alarms caused by mirror multipath caused by metal material in front of the radar. Attached Figure Description

[0054] The technical solution and other beneficial effects of the present invention will become apparent from the following detailed description of specific embodiments of the invention, in conjunction with the accompanying drawings.

[0055] Figure 1 A flowchart illustrating the steps of a multipath false alarm suppression method for radar provided in this embodiment of the invention;

[0056] Figure 2 A schematic diagram illustrating the mechanism of false alarm generation in mirror multipath propagation;

[0057] Figure 3 This is a schematic diagram of the distribution of typical distance-pitch angle measurement data;

[0058] Figure 4 is a comparison chart of data before and after data normalization;

[0059] Figure 5 A schematic diagram of density center point search provided in an embodiment of the present invention;

[0060] Figure 6 A schematic diagram of a density center point set provided in an embodiment of the present invention;

[0061] Figure 7 This is a schematic diagram of a density center curve fitting result provided in an embodiment of the present invention;

[0062] Figure 8 A schematic diagram of an upper boundary curve and a lower boundary curve provided for an embodiment of the present invention;

[0063] Figure 9 This is a schematic diagram of a multipath false alarm suppression device for radar provided in an embodiment of the present invention. Detailed Implementation

[0064] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0065] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0066] Figure 1 The diagram shows a flowchart of the steps in the multipath false alarm suppression method for radar provided by an embodiment of the present invention. The multipath false alarm suppression method for radar includes:

[0067] Step 101: Obtain the echo signal fed back by the moving target to the radar detection signal to obtain the corresponding set of detection points;

[0068] Step 102: Cluster the set of detection points to obtain the set of density center points of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin;

[0069] Step 103: Fit the set of density center points to obtain a density center curve;

[0070] Step 104: Search the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and remove multipath target points in the detection point set according to the upper boundary curve and the lower boundary curve.

[0071] In radar systems, the principle of radar target tracking is based on the time difference between the transmitted and received pulses and the propagation speed of electromagnetic waves (speed of light) to obtain the precise distance between the radar and the target. The principle of target angular position measurement utilizes the antenna's directivity; when the antenna beam is aligned with the target, the echo signal is strongest. Based on the direction of the antenna beam when the received echo is strongest, the target's direction can be determined. The principle of velocity measurement is based on the Doppler effect caused by the relative motion between the radar and the target. The frequency of the target echo received by the radar differs from the radar's transmitted frequency; the difference between the two is called the Doppler frequency. One of the main pieces of information that can be extracted from the Doppler frequency is the rate of change of distance between the radar and the target, which in turn determines the target's velocity. In summary, information such as the target's velocity, direction of motion, and distance can be obtained.

[0072] Intelligent traffic radars can be categorized by installation location into roadside fixed-point radars and vehicle-mounted radars. These radars track targets. The tracker on the radar, through hardware and software coordination under the control of the main processor, detects, acquires, and tracks targets, establishes their trajectory, warns of dangerous targets, and assists in providing collision avoidance measures. Therefore, radar tracking of acquired targets needs to achieve high accuracy to obtain predicted target motion and enter a stable target tracking state. When the radar detects a large number of multipath targets that interfere with measurement accuracy, the target tracking accuracy will decrease.

[0073] Since multipath targets are caused by the echoes of real targets being reflected back to the radar receiver by metallic objects on the road, this method estimates the distribution range of the elevation angle of the real targets and uses the difference information between multipath targets and real targets in the elevation dimension to identify and eliminate multipath targets.

[0074] In the field of traffic radar, millimeter-wave radar used for vehicle detection and tracking on highways is inevitably obstructed by highway warning signs when deployed. Since these signs are typically made of metal and have a smooth, flat surface, they strongly reflect radar waves. Therefore, target echoes are prone to false alarms due to multiple reflections from the signs and the resulting image multipath propagation.

[0075] Figure 2 The diagram illustrates the mechanism of false alarms generated by mirror multipath propagation. Paths ① to ④ represent the radar wave propagation paths between the radar and the real target 1. The radar pulse signal is refracted by the warning sign before reaching the real target 1. The echo from the real target 1 is refracted by the warning sign before reaching the radar's receiving antenna. The radar system analyzes the target's position based on the emitted pulse signal and the received echo signal. Therefore, the radar determines the position of the real target 1 as follows: Figure 2 The location of multipath target 1 is a false alarm mirrored by the real target 1; in reality, multipath target 1 does not exist.

[0076] The multipath target 1 has almost the same radial velocity as its corresponding real target 1, but the velocities are opposite, and their position coordinates are symmetrically distributed around the reflection point. However, compared to the real target 2 at the same position on the road, the height of multipath target 1 is obviously greater than that of real target 2. Since the real target 1 may appear behind the radar in this scenario, the radar cannot detect or acquire information about the real target 1, and traditional multipath false detection methods are no longer applicable. Therefore, this method estimates the distribution range of the real target's pitch angle and uses the difference in pitch dimension between the multipath target and the real target to identify and eliminate the multipath target.

