Dynamic and static state identification method for target point, radar signal processing method, integrated circuit, and electromagnetic wave device

Abstract

A dynamic and static state identification method for a target point, a radar signal processing method, an integrated circuit, and an electromagnetic wave device. The dynamic and static state identification method for a target point is applied to a radar system, and comprises: on the basis of first radar attitude information, selecting a plurality of candidate stationary target points from among target points identified within a current frame (110); using the first radar attitude information as an initial fitting value, and on the basis of radar detection information of the plurality of candidate stationary target points, performing radar attitude model fitting to determine radar attitude information corresponding to the current frame (120); and on the basis of the radar attitude information corresponding to the current frame, performing dynamic and static state identification on the target points identified within the current frame (130). By performing model fitting on the basis of the screened candidate stationary target points, the fitting speed can be effectively increased, and the fitting accuracy is improved, thereby improving the accuracy and performance of dynamic and static state identification of the target points.

Description

Target point dynamic and static identification, radar signal processing methods, integrated circuits and electromagnetic wave devices

[0001] This application claims priority to Chinese patent application No. 202411740580.8, filed on November 28, 2024, entitled “Signal Processing Method and Apparatus for Antenna Array, Integrated Circuit, Radio Device and Equipment”, the contents of which shall be construed as incorporated herein by reference. Technical Field

[0002] This article relates to the field of electromagnetic wave technology, specifically a method for identifying dynamic and static target points, a radar signal processing method, integrated circuits, and electromagnetic wave devices. Background Technology

[0003] Radar target detection technology is widely used in both military and civilian fields, especially in the field of autonomous driving. Using radar sensors to perceive the surrounding environment to ensure driving safety and stability has become an important component of autonomous driving solutions. Radar systems transmit radio waves and receive the reflected echo signals, using the time delay and Doppler shift of the signals to identify targets and further determine information such as the target's distance, position, and speed.

[0004] Taking autonomous driving applications as an example, vehicle-mounted radar systems continuously monitor the surrounding environment, needing to effectively identify surrounding buildings, trees, traffic signs, vehicles, pedestrians, and other objects, including both static and dynamic targets. Considering various aspects of driving safety and comfort, the identified static and dynamic targets require different data and processing solutions. Therefore, efficient and accurate identification of both static and dynamic targets is a crucial aspect of radar target detection solutions, laying a solid data foundation for more intelligent and reliable autonomous driving solutions. Summary of the Invention

[0005] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0006] This disclosure provides a method for identifying the dynamic and static state of target points, a radar signal processing method, an integrated circuit, and an electromagnetic wave device. Based on the screening of candidate static target points, model fitting can effectively accelerate the fitting speed and improve the fitting accuracy, thereby enhancing the accuracy and performance of target point dynamic and static identification.

[0007] This disclosure provides a method for identifying the dynamic and static states of a target point, specifically a method for identifying the dynamic and static states of a target point, applied to a radar system. This method may include:

[0008] Based on the first radar attitude information, select multiple candidate stationary target points from the target points detected in the current frame;

[0009] Using the first radar attitude information as the initial value for fitting, the radar attitude model is fitted according to the radar detection information of the multiple candidate stationary target points to determine the radar attitude information corresponding to the current frame.

[0010] Based on the radar attitude information corresponding to the current frame, the target points detected in the current frame are identified as dynamic or static.

[0011] The first radar attitude information is obtained from at least one of the following: radar attitude information corresponding to historical frames, and radar attitude information collected in real time by the radar system carrying equipment.

[0012] In some exemplary embodiments, the radar system includes: a vehicle-mounted frequency-modulated continuous wave millimeter-wave radar system.

[0013] In some exemplary embodiments, radar attitude model fitting can be performed based on methods such as direct least squares fitting or robust fitting based on random sampling (such as RANSAC).

[0014] This disclosure also provides a method for dynamic and static target point identification, applied in a radar. The radar includes a radio frequency transceiver and a processor. The processor is configured to process signals received by the radio frequency transceiver to obtain target detection data. The target detection data includes the Doppler velocity and elevation angle of multiple target points. The method includes:

[0015] Obtain the target detection data for the current frame;

[0016] Based on the first radar attitude information, multiple candidate stationary target points are selected from the target detection data of the current frame, according to a set pitch angle range and a set Doppler velocity range:

[0017] Using the first radar attitude information as the initial fitting value, a preset model is used to fit the radar attitude model to the plurality of candidate stationary target points, generating the radar attitude information corresponding to the current frame; and,

[0018] Based on the generated radar attitude information, all target points in the current frame are classified into dynamic and static categories, and target point cloud data containing dynamic and static labels are output.

[0019] This disclosure also provides a radar signal processing method, including: classifying target points; and performing post-processing on target points of different categories using different processing methods.

[0020] This disclosure also provides a radar system including a processor and a memory, which is configured to execute the radar signal processing method described above based on the processor and memory, so as to reduce computing resource consumption and improve radar real-time performance through dynamic and static identification.

[0021] This disclosure also provides a radar signal processing method applied to a frequency modulated continuous wave (FMCW) millimeter-wave radar. The FMCW millimeter-wave radar includes a radio frequency (RF) front-end and a processor. The RF front-end is configured to transmit and receive radar signals, and the processor is configured to perform signal processing on the received radar signals to obtain point cloud data. The method includes:

[0022] The point cloud data is classified into target points to generate a set of moving target points and a set of stationary target points;

[0023] Perform occupancy raster map construction on a set of stationary target points, and output location information that can be used for map construction; and

[0024] Perform Doppler velocity deblurring on the set of moving target points, apply the DBSCAN clustering algorithm for density-based noisy spatial clustering, and perform Kalman filter tracking to output one or more of the following: target identifier, position, velocity, and trajectory history.

[0025] This disclosure also provides an integrated circuit, including a processor configured to implement the target point dynamic and static identification method as described in any embodiment of this disclosure; or to implement the radar signal processing method as described in any embodiment of this disclosure.

[0026] In some exemplary embodiments, the processor may be an embedded microcontroller (MCU) or a digital signal processor (DSP), which may be integrated into a communication or radar SoC.

[0027] This disclosure also provides an electromagnetic wave device, comprising: a carrier; an integrated circuit as described in any embodiment of this disclosure, disposed on the carrier; an antenna, disposed on the carrier, or the antenna and the integrated circuit are integrated into a single device disposed on the carrier; wherein the integrated circuit is connected to the antenna, and the antenna is used to transmit radio frequency signals and receive radio frequency signals.

[0028] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the solutions described in the description and the accompanying drawings.

[0029] After reading and understanding the accompanying diagrams and detailed descriptions, the other aspects can be understood.

[0030] Overview of the attached figures

[0031] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0032] The embodiments of the present disclosure will be described in further detail below with reference to the accompanying drawings:

[0033] Figure 1 is a flowchart of a target point motion and static identification method provided in an embodiment of this application;

[0034] Figure 2 is a flowchart of another target point motion and static recognition method provided in an embodiment of this application;

[0035] Figure 3 is a flowchart of another target point motion and static recognition method provided in an embodiment of this application;

[0036] Figure 4 is a flowchart of a radar signal processing method provided in an embodiment of this application;

[0037] Figure 5 is a schematic diagram of the structure of an integrated circuit provided in an embodiment of this application.

