A radar ray tracing method, system and terminal based on data drive

Abstract

The present invention discloses a radar ray tracing method, system and terminal based on data-driven, the method comprising: constructing a data-driven model based on a data-driven method, the data-driven model comprising a target state classification model and a parameter regression model; inputting first scene information into the data-driven model to output a false alarm target; reversely generating multipath rays that generate false alarm targets according to the false alarm targets, and calculating parameters such as the energy value of the multipath rays; constructing a ray tracing model, inputting second scene information into the ray tracing model, and outputting a ray set composed of a plurality of original rays; adding multipath rays and parameters such as the energy value of the multipath rays to the ray set, and obtaining a ray set that returns to the radar in the scene; the above processes together constitute a radar simulation system, and the simulation system is used to perform radar ray tracing tasks. The present invention is based on data-driven inference of multipath information, and effectively improves the accuracy of radar ray tracing.

Description

Technical Field

[0001] The present invention relates to the field of radar ray tracing technology, and in particular to a data-driven radar ray tracing method, system, terminal and computer-readable storage medium. Background Art

[0002] Radar ray tracing is the process of using the propagation characteristics of electromagnetic waves to simulate the propagation path of electromagnetic waves in space to predict and calculate the position and intensity of radar echoes. This process involves the propagation, reflection, and refraction of electromagnetic waves, and usually requires the help of computer simulation and algorithms to achieve this.

[0003] Radar ray tracing involves the multipath propagation of radar signals. The multipath propagation of radar signals refers to the phenomenon that during the propagation of signals, due to the influence of obstacles, multiple reflections or scattering occur, causing the signals to propagate along multiple paths to the receiving end. The lengths and delays of these paths are different, so that the time for each component field to arrive at the receiving end is also different, resulting in interference, causing signal distortion or errors.

[0004] However, the signal of multipath propagation is weak and can be easily ignored by ray tracing. Ray tracing has a long tail effect. If you want to accurately calculate the process, you need an extremely sophisticated calculation process, which brings an unacceptable amount of calculation. When existing methods simplify the ray propagation process to improve calculation efficiency, they often miss the critical information of ray multipath propagation, resulting in insufficient radar ray tracing accuracy.

[0005] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0006] The main purpose of the present invention is to provide a data-driven radar ray tracing method, system, terminal and computer-readable storage medium, aiming to solve the problem in the prior art that when simplifying the ray propagation process to improve calculation efficiency, critical ray multipath propagation information is often omitted, resulting in insufficient radar ray tracing accuracy.

[0007] To achieve the above object, the present invention provides a radar ray tracing method based on data driving, and the radar ray tracing method based on data driving comprises the following steps:

[0008] Build data-driven models based on data-driven methods;

[0009] Acquire first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target;

[0010] Reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate parameters of the multipath rays;

[0011] Constructing a ray tracing model, and acquiring second scene information, inputting the second scene information into the ray tracing model for calculation, and outputting a ray set consisting of a plurality of original rays;

[0012] Adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set;

[0013] Radar simulation modeling is performed according to the target ray set to obtain a simulation system, and the radar ray tracing task is performed using the simulation system.

[0014] Optionally, in the data-driven radar ray tracing method, the data-driven model includes a target state classification model and a parameter regression model.

[0015] Optionally, in the data-driven radar ray tracing method, the data-driven model is used to predict false alarm targets, and the target state classification model and the parameter regression model are both constructed based on artificial neural networks.

[0016] Optionally, the data-driven radar ray tracing method, wherein the acquiring of first scene information, inputting the first scene information into the data-driven model for identification and prediction, and outputting a false alarm target, specifically includes:

[0017] Acquire first scene information, wherein the first scene information includes a speed of a scene target, an angle of the scene target, a distance of the scene target, and a heading angle of the scene target;

[0018] Inputting the first scene information into the target state classification model for identification, and outputting the identified false alarm target;

[0019] The first scene information is input into the parameter regression model for prediction, and the variable value of the false alarm target is output.