[0077] The following is a detailed description of steps 101 to 104 above.

[0078] In step 101 above, the echo signal fed back by the moving target to the radar detection signal is obtained to obtain the corresponding set of detection points.

[0079] For example, when a radar tracks a target, it transmits signals frame by frame to all targets within its detection range. One frame consists of all the signals transmitted within a pulse sequence. In a pulse sequence, one or more transmitting antennas each emit pulse signals and receive the echo signals reflected back from the target. Different positions of the moving target will return echo signals, and each echo signal corresponds to a detection point. Therefore, one moving target corresponds to many detection points, forming a set of detection points.

[0080] In step 102 above, the set of detection points is clustered to obtain the set of density center points of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin.

[0081] For example, there exists a certain positional relationship between multipath targets, real targets, reflection points, and radar. Based on this fixed positional relationship between the real target and the radar, a formula relating range and elevation angle can be determined. Detection points that satisfy this formula can be identified as real targets; otherwise, they are considered false multipath targets.

[0082] Therefore, in the set of detection points, the number of actual detection points that satisfy the relationship is the largest, which translates to the highest density of coordinate points corresponding to the actual detection points in the Cartesian coordinate system. To determine the relationship between range and elevation angle, a Cartesian coordinate system needs to be established based on range and elevation angle. Then, the set of detection points is clustered to obtain the set of density center points corresponding to the coordinates of the point with the highest density. The Cartesian coordinate system uses the location of the radar as its origin.

[0083] In step 103 above, the set of density center points is fitted to obtain a density center curve.

[0084] For example, a mathematical optimization method is used to fit the set of density center points. The optimization problem refers to determining the values ​​of certain selectable variables under certain constraints to make the selected objective function optimal. This invention processes the set of density center points using a mathematical optimization method, and determines the relationship representing the density center curve based on the coordinate points corresponding to the set of density center points.

[0085] like Figure 2 As shown, assuming all real targets move in the same plane, their pitch angle and radial distance have the following relationship:

[0086]

[0087] in, Let r be the pitch angle, r be the radial distance to the target, and h be the pitch angle. m For the installation height of the radar, Set the tilt angle for the radar's elevation, where r > h. m .

[0088] Based on mathematical optimization methods and according to the relationship and the set of density center points, the correlation coefficient can be determined, and thus the exact relationship can be derived.

[0089] In step 104 above, the detection point set is searched based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and multipath target points in the detection point set are removed according to the upper boundary curve and the lower boundary curve.

[0090] For example, the detection points corresponding to the density center curve are the real target points. Correspondingly, the detection points near the density center curve are also detection points generated by the real target. Therefore, it is necessary to search for the range of detection point density that meets the preset conditions based on the density center curve, and determine the upper boundary curve and the lower boundary curve. Detection points falling between the upper boundary curve and the lower boundary curve are determined as real detection points, and other detection points are determined as multipath target points and are eliminated.

[0091] According to equation (1) in step 103 above, the two-dimensional data of the real target's range-pitch angle are distributed on the curve. Nearby. Therefore, it is necessary to... Perform curve fitting to estimate the unknown parameters. Only then can the distance-pitch curve equation of the true target be determined, and then a rejection threshold can be set based on this curve equation to reject multipath targets.

[0092] Figure 3 The diagram shows a typical distribution of range-elevation angle measurement data. If the range-elevation angle data acquired by the radar is directly fitted, the presence of false alarms will inevitably affect the fitting accuracy. Therefore, it is necessary to identify the center of the range-elevation angle data distribution density of the actual target and use the data from this center for fitting to improve the curve fitting accuracy.

[0093] Furthermore, the step of acquiring the echo signal fed back by the moving target to the radar detection signal to obtain the corresponding set of detection points includes:

[0094] Multiple original detection points are obtained based on the echo signal;

[0095] The distance information of each original detection point is normalized based on the maximum distance value of the plurality of original detection points, and the pitch angle information of each original detection point is normalized based on the maximum pitch angle value of the plurality of original detection points, so as to obtain the detection point set.

[0096] For example, normalization is used to eliminate the influence of different dimensions between indicators, and to standardize data to ensure comparability between data indicators. In short, the purpose of normalization is to confine the preprocessed data to a certain range, such as [0, 1] or [-1, 1]. For example, mapping data to the range of 0 to 1 is mainly for ease of data processing, thereby eliminating the adverse effects caused by outlier samples. After data standardization, the original data is at the same order of magnitude, making it suitable for comprehensive comparative evaluation.