[0038] Detailed Explanation

[0039] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings. The implementation can be carried out in many different forms. Those skilled in the art will readily understand that the methods and content can be transformed into other forms without departing from the spirit and scope of this disclosure. Therefore, this disclosure should not be construed as limited to the content described in the following embodiments. Unless otherwise specified, the embodiments and features in the embodiments of this disclosure can be arbitrarily combined with each other.

[0040] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed herein can also be combined with any conventional features or elements to form a unique inventive scheme as defined by the claims. Any feature or element of any embodiment can also be combined with features or elements from other inventive schemes to form another unique inventive scheme as defined by the claims. Therefore, it should be understood that any feature shown or discussed in the embodiments of this application can be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes can be made within the scope of the appended claims.

[0041] Furthermore, in describing representative embodiments, the specification may have presented the method or process as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims relating to the method or process should not be limited to the steps performed in the order written, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0042] Taking intelligent driving applications as an example, vehicles traveling at high speeds place higher demands on the performance of radar target detection. In solutions for radar signal processing to obtain the identification and detection results of multiple target points, steps such as velocity fuzzy resolution, target point clustering, and tracking are involved. A large amount of target point data participates in these computational steps, consuming significant computational resources and directly impacting radar target detection performance. Research has found that in most application scenarios, the necessity for velocity fuzzy resolution or tracking clustering of stationary target points is low. In some feasible radar target detection algorithms, more than 50% of the target points involved in the calculation are often stationary. For example, in some 4D radar target detection solutions, more than 80% of the target points involved in the calculation are stationary. It can be seen that processing a large number of stationary target point data requires substantial computational resources, significantly impacting the overall performance of radar detection.

[0043] It is known that in radar-based target detection and environmental perception schemes, stationary target points are very useful for free-space identification, while moving target points require more refined calculation, clustering, and classification processing. Distinguishing between stationary and moving target points and providing separate input data for different detection function requirements greatly helps reduce receiver computational resource consumption and improve processing performance.

[0044] This application provides a method for identifying the static and dynamic targets detected in each frame of radar signal. The method distinguishes between stationary and moving target points. Specifically, in each frame of radar signal, after determining the moving target points (dynamic target points) according to a relevant algorithm, the remaining target points are considered stationary target points (static target points); or, after determining the stationary target points, the remaining target points are considered moving target points.

[0045] This application provides a method for identifying the dynamic and static states of target points, as shown in Figure 1, including:

[0046] Step 110: Select multiple candidate stationary target points from the target points detected in the current frame based on the first radar attitude information;

[0047] Step 120: Using the first radar attitude information as the initial fitting value, the radar attitude model is fitted according to the radar detection information of the multiple candidate stationary target points to determine the radar attitude information corresponding to the current frame.

[0048] Step 130: Based on the radar attitude information corresponding to the current frame, perform dynamic and static identification on the target points detected in the current frame;

[0049] The first radar attitude information is obtained based on at least one of the following:

[0050] The radar attitude information corresponding to historical frames and the radar attitude information collected in real time by the radar system's carrying equipment.

[0051] In some exemplary embodiments, the target point dynamic and static identification method is applied to a radar system, such as a vehicle-mounted frequency-modulated continuous wave millimeter-wave radar. The initial fitting value is the initial value of the radar attitude information used for fitting the radar attitude model. The model fitting can be performed using methods such as direct least squares fitting or robust fitting based on random sampling (e.g., RANSAC).

[0052] A radar attitude model describes the relationship between radar attitude and radar detection information of multiple target points, representing a mathematical model of the relative motion between the radar and the target points. This model can include multiple model parameters. Fitting the radar attitude model determines the values ​​of these model parameters. All or some of these parameters correspond to and determine the radar attitude information of the current frame; that is, the model parameters can be radar attitude information. Different models correspond to different model parameters. The process of fitting the model and determining the corresponding radar attitude information is also called radar attitude information estimation.

[0053] In practical applications, the target point dynamic and static identification method in this embodiment can utilize the radar's radio frequency front-end to receive echo signals. These echo signals are then processed by the radar's processor to generate target point cloud data, thereby performing the dynamic and static identification steps. This method not only enables both dynamic and static identification of radar target points but also reduces the computational load.

[0054] In some exemplary embodiments, radar attitude information includes: radar velocity, radar azimuth angle; or, radar velocity, radar azimuth angle, and radar elevation angle, etc. In the embodiments of this application, the first radar attitude information, the radar attitude information corresponding to the current frame, the radar attitude information corresponding to the historical frame, and the radar attitude information collected by sensors outside the radar system are all radar attitude information, corresponding to data with the same or different values ​​determined at different times, from different sources, or by different methods.

[0055] In some exemplary embodiments, the model includes multiple model parameters: radar velocity and radar azimuth angle; or, the model includes multiple model parameters: radar velocity, radar azimuth angle, and radar elevation angle. Successful fitting means that the fitting has determined these model parameter values, thereby determining the radar's attitude information. It can be understood that the model parameter values ​​are all fitting results, also called estimated values, corresponding to: radar velocity estimate, radar azimuth angle estimate, and radar elevation angle estimate.

[0056] For example, the radar attitude model is a target point velocity-azimuth cosine function, and the model parameters include: radar velocity estimate. Radar azimuth estimate

[0057] Where, x i The radar detection information for the i-th target point includes: elevation angle. Azimuth θ i Doppler velocity v i .

[0058] In some exemplary embodiments, the above model can be extended to three-dimensional attitude estimation when the radar attitude includes pitch angle.

[0059] Each frame of radar signal corresponds to radar detection information for multiple target points, x i,t The radar detection information for the i-th target point in the t-th frame includes: elevation angle. Azimuth θ i,t Doppler velocity v i,t x i,t-1 The radar detection information for the i-th target point in the (t-1)-th frame includes: elevation angle. Azimuth θ i,t-1 Doppler velocity v i,t-1 If frame t is the current frame, then frame t-1 is the previous frame. Without distinguishing which frame number it is, all frames represent the current frame data.

[0060] Optionally, the radar attitude model can also be other forms of models or functions, not limited to the cosine function exemplified in the embodiments of this application. In this embodiment, the target point velocity-azimuth cosine function is used as an example to illustrate aspects related to model fitting; this fitting process is also called cosine fitting.

[0061] In some exemplary embodiments, radar point cloud data is obtained by detecting radar signals in each frame, including radar detection information corresponding to multiple target points.

[0062] In some exemplary embodiments, the radar detection information of the target point includes one or more of the following: elevation angle, azimuth angle, and Doppler velocity;

[0063] The plurality of candidate stationary target points are multiple target points among the target points detected in the current frame that satisfy one or more of the following constraints:

[0064] Constraint 1: The pitch angle of the target point falls within the set pitch angle threshold range;

[0065] Constraint 2: The Doppler velocity of the target point is less than the difference between the set Doppler velocity threshold and the first velocity; wherein, the first velocity is the ground motion velocity of the target point determined based on the first radar attitude information;

[0066] Constraint 3: The azimuth angle of the target point falls within the first azimuth angle range; wherein, the first azimuth angle range is the azimuth angle range of a stationary target determined based on the first radar attitude information under the set maximum unambiguous velocity condition.