[0097] For example, the distance information of the detection point is represented by a column vector r, and the pitch angle information of the detection point is represented by e. The two-dimensional data coordinates are represented in the form of (r, e), where rmax The maximum distance value, i.e., the maximum value of |r|, is e. max This is the maximum pitch angle value, which is the maximum value of |e|.

[0098] Assuming the distance-pitch data length is N r The data coordinates are normalized using the following formula:

[0099]

[0100] Among them, (x norm (n), y norm (n) represents the normalized coordinates of the points, x norm (n) represents the normalized distance information, y norm (n) represents the normalized pitch angle information, (r(n), e(n)) represents the coordinates of the original detection point, and r(n) represents the distance information of the original detection point. e(n) r represents the pitch angle information corresponding to the original detection point. max e is the maximum distance value. max N is the maximum pitch angle value. r This is the length of the distance-pitch angle data.

[0101] Figure 4 is a comparison chart of the data before and after data normalization. Figure 4-a The data is normalized, so the comparison between the data is not obvious. Figure 4-b It is normalized data, which eliminates the adverse effects caused by outlier samples. After the original data has been processed by data standardization, all indicators are on the same order of magnitude, making it suitable for comprehensive comparison and evaluation.

[0102] Further, the step of clustering the detection point set to obtain the density center set of the detection point set in a Cartesian coordinate system with the radar location as the origin includes:

[0103] Determine the coordinates of the cluster starting point, which serves as the first density center point; the coordinates include distance coordinates and angular coordinates.

[0104] Based on the coordinates of the clustering starting point and a preset search window, multiple density center points are determined iteratively according to the principle of maximizing the density of detection points, so as to obtain the set of density center points.

[0105] For example, to obtain the density center set, it is first necessary to determine the first density center point of the density center set, that is, the clustering starting point. The clustering starting point is obtained by performing a density search on the detection points, and its corresponding distance dimension coordinate value and angle dimension coordinate value are obtained.

[0106] Then, based on the first density center point, a second density center point is found, and so on, until the iteration ends. Specifically, a search window needs to be preset. Based on the coordinates of the clustering starting point and the preset search window, the density center point in each search window is determined. Then, the next density center point is found iteratively. Each time, multiple density center points are determined based on the principle of maximizing the density of detection points, and a set of density center points is obtained based on these multiple density center points.

[0107] Furthermore, determining the coordinates of the clustering starting point, which serves as the first density center point, includes:

[0108] Obtain the preset detection distance range;

[0109] Within the detection distance range, the region with the highest detection point density is searched in a stepwise manner based on a preset search window size, and the coordinates of the clustering starting point are determined according to the location of the region with the highest detection point density.

[0110] For example, first, a preset detection distance range is determined, such as around 20 or 30 meters. Then, within this fixed detection distance range, the density of detection points at each segment is searched in a step-by-step manner based on a preset search window size. After all search windows have been searched, the region with the highest detection point density is obtained, and then the coordinates of the clustering starting point are determined based on the coordinates of the region with the highest detection point density.

[0111] For example, a vertical search method can be used to determine the starting coordinates for clustering. For instance, at a distance r... c At (0) = 30m, within a range of ±1m, search in increments of 0.5° and a search window size of 1° along the data pitch angle dimension to find the position e with the most data within the search window. c (0).

[0112] Further, the step of iteratively determining multiple density center points based on the coordinates of the clustering starting point and a preset search window, and sequentially determining multiple density center points according to the principle of maximizing the detection point density, to obtain the set of density center points includes:

[0113] The clustering starting point is used as the first latest density center point, and the following operation is performed iteratively to obtain a density center point in each iteration until the iteration termination condition is met:

[0114] (1) Take the distance coordinate value in the coordinates of the latest normalized density center point as the center distance, take the angle coordinate value in the coordinates of the previous normalized density center point as the center drift direction angle, and determine the fan-shaped search area according to the center distance, the center drift direction angle, the preset search distance range, and the preset search angle range.

[0115] (2) The number of searches is determined based on the preset angle size of the fan-shaped search window and the preset search angle range, and the detection points in the fan-shaped search area are searched in a step-by-step manner based on the fan-shaped search window.

[0116] (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

[0117] For example, Figure 5 This diagram illustrates the search for density centroids. The clustering start point is used as the first and most recent density centroid to begin iteration. Each iteration yields one density centroid, and the clustering iteration stops when the iteration termination condition is met. The specific steps are as follows:

[0118] (1) Determine the sector search area.

[0119] For example, such as Figure 5 As shown, the relevant parameters of the sector search area include the center, radius, central angle, and direction of the sector.

[0120] Wherein, the center of the sector represents the coordinates of the latest normalized density center point. Figure 5 The center of the sector shown is the data coordinate of the first density center, i.e., the starting point of the clustering. Its normalized data coordinates are (x... c (0), y c (0));

[0121] The radius of the sector is the preset search distance range r. w The center distance is determined based on the distance dimension coordinates. Figure 5 The normalized distance range is [r] nc (0), r nc (0)+r w ], where r nc (0) is the normalized distance from the density center point to the origin.

[0122] The direction of the sector is determined by the center drift direction angle, which can be understood as the slope of the straight line corresponding to two density center points, since the slopes of the straight lines between three adjacent density center points are similar. Therefore, the center drift direction angle when searching for the third density center point can be determined by the slope between the first and second density center points, that is, the angular coordinate value in the normalized coordinates of the previous density center point is used as the center drift direction angle. During the density center search process, the size of the sector region remains unchanged, and its direction is determined by the center drift direction angle. The update formula is as follows:

[0123]

[0124] in, Figure 5 The center is the first density center point, and its center drift direction angle is a preset angle value, for example,

[0125] The central angle of the sector is determined based on a preset angle value.

[0126] After all relevant parameters are determined, based on the center distance and center drift direction angle... The preset search distance range and the preset search angle range a w To determine the sector search area.

[0127] For example, such as Figure 5 As shown, with (x c (0), y c (0)) is the center of the sector, and r is the radius of the sector. w Radius, center drift direction angle As the direction, at a preset angle a w A normalized sector-shaped search area was established to define the search angle range. Perform a density center search on the data falling within this region.

[0128] (2) Determine the number of searches and search sequentially.

[0129] For example, after determining the sector-shaped search area, it is further divided into several sector-shaped search windows. Specifically, the angle of each sector-shaped search window is a preset value, and the number of searches is determined based on the angle size and the search angle range. Then, the detection points within the sector-shaped search area are searched sequentially and stepwise based on the sector-shaped search windows.

[0130] Of course, the number of searches can also be preset, and the angle of the fan-shaped search window can be determined based on the number of searches. Both methods can achieve the same effect and are within the scope of protection of this invention.

[0131] For example, Figure 5 In the middle, the fan-shaped search area is The search angle range is a w The angle of the fan-shaped search window is Then the number of searches

[0132] (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

[0133] For example, in L s Within each search, the search window with the highest detection point density is identified, indicating that the data in that area is the densest. This area is then designated as the target area. The coordinates of all detection points within the target area are averaged, and the resulting coordinates are used as the next latest density center point.

[0134] For example, Figure 5 The second density center shown is (r) c (1), e c (1)), its normalized coordinates are (x) c (1), y c (1)). Thus, the latest density center point (x) can be obtained. c (1), y c (1) As a new starting point, return to step (1) to establish a new sector search area, and repeat the above method to start the next round of density center search.

[0135] Furthermore, the method also includes:

[0136] The maximum lateral search range is determined based on the latest distance coordinate value of the density center point, the preset search distance range, and the current center drift direction angle. When the maximum lateral search range is greater than 1, the iteration termination condition is satisfied, and the iteration ends.

[0137] For example, when When x reaches this point, the search for the density center of the data is considered complete. c (i) represents the distance dimension coordinates of the latest normalized density center points, r w , which is the preset search distance range. The current center drift direction angle is used. Based on the preset search distance range and the current center drift direction angle, the maximum lateral position corresponding to the sector search area can be determined. Then, based on the latest distance coordinate value of the density center point and the maximum lateral position corresponding to the sector search area, the current maximum lateral search range can be determined. In the normalized coordinate system, the maximum value of the distance coordinate is 1. Therefore, the x-coordinates of all detection points are less than or equal to 1. Thus, when the maximum lateral search range is greater than 1, there are no more detection points, and the iteration ends.

[0138] Figure 6The diagram shows a set of density center points. After the search ends, the search results are multiple coordinate points. During the search process, the saved data are the coordinate data corresponding to the unnormalized density center points, denoted as x = [r c (1) r c (2) ... r c (N)] T y = [e c (1) e c (2) ... e c (N)] T , where N is the length of the density center data.

[0139] Further, fitting the set of density center points to obtain a density center curve includes:

[0140] Obtain the distance and angle coordinates of each density center point in the set of density center points to construct a distance vector and an angle vector respectively;

[0141] The distance vector and the angle vector are optimized and fitted based on the LM algorithm to obtain the density center curve.

[0142] For example, Figure 7 The image shows a schematic diagram of the density center curve fitting results. Figure 6 The set of density center points in the data was fitted to a density center curve. Specifically, the data was processed using the LM algorithm for mathematical optimization. First, the set of density center points was processed into distance and angle vectors, and then the density center curve was obtained using the LM algorithm.

[0143] The Levenberg-Marquardt (LM) algorithm is an optimization algorithm that finds the parameter vector that minimizes the function value. It is widely used in data analysis, optimal design, and electronic design. The LM algorithm is the most widely used nonlinear least squares algorithm. It is an algorithm that uses gradients to find the maximum (minimum) value. Figuratively speaking, it belongs to the "hill climbing" method and has the advantages of both the gradient method and Newton's method, resulting in a relatively fast optimization speed.