[0067] In some exemplary embodiments, the pitch angle threshold range refers to a pitch angle whose absolute value is less than a set pitch angle threshold. For example, the pitch angle threshold is 10°, or other set values. The pitch angle threshold range is primarily related to the pitch angle to be measured. It is understood that, generally speaking, distant target points typically have relatively small pitch angles.

[0068] For example, the pitch angle threshold is Constraint 1 corresponds to:

[0069] Doppler velocity threshold v c,th Constraint 2 corresponds to:

[0070] As can be seen, constraint 2 is determined based on the relationship between the target point's azimuth and Doppler velocity, and the radar's velocity and azimuth. Using the radar's velocity from the previous frame... and azimuth An approximation of the velocity and azimuth of the current frame, obtained through this approximation and a more relaxed Doppler threshold v. c,thThis allows for a rough selection of target points. The Doppler threshold is set based on experience or application requirements, for example, v. c,th =0.5m / s, or other set values.

[0071] Maximum unambiguous speed v max Constraint 3 corresponds to:

[0072] or

[0073] Where, θ tol The preset azimuth angle estimation error is determined based on analysis of actual measurement data or experience. For example, θ tol =3°, or other settings.

[0074] That is, the range of the first azimuth angle corresponds to greater than or less The point whose azimuth angle satisfies condition inequality (4) in frame t (current frame) satisfies constraint condition 3.

[0075] As can be seen, by filtering all target points detected in the current frame based on one or more of the constraints, more targeted target points are selected for model fitting, which can effectively limit the number of points involved in the fitting. Among them, constraint 2 filters the target points based on the Doppler velocity, and constraint 3 filters the target points based on the azimuth angle, selecting approximately stationary target points for subsequent fitting.

[0076] In some exemplary embodiments, the method further includes:

[0077] Step 140: If the fitting fails, perform dynamic and static identification of the target points detected in the current frame based on the first radar attitude information.

[0078] Accordingly, if the fit is successful, proceed to step 130.

[0079] It is understandable that when the model fitting fails to meet the judgment condition for successful fitting, it is determined that the model fitting for the current frame has failed, and thus the latest radar attitude information cannot be determined. In this case, the attitude information obtained by fitting based on historical frame data or the attitude information collected from outside the radar system is used to participate in the dynamic and static identification of the target point in the current frame.

[0080] In some exemplary embodiments, as shown in FIG2, step 120 includes:

[0081] Repeat the following fitting steps until the loop exit condition is met:

[0082] Step 1210: Using the first radar attitude information as the initial fitting value, the radar attitude model is fitted according to the radar detection information of the multiple candidate stationary target points to determine the first model parameter value corresponding to the current frame.

[0083] Step 1220: Based on the fitted radar attitude model, determine the in-class points among the multiple candidate stationary target points;

[0084] Step 1240: Using the first radar attitude information as the initial fitting value, the radar attitude model is fitted again based on the in-class points to determine the second model parameter value corresponding to the current frame, and the error index of the current radar attitude model is calculated based on the in-class points.

[0085] The loop exit condition includes one of the following: reaching a preset maximum number of loops; or the error index being less than or equal to a preset error index threshold.

[0086] In some exemplary embodiments, step 110 further includes: determining whether the number of the plurality of candidate stationary target points is greater than or equal to a candidate point number threshold; if so, continuing to execute step 120.

[0087] In some exemplary embodiments, step 110 further includes: if it is determined that the number of the plurality of candidate stationary target points is less than a candidate point count threshold, relaxing the constraint conditions and selecting a plurality of candidate stationary target points again from the target points detected in the current frame. Thus, by relaxing the constraint conditions, the number of (candidate stationary) target points that meet the conditions reaches a preset candidate point count threshold.

[0088] In this embodiment of the disclosure, exiting the loop when the loop exit condition is met indicates the execution result of the entire loop corresponding to step 120; exiting the current loop indicates exiting the current single iteration, the current loop ends, and the next loop will be executed again.

[0089] In some exemplary embodiments, the fitting step, which is performed cyclically, further includes:

[0090] Step 1230: Determine whether the number of intra-class points among the multiple candidate static target points is less than the intra-class point number threshold d.

[0091] Accordingly, if the number of intra-class points among the multiple candidate static target points is less than the intra-class point number threshold d, the current loop is exited, that is, the current single iteration is exited, and the fitting step is executed again in the next loop; if it is greater than or equal to, step 1240 is executed.

[0092] In some exemplary embodiments, the fitting step, which is performed cyclically, further includes:

[0093] Step 1250: Determine whether the error index determined in this iteration is greater than the current minimum value of the error index, also known as the optimal value;

[0094] Accordingly, if the error index determined in this iteration is not greater than the current minimum value, exit the current iteration and proceed to the next iteration to execute the fitting step again; if it is greater, continue with the subsequent steps 1260.

[0095] Step 1260: Update the minimum value of the error index to the error index determined in this iteration.

[0096] In some exemplary embodiments, if the error index determined in the current loop is less than the current minimum value, the minimum value of the error index is updated to the error index determined in the current loop. It can be understood that in each loop executed for the current frame data, after calculating the error index for the current loop, it is compared with the original minimum value. After the process judgment in step 1250, the latest minimum value of the error index is updated for subsequent judgment on whether to exit the current loop.

[0097] In some exemplary embodiments, the fitting step, which is performed cyclically, further includes:

[0098] Step 1200: Determine if the loop exit condition is met. If not, proceed to step 1210. If met, proceed to step 1270.

[0099] Step 1270: Determine whether the model fitting for the current frame is successful, and then exit the loop.

[0100] Step 1200 can be executed before the start of each loop or after the end of each loop, and is not limited to the aspect shown in Figure 2.

[0101] In some exemplary embodiments, step 1270 includes: determining that the fitting is successful if the current error index is less than or equal to a preset error index threshold, and determining that the fitting is unsuccessful if the current error index is greater than the preset error index threshold.

[0102] In the iterative fitting process of fitting a model to determine the corresponding model parameters for the current frame, the loop can exit either upon reaching an error threshold or upon reaching the maximum number of iterations. Reaching the error threshold and exiting the loop signifies successful fitting, meaning the radar attitude model obtained from the last iteration of the fitting for the current frame meets the error threshold requirement; correspondingly, the radar attitude information for the current frame can be determined based on the model parameters. Exiting the loop simply because the maximum number of iterations has been reached indicates unsuccessful fitting.

[0103] In some exemplary embodiments, the error index is the mean squared error (MSE). The error index of the current radar attitude model is calculated based on the in-class points, i.e., the MSE of the radar-target point operational relationship model obtained by refitting is calculated on these in-class points. The minimum MSE is recorded. It can be seen that the smaller the MSE, the more accurate the fitted model. The fitting step is executed iteratively until the minimum error index meets the set error index threshold, or when the maximum number of iterations has been reached, the loop exits. In some exemplary embodiments, the model parameters of the radar attitude model include: radar attitude information;

[0104] Step 1210 includes: setting the solution space range of the model parameters of the radar attitude model based on the first radar attitude information and the attitude change amplitude threshold;

[0105] Based on the radar detection information of the multiple candidate stationary target points, a radar attitude model is fitted to determine the first model parameter value corresponding to the current frame.