[0144] Based on the above, it can be seen that there is a certain relationship between the distance to the detection point and the pitch angle, and the curve equation corresponding to this relationship is:

[0145]

[0146] Where n = 1, 2, ..., N, N is the number of samples, and {a, b} are the parameters of the equation. Using the parameters of the equation, we construct an estimation vector P for the parameters, and let p = [ab]. Then, we construct the first error function as shown in the following equation:

[0147]

[0148] Among them, let Correspondingly, the first error function is expressed as follows:

[0149]

[0150] The first error function f(p) can be obtained by performing a first-order Taylor expansion as follows:

[0151] f(p+h)=f+Jh+O(h T h),

[0152] Where h is the step size, and the Jacobian matrix J is given by [formula missing]. Further, the Jacobian matrix J is shown below:

[0153]

[0154] After removing the higher-order term O(h) of f(p+h) T h) after which a damping term is added. The second error function can then be approximated as follows:

[0155]

[0156] Differentiating the error function G(h) with respect to h yields G′(h), and setting G′(h) = 0, we obtain the following equation:

[0157] (J T J+μI)h=-J T f,

[0158] This leads to the expression for the step size h, as shown in the following formula:

[0159] h = -(J T J+μI) -1 J T f,

[0160] Where μ is the damping factor, and μ > 0.

[0161] It can be seen that, in the second error function G(h), besides h T J T If all other terms in term f are greater than 0, then this term determines the direction of descent of the second error function G(h). When the damping factor μ > 0, the following equation can be obtained:

[0162] h T J T f = -h T (J T J+μI)h<0

[0163] Therefore, setting the damping factor μ > 0 ensures that the step size h can represent the descent direction of the second error function G(h).

[0164] During the iterative process, the change in the magnitude of the first error function F(p) is used to determine whether the current step size is acceptable. If F(p+h)≤F(p), it means the step size h is acceptable, so the damping factor μ is reduced to accelerate the convergence speed. Otherwise, it means the step size is inappropriate, causing the optimal solution to be skipped, so the damping factor μ is increased and the step size is reduced. This ensures that the optimal solution is quickly approximated with an appropriate step size.

[0165] When the step size ||h||2 is small enough, let p opt =p+h, then p is considered opt This is the optimal estimate of the parameter. Correspondingly, based on the parameter p... opt =[a opt b opt ] T The following fitted curve can be obtained:

[0166]

[0167] The correlation coefficient of the function was determined from the curve, which in turn determined the relationship between the distance of the detection point and the pitch angle.

[0168] Further, the step of searching the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set includes:

[0169] Based on the density center curve, and according to a preset angle search step size and a preset number of search times, a translational search is performed on the detection point set to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set.

[0170] For example, due to measurement errors, the distance-pitch angle data corresponding to the true target cannot all fall on the density center curve, but rather on its density center curve y m The values ​​of the pitch angles of the actual target are uniformly distributed within a certain range around (x). It can be assumed that the measured pitch angles of the actual target are distributed along the lower boundary curve y. l (x) and the upper boundary curve y h Between (x), we can base our conclusion on the density center curve y. m (x), the lower boundary curve y is determined using a translation search method. l (x) and the upper boundary curve y h (x). After determining the distribution boundaries of the true target data, false alarm data outside the upper and lower boundary curves can be eliminated.

[0171] Since a translational search method is used, the density center curve is moved by a preset angle and search step size. The position of the density center curve in the coordinate axis after translation can be determined according to the preset number of searches. Then, the number of detection points within a specific range can be determined. Based on the number of detection points, the upper boundary curve and the lower boundary curve can be determined.

[0172] Further, the step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search iterations to obtain the upper and lower boundary curves corresponding to the detection point set includes:

[0173] Perform the following operations to obtain the upper boundary curve:

[0174] The angle search step size is added to the density center curve as a positive value to determine the search curve corresponding to each search;

[0175] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0176] The critical value for the number of upper boundary detection points is determined based on the maximum number of effective detection points and the preset upper boundary decision threshold;

[0177] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of upper boundary detection points is determined as the upper boundary curve.

[0178] For example, Figure 8 This diagram illustrates the upper and lower boundary curves. When determining the boundary curves, let the angle search step size be Δy, and the search count be i = 1, 2, ..., N. s , where N s Let be the number of searches. Then, the function expression corresponding to the curve at the i-th search is:

[0179]

[0180] make The angle search step size is added to the density center curve as a positive value to determine the search curve corresponding to each search. The number of valid detection points satisfying preset conditions in each search is then determined based on the search curve. The number of valid targets L(i) in the i-th search is shown in the following formula:

[0181]

[0182] Here, the function number{·} represents the calculation of the number of measured values ​​that satisfy the conditions, i.e., the curve. and curve The number of detection points between them.

[0183] The critical value for the number of upper boundary detection points is determined based on the maximum number of effective detection points and the preset upper boundary decision threshold. Specifically, the maximum number of effective detection points L in the entire search is first determined. max and the pre-set upper boundary decision threshold T H Then, the critical value for the number of detection points at the upper boundary is calculated.

[0184] Then, it is determined when the number of effective detection points first falls below the upper boundary detection point threshold. At this point, the corresponding search curve is determined as the upper boundary curve. Mathematically, this means determining whether the number of effective detection points in the m-th search satisfies the following relationship:

[0185] L(m)<L max T H ,

[0186] If this relation is satisfied, then it is considered that... Let L be the upper boundary curve. max For N s The maximum number of valid measurements in the search, T H This is the upper boundary judgment threshold.

[0187] The search curve corresponding to the condition being met is determined, and this curve is the upper boundary curve, whose mathematical expression is as follows:

[0188]

[0189] Furthermore, the step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search iterations to obtain the upper and lower boundary curves corresponding to the detection point set also includes:

[0190] Perform the following operations to obtain the lower boundary curve:

[0191] The angle search step size is added to the density center curve as a negative value to determine the search curve corresponding to each search;

[0192] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0193] The critical value for the number of lower boundary detection points is determined based on the maximum number of effective detection points and the preset lower boundary decision threshold;

[0194] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of lower boundary detection points is determined as the lower boundary curve.

[0195] For example, the lower boundary curve y can be obtained using the same method as the method used to determine the upper boundary curve. l (x).

[0196] Further, the step of removing multipath target points from the detection point set based on the upper boundary curve and the lower boundary curve includes:

[0197] Detection points located outside the upper and lower boundary curves are discarded as multipath target points.

[0198] For example, multipath target data can be eliminated based on the upper and lower boundary curves. Specifically, the multipath target point y can be identified based on the following discriminant. f (x):

[0199] y f (x)∈(-∞, y) l (x))∪(y h (x), +∞),

[0200] Determine the upper boundary curve y h The detection points between (x) and the lower boundary curve y(x) are the detection points corresponding to the true target, and multipath target points y are removed. f (x), the multipath target point is a multipath target and will not participate in subsequent data processing.

[0201] Through one or more embodiments of the above embodiments of the present invention, at least the following technical effects can be achieved:

[0202] In the technical solution disclosed in this invention, by estimating the distribution range of the elevation angle of the real detection point, and using the difference information between the false multipath target and the real target in the elevation dimension, the multipath target can be identified and eliminated, which can solve the technical problem of false alarms caused by mirrored multipath caused by metal material in front of the radar.

[0203] Based on the same inventive concept as the multipath false alarm suppression method for radar in the embodiments of the present invention, the embodiments of the present invention provide a multipath false alarm suppression device for radar. Please refer to... Figure 9 The device includes:

[0204] The data acquisition unit 201 is used to acquire the echo signal fed back by the moving target to the radar detection signal, so as to obtain the corresponding set of detection points;

[0205] Density center point set determination unit 202 is used to cluster the detection point set to obtain the density center point set of the detection point set in the Cartesian coordinate system with the location of the radar as the origin.

[0206] Density center curve determination unit 203 is used to fit the set of density center points to obtain a density center curve;

[0207] The multipath target point elimination unit 204 is used to search the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and to eliminate multipath target points in the detection point set according to the upper boundary curve and the lower boundary curve.

[0208] Furthermore, the data acquisition unit 201 is also used for:

[0209] Multiple original detection points are obtained based on the echo signal;

[0210] The distance information of each original detection point is normalized based on the maximum distance value of the plurality of original detection points, and the pitch angle information of each original detection point is normalized based on the maximum pitch angle value of the plurality of original detection points, so as to obtain the detection point set.

[0211] Furthermore, the density center point set determination unit 202 is also used for:

[0212] Determine the coordinates of the cluster starting point, which serves as the first density center point; the coordinates include distance coordinates and angular coordinates.

[0213] Based on the coordinates of the clustering starting point and a preset search window, multiple density center points are determined iteratively according to the principle of maximizing the density of detection points, so as to obtain the set of density center points.

[0214] Furthermore, the density center point set determination unit 202 is also used for:

[0215] Obtain the preset detection distance range;

[0216] Within the detection distance range, the region with the highest detection point density is searched in a stepwise manner based on a preset search window size, and the coordinates of the clustering starting point are determined according to the location of the region with the highest detection point density.

[0217] Furthermore, the density center point set determination unit 202 is also used for:

[0218] The clustering starting point is used as the first latest density center point, and the following operation is performed iteratively to obtain a density center point in each iteration until the iteration termination condition is met:

[0219] (1) Take the distance coordinate value in the coordinates of the latest normalized density center point as the center distance, take the angle coordinate value in the coordinates of the previous normalized density center point as the center drift direction angle, and determine the fan-shaped search area according to the center distance, the center drift direction angle, the preset search distance range, and the preset search angle range.