[0106] The attitude change threshold includes one or more of the following: velocity change threshold and azimuth change threshold. Correspondingly, the set solution space range includes one or more of the following: velocity solution range and azimuth solution range. It can be understood that radar attitude information changes little in a short time (e.g., within 50ms), the radar signal continuously detects targets, and the time interval between consecutive data frames is short. Therefore, setting the solution space range for model fitting based on the radar attitude information obtained in a short time (first radar attitude information) combined with the variable range in a short time can effectively improve fitting accuracy and stability. It can be understood that the solution space range reflects the constraints on the actual velocity and heading of the radar (or vehicle), and can also be understood as reflecting the constraints on the actual velocity and heading change range between two consecutive radar (or vehicle) frames, representing a suitable solution space range. For example, at lower speeds, the velocity change between two consecutive radar (or vehicle) frames is ±5m / s, and the azimuth change is ±5°. At higher speeds, other ranges correspond. Specific values ​​can be dynamically determined based on relevant factors and are not limited to the examples.

[0107] In some exemplary embodiments, the velocity change threshold is a maximum velocity increase or decrease value or a maximum increase or decrease percentage; the azimuth change threshold is a set maximum angle value or a maximum increase or decrease percentage. The velocity solution range corresponds to the velocity range determined based on the velocity and velocity change threshold in the first radar attitude information, and the azimuth solution range corresponds to the azimuth range determined based on the azimuth and azimuth change threshold in the first radar attitude information.

[0108] In some exemplary embodiments, the attitude change magnitude threshold also includes a pitch angular velocity change threshold.

[0109] In some exemplary embodiments, step 1210, which involves fitting a radar attitude model based on the radar detection information of the plurality of candidate stationary target points, includes:

[0110] From the P candidate stationary target points selected in step 110, Q candidate stationary target points are randomly selected, and radar attitude model fitting is performed based on the radar detection information of the Q candidate stationary target points; where P is greater than Q and both are greater than 1.

[0111] During the multiple iterations, in each iteration, step 1210 randomly selects a predetermined number Q candidate stationary target points from the multiple candidate stationary target points obtained in step 110. The number Q of randomly selected candidate stationary target points satisfies the target point requirement of the model fitting algorithm, and the specific number may vary for different algorithms.

[0112] In some exemplary embodiments, radar attitude model fitting is performed in steps 1210 and 1240 using one of the following methods:

[0113] Newton's algorithm, LM algorithm, and Doggleg algorithm.

[0114] In some exemplary embodiments, the radar attitude model is a target point velocity-azimuth cosine function, and the Newton algorithm, the LM (Levenberg-Marquardt) algorithm, or the Doglg algorithm is used to perform cosine fitting to determine the model parameters of the target point velocity-azimuth cosine function.

[0115] In some exemplary embodiments, step 1220 employs the Random Sample Consensus (RANSAC) algorithm for in-class point identification. In-class points are stationary target points, and non-in-class points are non-stationary target points. That is, step 1220 further determines more accurate candidate stationary target points from the candidate stationary target points for model fitting again. It can be seen that both model fitting and in-class point identification are based on candidate stationary target points. Compared to randomly using radar-identified target point data to execute the corresponding algorithm, this method helps to accelerate convergence and reduce computational resource consumption.

[0116] In some exemplary embodiments, when the Hessian matrix of the target point velocity-azimuth cosine function is positive definite, Newton's method is used for fitting: P new =P old -(J T J) -1 J T f(P old (5);

[0117] In the case of non-positive definiteness, the LM method is used for fitting: P new =P old -(J T J+λI) -1 J T f(P old (6);

[0118] Among them, P old Let P be the initial vector. new The new vector determined after fitting is f(), where f() is the target point velocity-azimuth cosine function, and J is the Jacob matrix of the target point velocity-azimuth cosine function. T λ is the transpose of the Jacob matrix, where I is the identity matrix and λ is the coefficient.

[0119] The target point velocity-azimuth cosine function is:

[0120] Matrix J is:

[0121] Where, x i The radar detection information for the i-th target point is known information; the radar attitude information is a variable, including v. i and θ i The radar detection information and radar attitude information form a 2×1 vector P, i.e., P old P new .

[0122] In some exemplary embodiments, the approximate matrix J of the Hessian matrix of the velocity-azimuth cosine function at the target point is... T When J is positive definite, the Newton method shown in (5) above is used for fitting; the approximate matrix J of the Hessian matrix of the target point velocity-azimuth cosine function is... T When J is non-positive definite, the LM method shown in (6) above is used for fitting.

[0123] In some exemplary embodiments, step 130 includes: determining a target point that satisfies the following conditions as a motion target point:

[0124] Maximum unambiguous speed v max The Leyen velocity v of the i-th target point in the t-th frame (current frame) i,t The azimuth angle θ of the i-th target point in the t-th frame (current frame) i,t The radar velocity estimate in the radar attitude information (first radar attitude information) of frame t. The radar azimuth estimate in the radar attitude information (first radar attitude information) of frame t. k is the maximum blur factor, v d,th The preset speed threshold is used to determine whether it is a stationary target point. If the target point satisfies (8), it is a moving target point; otherwise, it is a stationary target point.

[0125] In some exemplary embodiments, k = 1 or 2, or other values.

[0126] This application also provides a method for identifying the dynamic and static states of target points, as shown in Figure 2 or Figure 3, including:

[0127] Step 310: Record the number of frames that failed to fit consecutively;

[0128] Step 320: If the number of frames is greater than or equal to a preset failure count threshold, relax the constraints to allow for candidate stationary target point selection in the next frame.

[0129] In some exemplary embodiments, the preset failure count threshold in step 320 is also referred to as the first failure count threshold.

[0130] The radar system continuously transmits data frames for target point detection, performing dynamic and static target point identification for each detected target point. This includes identifying dynamic and static target points using the radar attitude information determined by the successfully fitted model when model fitting is successful, and identifying dynamic and static target points using the radar attitude information determined by the successfully fitted model in a previous frame when model fitting fails. If model fitting fails for multiple consecutive frames, and the number of consecutive failures reaches a failure threshold, the constraints in step 110 for candidate target point selection are adjusted to increase the number of candidate static target points for processing the next frame of data.

[0131] In some exemplary embodiments, relaxing the constraints includes at least one of the following:

[0132] Increase the pitch angle threshold range in constraint 1; for example, increase the pitch angle threshold to...

[0133] Increase the Doppler velocity threshold v in constraint 2 c,th ;

[0134] Increase the threshold d for the number of points within a class;

[0135] Increase the maximum number of loops.