[0220] (2) The number of searches is determined based on the preset angle size of the fan-shaped search window and the preset search angle range, and the detection points in the fan-shaped search area are searched in a step-by-step manner based on the fan-shaped search window.

[0221] (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

[0222] Furthermore, the method also includes:

[0223] The maximum lateral search range is determined based on the latest distance coordinate value of the density center point, the preset search distance range, and the current center drift direction angle. When the maximum lateral search range is greater than 1, the iteration termination condition is satisfied, and the iteration ends.

[0224] Furthermore, the density center curve determination unit 203 is also used for:

[0225] Obtain the distance and angle coordinates of each density center point in the set of density center points to construct a distance vector and an angle vector respectively;

[0226] The distance vector and the angle vector are optimized and fitted based on the LM algorithm to obtain the density center curve.

[0227] Furthermore, the multipath target point elimination unit 204 is also used for:

[0228] Based on the density center curve, and according to a preset angle search step size and a preset number of search times, a translational search is performed on the detection point set to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set.

[0229] Furthermore, the multipath target point elimination unit 204 is also used for:

[0230] Perform the following operations to obtain the upper boundary curve:

[0231] The angle search step size is added to the density center curve as a positive value to determine the search curve corresponding to each search;

[0232] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0233] The critical value for the number of upper boundary detection points is determined based on the maximum number of effective detection points and the preset upper boundary decision threshold;

[0234] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of upper boundary detection points is determined as the upper boundary curve.

[0235] Furthermore, the multipath target point elimination unit 204 is also used for:

[0236] Perform the following operations to obtain the lower boundary curve:

[0237] The angle search step size is added to the density center curve as a negative value to determine the search curve corresponding to each search;

[0238] The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles.

[0239] The critical value for the number of lower boundary detection points is determined based on the maximum number of effective detection points and the preset lower boundary decision threshold;

[0240] The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of lower boundary detection points is determined as the lower boundary curve.

[0241] Furthermore, the multipath target point elimination unit 204 is also used for:

[0242] Detection points located outside the upper and lower boundary curves are discarded as multipath target points.

[0243] Other aspects and implementation details of the multipath false alarm suppression device for radar described herein are the same as or similar to the multipath false alarm suppression method for radar described above, and will not be repeated here.

[0244] According to another aspect of the invention, the invention also provides a storage medium storing a plurality of instructions adapted for loading by a processor to execute any of the multipath false alarm suppression methods for radar as described above.

[0245] In summary, although the present invention has been disclosed above with reference to preferred embodiments, the above preferred embodiments are not intended to limit the present invention. Those skilled in the art can make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the scope defined in the claims.

Claims

1. A method for suppressing multipath false alarms in radar, characterized in that, The method includes: acquiring the echo signal fed back by the moving target to the radar detection signal, so as to obtain the corresponding set of detection points; Cluster the set of detection points to obtain the set of density center points of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin; The density center point set is fitted to obtain the density center curve; The detection point set is searched based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and multipath target points in the detection point set are removed according to the upper boundary curve and the lower boundary curve. The step of clustering the set of detection points to obtain the set of density centers of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin includes: determining the coordinates of the clustering starting point as the first density center point, wherein the coordinates include distance coordinates and angular coordinates. Based on the coordinates of the clustering starting point and a preset search window, multiple density center points are determined iteratively according to the principle of maximizing the density of detection points, so as to obtain the set of density center points; The step of determining multiple density center points in an iterative manner based on the coordinates of the clustering starting point and a preset search window, and in accordance with the principle of maximum detection point density, to obtain the set of density center points includes: taking the clustering starting point as the first latest density center point, and performing the following operations in an iterative manner to obtain a density center point in each iteration until the iteration end condition is met: (1) taking the distance dimension coordinate value in the coordinates of the normalized latest density center point as the center distance, taking the angle dimension coordinate value in the coordinates of the normalized previous density center point as the center drift direction angle, and determining the fan-shaped search area according to the center distance, the center drift direction angle, the preset search distance range, and the preset search angle range; (2) The number of searches is determined based on the preset angle size of the fan-shaped search window and the preset search angle range, and the detection points in the fan-shaped search area are searched in a step-by-step manner based on the fan-shaped search window. (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

2. The method as described in claim 1, characterized in that, The step of obtaining the echo signal fed back by the moving target to the radar detection signal to obtain the corresponding set of detection points includes: obtaining multiple original detection points based on the echo signal; The distance information of each original detection point is normalized based on the maximum distance value of the plurality of original detection points, and the pitch angle information of each original detection point is normalized based on the maximum pitch angle value of the plurality of original detection points, so as to obtain the detection point set.

3. The method as described in claim 2, characterized in that, Determining the coordinates of the clustering starting point, which serves as the first density center point, includes: obtaining a preset detection distance range; Within the detection distance range, the region with the highest detection point density is searched in a stepwise manner based on a preset search window size, and the coordinates of the clustering starting point are determined according to the location of the region with the highest detection point density.

4. The method as described in claim 1, characterized in that, The method further includes: determining the maximum lateral search range based on the latest distance coordinate value of the density center point, the preset search distance range, and the current center drift direction angle; when the maximum lateral search range is greater than 1, it is determined that the iteration termination condition is met, and the iteration ends.

5. The method as described in claim 1, characterized in that, The step of fitting the set of density center points to obtain the density center curve includes: obtaining the distance coordinate value and the angle coordinate value of each density center point in the set of density center points, so as to construct the distance vector and the angle vector respectively; The distance vector and the angle vector are optimized and fitted based on the LM algorithm to obtain the density center curve.

6. The method as described in claim 1, characterized in that, The step of searching the detection point set based on the density center curve to obtain the upper and lower boundary curves corresponding to the detection point set includes: performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset search number to obtain the upper and lower boundary curves corresponding to the detection point set.

7. The method as described in claim 6, characterized in that, The step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search times to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set includes: performing the following operations to obtain the upper boundary curve: The angle search step size is added to the density center curve as a positive value to determine the search curve corresponding to each search; The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles. The critical value for the number of upper boundary detection points is determined based on the maximum number of effective detection points and the preset upper boundary decision threshold. The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of upper boundary detection points is determined as the upper boundary curve.

8. The method as described in claim 6, characterized in that, The step of performing a translational search on the detection point set based on the density center curve and according to a preset angle search step size and a preset number of search times to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set further includes: performing the following operation to obtain the lower boundary curve: The angle search step size is added to the density center curve as a negative value to determine the search curve corresponding to each search; The number of valid detection points that meet the preset conditions in each search is determined based on the search curve, and the maximum number of valid detection points in all searches is determined by the number of search cycles. The critical value for the number of lower boundary detection points is determined based on the maximum number of effective detection points and the preset lower boundary decision threshold; The search curve corresponding to the first occurrence of the number of effective detection points being less than the critical value of the number of lower boundary detection points is determined as the lower boundary curve.

9. The method as described in claim 1, characterized in that, The step of removing multipath target points from the detection point set based on the upper boundary curve and the lower boundary curve includes: removing detection points located outside the upper boundary curve and the lower boundary curve as multipath target points.

10. A multipath false alarm suppression device for radar, characterized in that, The device includes: a data acquisition unit, used to acquire the echo signal fed back by the moving target to the radar detection signal, so as to obtain the corresponding set of detection points; A density center point set determination unit is used to cluster the detection point set to obtain the density center point set of the detection point set in a Cartesian coordinate system with the location of the radar as the origin. A density center curve determination unit is used to fit the set of density center points to obtain a density center curve; The multipath target point elimination unit is used to search the detection point set based on the density center curve to obtain the upper boundary curve and lower boundary curve corresponding to the detection point set, and to eliminate the multipath target points in the detection point set according to the upper boundary curve and the lower boundary curve. The step of clustering the set of detection points to obtain the set of density centers of the set of detection points in a Cartesian coordinate system with the location of the radar as the origin includes: determining the coordinates of the clustering starting point as the first density center point, wherein the coordinates include distance coordinates and angular coordinates. Based on the coordinates of the clustering starting point and a preset search window, multiple density center points are determined iteratively according to the principle of maximizing the density of detection points, so as to obtain the set of density center points; The step of determining multiple density center points in an iterative manner based on the coordinates of the clustering starting point and a preset search window, and in accordance with the principle of maximum detection point density, to obtain the set of density center points includes: taking the clustering starting point as the first latest density center point, and performing the following operations in an iterative manner to obtain a density center point in each iteration until the iteration end condition is met: (1) taking the distance dimension coordinate value in the coordinates of the normalized latest density center point as the center distance, taking the angle dimension coordinate value in the coordinates of the normalized previous density center point as the center drift direction angle, and determining the fan-shaped search area according to the center distance, the center drift direction angle, the preset search distance range, and the preset search angle range; (2) The number of searches is determined based on the preset angle size of the fan-shaped search window and the preset search angle range, and the detection points in the fan-shaped search area are searched in a step-by-step manner based on the fan-shaped search window. (3) Determine the area where the search window with the highest detection point density is located as the target area, and perform an average calculation on the coordinates of the detection points in the target area to obtain the coordinates of the latest density center point and return to step (1).

11. A storage medium, characterized in that, The storage medium stores a plurality of instructions adapted for loading by a processor to execute the multipath false alarm suppression method for radar as described in any one of claims 1 to 9.

Images