[0136] In some exemplary embodiments, the radar detection information further includes: distance;

[0137] Step 120 includes:

[0138] Step 1201: Sort the multiple candidate stationary target points according to the distance to the target point, and select the nearest set number of candidate stationary target points;

[0139] Step 1210: Using the first radar attitude information as the initial value for fitting, the radar attitude model is fitted based on the radar detection information of the selected multiple candidate stationary target points.

[0140] or,

[0141] Step 1201: Select multiple candidate stationary target points from the multiple candidate stationary target points whose distance is less than a set distance threshold;

[0142] Step 1210: Using the first radar attitude information as the initial value for fitting, the radar attitude model is fitted based on the radar detection information of the selected multiple candidate stationary target points.

[0143] As can be seen, for the candidate stationary target points initially obtained in step 110, further prioritizing the selection of target points that are closer means further selecting points with larger azimuth angles, which helps the model fit converge faster and accelerates the fitting process.

[0144] In step 120 of some exemplary embodiments, during the model fitting process based on candidate stationary target points, one or more of the following reselection steps may be included, without conflict:

[0145] Randomly select some target points, select some target points based on a distance threshold, or select some target points sorted by distance.

[0146] In cases involving multiple reselection steps, the order of these steps is not limited to any particular aspect; subsequent reselection steps can be performed based on the results of previous selections. After reselection, the target point data participating in model fitting should meet the required number of points for model fitting.

[0147] In some exemplary embodiments, the radar system carrier is a vehicle, and the first radar attitude information is determined based on vehicle motion information collected by sensors configured on the vehicle;

[0148] or,

[0149] The first radar attitude information is the radar attitude information corresponding to the latest successfully fitted frame before the current frame;

[0150] or,

[0151] The first radar attitude information is obtained by filtering the radar attitude information corresponding to multiple successfully fitted frames before the current frame.

[0152] The determination based on vehicle motion information collected by sensors mounted on the vehicle includes: determining the first radar attitude information based on information such as vehicle speed / heading collected by the sensors, combined with radar installation information.

[0153] In some exemplary embodiments, filtering is performed on the radar attitude information corresponding to multiple successfully fitted historical frames, including:

[0154] After applying Kalman or particle filtering to the radar attitude information corresponding to multiple historical frames, the first radar attitude information is obtained.

[0155] or,

[0156] After smoothing and filtering the radar attitude information corresponding to multiple historical frames, the first radar attitude information is obtained.

[0157] In some exemplary embodiments, the first radar attitude information is obtained by weighted averaging the radar attitude information determined by fitting and the radar attitude information determined by vehicle motion information.

[0158] The radar attitude information determined by the fitting includes: the first radar attitude information is the radar attitude information corresponding to the latest successfully fitted frame before the current frame; or, the first radar attitude information is the radar attitude information obtained after filtering the radar attitude information corresponding to multiple successfully fitted frames before the current frame.

[0159] This application also provides a method for dynamic and static target point identification, applied in a radar. The radar includes a radio frequency transceiver and a processor. The processor is configured to process signals received by the radio frequency transceiver to obtain target detection data. The target detection data includes the Doppler velocity and elevation angle of multiple target points. The method includes:

[0160] Obtain the target detection data for the current frame;

[0161] Based on the first radar attitude information, multiple candidate stationary target points are selected from the target detection data of the current frame, according to a set pitch angle range and a set Doppler velocity range:

[0162] Using the first radar attitude information as the initial fitting value, a preset algorithm is used to fit the radar attitude model to the plurality of candidate stationary target points, generating the radar attitude information corresponding to the current frame; and,

[0163] Based on the generated radar attitude information, all target points in the current frame are classified into dynamic and static categories, and target point cloud data containing dynamic and static labels are output.

[0164] In some exemplary embodiments, the set pitch angle range is a range where the absolute value of the pitch angle is less than 10°, and the Doppler velocity range is a range where the absolute value of the Doppler velocity is less than 0.5 m / s. Optionally, other ranges may be determined as needed.

[0165] In some exemplary embodiments, the preset algorithm includes the LM algorithm.

[0166] This application also provides a radar signal processing method, as shown in Figure 4, including:

[0167] Step 410, classify the target points; and

[0168] Step 420: Different processing methods are used for post-processing of target points of different categories.

[0169] This radar signal processing method is applied to radar systems, where target points are classified based on the fitting results of radar attitude models.

[0170] Accordingly, this disclosure also provides a radar system, which may include a processor and a memory. The radar system can be configured to execute the radar signal processing method described above based on the processor and memory, so as to reduce the consumption of computing resources and improve the real-time performance of the radar through dynamic and static identification.

[0171] In some exemplary embodiments, the target points are categorized as: moving target points and stationary target points;

[0172] The post-processing of different categories of target points using different processing methods includes at least one of the following:

[0173] Perform one or more of the following post-processing steps on the moving target point: de-velocity fuzzing, clustering, and tracking;

[0174] For stationary target points, none of the following post-processing steps are performed: de-velocity fuzzing, clustering, and tracking.

[0175] It is understandable that for functions such as target velocity detection and target tracking, the main focus is on the motion characteristics of the target, and correspondingly, more data processing is essentially focused on moving target points. Conversely, for functions such as contour recognition and spatial construction of stationary targets, more data processing is essentially focused on stationary target points. Therefore, after identifying the motion / static status of the target points detected in each frame, different post-processing steps can be performed depending on the application's functional requirements. In some embodiments, only the identified moving target points undergo velocity de-fuzzing, clustering, and tracking processing, while these post-processing steps are not performed on stationary target points. This effectively reduces the number of points involved in velocity de-fuzzing, reduces computational resource requirements, and improves computational speed. In other embodiments, only the identified stationary target points undergo pre-defined processing steps, while these pre-defined processing steps are not performed on moving target points. As can be seen, the above post-processing steps have high computational resource requirements, especially for radar point cloud data. Performing radar signal processing such as de-ambiguation, clustering, and tracking on all target point data detected in the current frame places high demands on computational resources and imposes high requirements on equipment cost and performance. By adopting the method provided in this application, which distinguishes between static and dynamic target points before performing post-processing, the amount of data processed in different post-processing steps can be effectively reduced, the computational resource requirements can be lowered, and the radar signal processing performance can be improved.

[0176] In some exemplary embodiments, step 410 includes: determining the radar attitude information corresponding to the current frame based on the first radar attitude information, and classifying the target point according to the radar attitude information corresponding to the current frame;

[0177] The first radar attitude information is obtained based on at least one of the following:

[0178] The radar attitude information corresponding to historical frames and the radar attitude information collected in real time by the radar system's carrying equipment.

[0179] When obtaining the first radar attitude information based on the radar attitude information corresponding to historical frames, at least one of the following methods is used to fit the radar attitude model in order to determine the radar attitude information corresponding to the current frame:

[0180] The radar attitude information is used as the initial value for fitting, and the radar attitude model is fitted.

[0181] The radar attitude model is fitted based on the set velocity or azimuth solution space range; or, the radar attitude model is fitted based on the set velocity and azimuth solution space range.

[0182] Select a subset of target points from the target points detected in the current frame to fit the radar attitude model;

[0183] If the radar attitude model fails to fit for multiple consecutive frames, the radar attitude model is fitted based on all target points detected in the current frame.

[0184] Using the first radar attitude information as the initial value for fitting can accelerate the fitting process, improve the accuracy and computational efficiency of radar attitude information estimation for the current frame, and thus improve the accuracy of motion / static identification. Fitting the radar attitude model based on a set velocity or azimuth solution space range, or based on a set velocity and azimuth solution space range, enables the fitting algorithm to converge quickly and improves the fitting success rate. Selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting, and reasonably selecting the target points involved in the fitting, can reduce the amount of data required for fitting calculations and lower the computational resource requirements.

[0185] If the number of consecutive failed fitting frames exceeds the second failure threshold, the radar attitude model fitting in step 410 is determined to be a failure. In this case, target point selection is not performed during the processing of the next frame's radar signal; instead, the radar attitude model is fitted based on all target points detected in the previous frame. The second failure threshold is greater than or equal to the first failure threshold.

[0186] In some exemplary embodiments, the step of selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting includes:

[0187] Based on the first radar attitude information, multiple candidate stationary target points are selected from the target points detected in the current frame for radar attitude model fitting.

[0188] The plurality of candidate stationary target points are multiple target points among the target points detected in the current frame that satisfy one or more of the following constraints:

[0189] The pitch angle of the target point is within the pitch angle threshold range;

[0190] The Doppler velocity of the target point is less than the difference between a set Doppler velocity threshold and a first velocity; wherein, the first velocity is the ground motion velocity of the target point determined based on the first radar attitude information;

[0191] The azimuth of the target point falls within the first azimuth range; wherein, the first azimuth range is the azimuth range of a stationary target determined based on the first radar attitude information under the set maximum unambiguous velocity condition.

[0192] In some exemplary embodiments, the step of selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting includes:

[0193] Select multiple candidate stationary target points from the target points detected in the current frame, and then randomly select some of the candidate stationary target points from them;

[0194] Based on the first radar attitude information, a radar attitude model is fitted using a selected subset of candidate stationary target points.

[0195] Optionally, the step of selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting includes:

[0196] Select multiple candidate stationary target points from the target points detected in the current frame, and then select the candidate stationary target points that are closest to each other from them;

[0197] Based on the first radar attitude information, a radar attitude model is fitted using a selected subset of candidate stationary target points.

[0198] During the process of fitting the model based on all or part of the target points detected in the current frame, the fitting steps are performed using one or more of the methods described above, provided there are no conflicts.

[0199] Among them, the nearest candidate stationary target points are selected, including: selecting candidate stationary target points whose distance is less than a set distance threshold; or selecting a set number of candidate stationary target points that are closest after being sorted by distance.

[0200] In some exemplary embodiments, step 410 includes classifying the target points using the target point dynamic and static recognition method described in any embodiment of this application.

[0201] This application also provides a radar signal processing method applied to a frequency modulated continuous wave (FMCW) millimeter-wave radar. The FMCW millimeter-wave radar includes a radio frequency (RF) front-end and a processor. The RF front-end is configured to transmit and receive radar signals, and the processor is configured to perform signal processing on the received radar signals to obtain point cloud data. The method includes:

[0202] The point cloud data is classified into target points to generate a set of moving target points and a set of stationary target points;

[0203] Perform occupancy raster map construction on a set of stationary target points, and output location information that can be used for map construction; and

[0204] Perform Doppler velocity deblurring, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm processing, and Kalman filter tracking processing on the set of moving target points, and output one or more of the following: target identification, position, velocity, and trajectory history.

[0205] In some exemplary embodiments, the Doppler velocity deblurring process can be performed using a velocity de-analog algorithm based on the radar's own velocity. In some exemplary embodiments, the clustering parameters of the DBSCAN clustering algorithm include: minimum number of points and neighborhood radius, etc.; for example, the minimum number of points is set to 3, and the neighborhood radius is set to 0.5m. Kalman filter tracking processing is used to output the target trajectory.

[0206] In some exemplary embodiments, the target point dynamic and static recognition method or the radar signal processing method can obtain the corresponding recognition result or achieve the corresponding processing target by executing relevant steps in an embedded system or a radar processor.

[0207] This application also provides an integrated circuit, including a processor 510, which is configured to implement the target point dynamic and static identification method as described in any embodiment of this application; or, the processor is configured to implement the radar signal processing method as described in any embodiment of this application.

[0208] In some exemplary embodiments, the integrated circuit is a system-on-a-chip (SoC) chip, and the processor includes an MCU (microcontroller unit) or a digital signal processor (DSP).

[0209] In some exemplary embodiments, as shown in FIG5, the integrated circuit further includes: a radio frequency module 520, an analog signal processing module 530, a digital signal processing module 540, and a radar digital function module 550 connected in sequence;

[0210] The radio frequency module 520 is used to generate radio frequency transmission signals and receive radio frequency reception signals; the analog signal processing module 530 is used to down-convert the radio frequency reception signals to obtain intermediate frequency signals; the digital signal processing module 550 is used to perform analog-to-digital conversion on the intermediate frequency signals to obtain digital signals.

[0211] The radar digital function module 550 includes a processor 510 for executing the target point dynamic and static identification method based on the digital signal, or executing the radar signal processing method based on the digital signal.

[0212] In some exemplary embodiments, the radar digital function module 550 is also used to process digital signals to achieve target detection or wireless communication. For example, the integrated circuit may be a radar or an ultra-wideband UWB chip or die.

[0213] In some exemplary embodiments, the integrated circuit is a millimeter-wave radar chip, and the types of radar digital function modules in the integrated circuit can be determined according to actual needs. For example, in a millimeter-wave radar chip, the radar digital function modules can be used for range Vidoff transform, velocity Vidoff transform, constant false alarm rate detection, direction of arrival detection, point cloud processing, etc.; or, they can be used to acquire information such as the target's range, horizontal angle, pitch angle, velocity, altitude, and micro-Doppler motion characteristics, and can also further generate information such as the target's shape, size, surface roughness, and dielectric properties based on the above-mentioned target-related information.

[0214] In some exemplary embodiments, the integrated circuit may be an AiP (Antenna-In-Package) chip structure, an AoP (Antenna on Package) chip structure, an AoC (Antenna on Chip) chip structure, or a RoP (Radiator on / through Package) structure. Optionally, the RoP structure may involve setting a radiating structure on the chip package and surrounding the radiating structure with solder balls to form an air waveguide structure. That is, the radio frequency (RF) signal generated by the chip can be transmitted to an external antenna through the aforementioned radiating structure, the air cavity waveguide structure surrounded by the solder balls, and the air waveguide built into the PCB board, so as to radiate towards the target area.

[0215] In some exemplary embodiments, the integrated circuit further includes: a monolithic microwave integrated circuit (MMIC) chip for modulation, transmission, reception of millimeter-wave signals, and demodulation of echo signals; the processor includes an MCU for executing the target point dynamic and static identification method based on the output data of the MMIC chip, or for executing the radar signal processing method based on the output data of the MMIC chip.

[0216] This application also provides an electromagnetic wave device, including:

[0217] A carrier; an integrated circuit as described in any embodiment of the present application is disposed on the carrier;

[0218] An antenna is disposed on the carrier, or the antenna and the integrated circuit are integrated into a single device and disposed on the carrier.

[0219] The integrated circuit is connected to the antenna, which is used to transmit radio frequency signals and receive radio frequency signals.

[0220] In some exemplary embodiments, the electromagnetic wave device may include: a carrier, an integrated circuit as described in any of the above embodiments, and an antenna, etc. The integrated circuit may be disposed on the carrier; the antenna may be disposed on the carrier (i.e., the antenna may be an antenna disposed on a PCB board in a structure such as RoP), or it may be integrated with the integrated circuit as a single device disposed on the carrier (i.e., the antenna may be an antenna disposed in a structure such as AiP, AoP, or AoC); wherein, the integrated circuit is connected to the antenna (i.e., the sensing chip or integrated circuit does not integrate an antenna, such as a conventional SoC), and is used to transmit and receive electromagnetic wave signals. The carrier may be a printed circuit board (PCB).

[0221] In some exemplary embodiments, the aforementioned electromagnetic wave signals may be centimeter-wave or millimeter-wave signals (such as 3.1GHz, 24GHz, 60GHz, 77GHz, 94GHz, 120GHz, 140GHz, 220GHz, 250GHz, etc.). Specifically, the 3.1GHz centimeter-wave signal may include 3.1GHz to 10.6GHz, such as 3.1GHz, 5GHz, 6GHz, 8GHz, 10.6GHz, etc., or 7.163-8.812GHz, etc.; the 77GHz millimeter-wave signal may include signals from 76GHz to 81GHz, such as the frequency ranges of 76GHz to 77GHz, 77GHz to 79GHz, 79GHz to 81GHz, etc., or fixed frequency points such as 76GHz, 77GHz, 78GHz, 79GHz, 80GHz, 81GHz, etc.

[0222] This application embodiment also provides a terminal device, including:

[0223] One or more processors;

[0224] Storage device for storing one or more programs.

[0225] When the one or more programs are executed by the one or more processors, the one or more processors implement the target point dynamic and static identification method as described in any embodiment of this application; or, are configured to implement the radar signal processing method as described in any embodiment of this application.

[0226] In some exemplary embodiments, the terminal device includes a millimeter-wave radar detection device. For example, the millimeter-wave radar detection device is integrated into an intelligent driving system and installed in a vehicle.

[0227] This application also provides a terminal device, including: a device body; and an electromagnetic wave device disposed on the device body as described in any embodiment of this application; wherein the electromagnetic wave device is used for target detection or communication to provide reference information to the operation of the device body or other electronic devices disposed in the device body.

[0228] In some exemplary embodiments, the electromagnetic wave device may be disposed outside or inside the device body. In other optional embodiments of this application, the electromagnetic wave device may be partially disposed inside the device body and partially disposed outside the device body. This application does not limit the specific implementation of the electromagnetic wave device; the choice depends on the specific circumstances.

[0229] In some alternative embodiments, the aforementioned device body can be a component or product applied in fields such as smart cities, smart homes, transportation, smart homes, consumer electronics, security monitoring, industrial automation, in-cabin detection (such as smart cockpits), medical devices, and healthcare. For example, the device body can be intelligent transportation equipment (such as automobiles, bicycles, motorcycles, ships, subways, trains, etc.), security equipment (such as cameras), liquid level / flow rate detection equipment, smart wearable devices (such as wristbands, glasses, watches, etc.), smart home devices (such as robot vacuum cleaners, door locks, televisions, air conditioners, smart lights, etc.), various communication and office equipment (such as mobile phones, tablets, computers, air mice, keyboards, projectors, etc.), as well as devices such as barriers, intelligent traffic lights, intelligent signs, traffic cameras, and various industrial robotic arms (or robots). It can also be various instruments used to detect vital signs parameters and various devices equipped with such instruments, such as in-cabin vital sign detection in automobiles, indoor personnel monitoring, smart medical devices, and consumer electronic devices.

[0230] For example, when the aforementioned device is a vehicle, the electromagnetic wave device used as an in-vehicle sensor can be used to assist ADAS systems in realizing in-vehicle applications such as adaptive cruise control, automatic braking assist (AEB), blind spot detection warning (BSD), lane change assist warning (LCA), rear cross traffic alert (RCTA), assisted / automatic parking assist, rear vehicle warning, collision avoidance, pedestrian detection, as well as door collision avoidance, automatic opening and closing of the trunk door with foot kick, etc. It can also be used as a digital key for the vehicle.

[0231] The target point motion / static identification scheme and the radar signal processing scheme based on the identification results provided in this application embodiment determine the radar attitude information by first screening candidate stationary target points based on historical frames or radar attitude information collected externally to the radar system for current frame model fitting. This effectively reduces the amount of data involved in the fitting and accelerates fitting convergence, improving the model fitting accuracy and success rate. According to the motion / static identification method provided in this application embodiment, identification can be performed before velocity ambiguity is resolved, avoiding the reliance on de-ambiguous data in some feasible schemes. This reduces the overall identification scheme's computational resource requirements and improves identification efficiency. Based on motion / static identification, different types of target points undergo differentiated digital signal processing, such as velocity ambiguity resolution, clustering, or tracking. This effectively reduces the execution speed of different application functions and improves the overall performance of the radar system.

[0232] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0233] Although embodiments of this disclosure have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for identifying the dynamic and static states of a target point, applied to a radar system, comprising: Based on the first radar attitude information, select multiple candidate stationary target points from the target points detected in the current frame; Using the first radar attitude information as the initial value for fitting, the radar attitude model is fitted according to the radar detection information of the multiple candidate stationary target points to determine the radar attitude information corresponding to the current frame. Based on the radar attitude information corresponding to the current frame, the target points detected in the current frame are identified as dynamic or static. The first radar attitude information is obtained based on at least one of the following: The radar attitude information corresponding to historical frames and the radar attitude information collected in real time by the radar system's carrying equipment.

2. The target point dynamic and static recognition method according to claim 1, The radar detection information includes one or more of the following: elevation angle, azimuth angle, and Doppler velocity; The plurality of candidate stationary target points are multiple target points among the target points detected in the current frame that satisfy one or more of the following constraints: The pitch angle of the target point is within the pitch angle threshold range; The Doppler velocity at the target point is less than the difference between the set Doppler velocity threshold and the first velocity; where, The first speed is the ground-based speed of the target point determined based on the first radar attitude information; The azimuth of the target point falls within the first azimuth range; wherein, the first azimuth range is the azimuth range of a stationary target determined based on the first radar attitude information under the set maximum unambiguous velocity condition.

3. The target point dynamic and static recognition method according to claim 1 or 2, The method further includes: If the fitting fails, the target points detected in the current frame are identified dynamically or statically based on the first radar attitude information.

4. The target point dynamic and static recognition method according to claim 1 or 2, The step of fitting a radar attitude model using the first radar attitude information as the initial value and based on the radar detection information of the plurality of candidate stationary target points includes: The following fitting steps are performed iteratively in the embedded system or radar processor until the loop exit condition is met: Using the first radar attitude information as the initial value for fitting, the radar attitude model is fitted according to the radar detection information of the multiple candidate stationary target points to determine the first model parameter value corresponding to the current frame. Based on the fitted radar attitude model, determine the in-class points among the multiple candidate stationary target points; Using the first radar attitude information as the initial fitting value, the radar attitude model is fitted again based on the in-class points to determine the second model parameter value corresponding to the current frame, and the error index of the current radar attitude model is calculated based on the in-class points. The loop exit condition includes one of the following: reaching a preset maximum number of loops; or the error index being less than or equal to a preset error index threshold.

5. The target point dynamic and static recognition method according to claim 4, The method further includes: If the number of in-class points among the multiple candidate static target points is less than the in-class point number threshold, the current loop is exited, and the fitting step is executed again in the next loop.

6. The target point dynamic and static recognition method according to claim 4, The method further includes: Update the minimum value of the error index; If the error index determined in this iteration is not greater than the current minimum value, exit the current iteration and proceed to the next iteration to execute the fitting step again.

7. The target point dynamic and static recognition method according to claim 2, The method further includes: Record the number of consecutive frames that failed to fit; If the number of frames is greater than or equal to a preset failure threshold, the constraints are relaxed to allow for candidate stationary target point selection in the next frame.

8. The target point dynamic and static recognition method according to claim 4, The model parameters of the radar attitude model include: Radar attitude information; The step of using the first radar attitude information as the initial fitting value, and fitting the radar attitude model based on the radar detection information of the multiple candidate stationary target points to determine the first model parameter value corresponding to the current frame includes: performing the following steps in an embedded system or radar processor: Based on the first radar attitude information and the attitude change amplitude threshold, the solution space range of the model parameters of the radar attitude model is set; Based on the radar detection information of a set number of randomly selected candidate stationary target points from the plurality of candidate stationary target points, a radar attitude model is fitted to determine the first model parameter value corresponding to the current frame.

9. The target point dynamic and static recognition method according to claim 1 or 2, The radar system is carried by a vehicle, and the first radar attitude information is determined based on the vehicle motion information collected by the sensors configured on the vehicle. or, The first radar attitude information is the radar attitude information corresponding to the latest successfully fitted frame before the current frame; or, The first radar attitude information is obtained by filtering the radar attitude information corresponding to multiple successfully fitted frames before the current frame.

10. A method for identifying dynamic and static target points, applied in a radar, the radar comprising a radio frequency transceiver and a processor, the processor being configured to process signals received by the radio frequency transceiver to obtain target detection data, the target detection data including Doppler velocities and elevation angles of multiple target points; the method comprising: Obtain the target detection data for the current frame; Based on the first radar attitude information, multiple candidate stationary target points are selected from the target detection data of the current frame, according to a set pitch angle range and a set Doppler velocity range: Using the first radar attitude information as the initial value for fitting, a preset algorithm is used to fit the radar attitude model to the multiple candidate stationary target points to generate the radar attitude information corresponding to the current frame. as well as, Based on the generated radar attitude information, all target points in the current frame are classified into dynamic and static categories, and target point cloud data containing dynamic and static labels are output.

11. A radar signal processing method, comprising: Classify the target points; as well as Different processing methods are used for post-processing of target points of different categories.

12. The radar signal processing method as described in claim 11, wherein the categories of the target points include: Moving target point and stationary target point; The post-processing of different categories of target points using different processing methods includes at least one of the following: Perform one or more of the following post-processing steps on the moving target point: de-velocity fuzzing, clustering, and tracking; For stationary target points, none of the following post-processing steps are performed: de-velocity fuzzing, clustering, and tracking.

13. The radar signal processing method as described in claim 11 or 12, The classification of target points includes: The radar attitude information corresponding to the current frame is determined based on the first radar attitude information, and the target points are classified according to the radar attitude information corresponding to the current frame. The first radar attitude information is obtained based on at least one of the following: The radar attitude information corresponding to historical frames and the radar attitude information collected in real time by the radar system's carrying equipment.

14. The radar signal processing method as described in claim 13, When obtaining the first radar attitude information based on the radar attitude information corresponding to historical frames, at least one of the following methods is used to fit the radar attitude model in order to determine the radar attitude information corresponding to the current frame: The radar attitude information is used as the initial value for fitting, and the radar attitude model is fitted. The radar attitude model is fitted based on the set velocity or azimuth solution space range; or, the radar attitude model is fitted based on the set velocity and azimuth solution space range. Select a subset of target points from the target points detected in the current frame to fit the radar attitude model; If the radar attitude model fails to fit for multiple consecutive frames, the radar attitude model is fitted based on all target points detected in the current frame.

15. The radar signal processing method as described in claim 14, The step of selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting includes: Based on the first radar attitude information, multiple candidate stationary target points are selected from the target points detected in the current frame for radar attitude model fitting. The plurality of candidate stationary target points are multiple target points among the target points detected in the current frame that satisfy one or more of the following constraints: The pitch angle of the target point is within the pitch angle threshold range; The Doppler velocity of the target point is less than the difference between a set Doppler velocity threshold and a first velocity; wherein, the first velocity is the ground motion velocity of the target point determined based on the first radar attitude information; The azimuth of the target point falls within the first azimuth range; wherein, the first azimuth range is the azimuth range of a stationary target determined based on the first radar attitude information under the set maximum unambiguous velocity condition.

16. The radar signal processing method as described in claim 14, The step of selecting a subset of target points from the target points detected in the current frame for radar attitude model fitting includes: Based on the distance information of the target points, select target points whose distance is less than the set distance threshold from the target points detected in the current frame, and perform radar attitude model fitting. or, Based on the distance information of the target points, the target points detected in the current frame are sorted, and the nearest set number of target points are selected for radar attitude model fitting.

17. The radar signal processing method as described in claim 11 or 12, wherein the target points are classified using the target point dynamic and static identification method as described in any one of claims 1-10.

18. A radar signal processing method applied to a frequency modulated continuous wave (FMCW) millimeter-wave radar, wherein the FMCW millimeter-wave radar includes a radio frequency front-end and a processor, the radio frequency front-end is configured to transmit and receive radar signals, and the processor is configured to perform signal processing on the received radar signals to obtain point cloud data. And, the method includes: The point cloud data is classified into target points to generate a set of moving target points and a set of stationary target points; Perform occupancy raster map construction on a set of stationary target points, and output location information that can be used for map construction; and Perform Doppler velocity deblurring on the set of moving target points, apply the DBSCAN clustering algorithm for density-based noisy spatial clustering, and perform Kalman filter tracking to output one or more of the following: target identifier, position, velocity, and trajectory history.

19. An integrated circuit including a processor configured to implement the target point dynamic and static identification method as claimed in any one of claims 1-10, or the radar signal processing method as claimed in any one of claims 11-18.

20. An electromagnetic wave device, comprising: Carrier; The integrated circuit as described in claim 19 is disposed on the carrier; An antenna is disposed on the carrier, or the antenna and the integrated circuit are integrated into a single device and disposed on the carrier. The integrated circuit is connected to the antenna, which is used to transmit radio frequency signals and receive radio frequency signals.

Images