[0020] Optionally, in the data-driven radar ray tracing method, the parameters of the multipath ray include: total distance, receiving direction, speed and energy value of the target;

[0021] The reversely generating multipath rays that generate the false alarm target according to the false alarm target, and calculating the parameters of the multipath rays specifically includes:

[0022] Obtaining variable values ​​of the false alarm target, the variable values ​​including: the distance of the false alarm target, the angle of the false alarm target, the speed of the false alarm target and the target scattering interface of the false alarm target;

[0023] The total distance, receiving direction and target speed of the multipath rays generating the false alarm target are calculated according to the distance, the angle and the speed, and the energy value of the multipath rays generating the false alarm target is calculated according to the target scattering interface of the false alarm target.

[0024] Optionally, the data-driven radar ray tracing method, wherein the step of calculating the energy value of the multipath ray that generates the false alarm target according to the target scattering interface of the false alarm target, specifically includes:

[0025] The energy value of the multipath ray is calculated according to the target scattering interface, ray parameters and radar parameters of the false alarm target:

[0026] ;

[0027] in, represents the energy value of the multipath ray, represents the target scattering interface, represents the emission power of the ray, represents the gain of the radar receiving antenna, represents the gain of the radar transmitting antenna, represents the radar's operating wavelength, Indicates the distance between the radar and the false alarm target.

[0028] Optionally, in the data-driven radar ray tracing method, the radar ray tracing tasks include autonomous driving tasks and drone tasks.

[0029] In addition, to achieve the above-mentioned purpose, the present invention further provides a data-driven radar ray tracing system, wherein the data-driven radar ray tracing system comprises:

[0030] A driving model building module is used to build a data-driven model based on a data-driven approach;

[0031] A false alarm target prediction module, used to obtain first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target;

[0032] A multipath information inversion module is used to reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate the parameters of the multipath rays;

[0033] An original ray calculation module, used to construct a ray tracing model, obtain second scene information, input the second scene information into the ray tracing model for calculation, and output a ray set consisting of a plurality of original rays;

[0034] A multipath information fusion module, used for adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set;

[0035] The simulation application module is used to perform radar simulation modeling according to the target ray set to obtain a simulation system, and use the simulation system to perform radar ray tracing tasks.

[0036] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a data-driven radar ray tracing program stored in the memory and executable on the processor, wherein the data-driven radar ray tracing program, when executed by the processor, implements the steps of the data-driven radar ray tracing method as described above.

[0037] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a data-driven radar ray tracing program, and when the data-driven radar ray tracing program is executed by a processor, the steps of the data-driven radar ray tracing method as described above are implemented.

[0038] In the present invention, a data-driven model is constructed based on a data-driven method, and the data-driven model includes a target state classification model and a parameter regression model; first scene information is obtained, and the first scene information is input into the data-driven model for identification and prediction, and a false alarm target is output;

[0039] According to the false alarm target, the multipath ray corresponding to the false alarm target is reversely generated, and the energy value of the multipath ray is calculated according to the target scattering interface of the false alarm target; a ray tracing model is constructed, and the second scene information is obtained, and the second scene information is input into the ray tracing model for calculation, and a ray set composed of a plurality of original rays and an energy value set composed of energy values ​​corresponding to the plurality of original rays are output; the multipath ray is added to the ray set to obtain the target ray set, and the energy value of the multipath ray is added to the energy value set to obtain the target energy value set; radar simulation modeling is performed according to the target ray set and the target energy value set to obtain a simulation system, and the radar ray tracing task is performed using the simulation system. The present invention is based on data-driven inference of multipath information, and effectively improves the radar ray tracing accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a flow chart of a preferred embodiment of the data-driven radar ray tracing method of the present invention;

[0041] Figure 2 It is an architecture diagram of a data-driven model in a data-driven radar ray tracing method of the present invention;

[0042] Figure 3 is a structural diagram of a preferred embodiment of a data-driven radar ray tracing system of the present invention;

[0043] Figure 4 It is a structural diagram of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0044] The present application provides a data-driven radar ray tracing method, system and terminal. To make the purpose, technical solution and effect of the present application clearer and more specific, the present application is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0045] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as those generally understood by those skilled in the art to which this application belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with those in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as here.

[0046] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0047] The radar ray tracing method based on data driving described in the preferred embodiment of the present invention is as follows: Figure 1 As shown, the data-driven radar ray tracing method includes the following steps:

[0048] Step S10: construct a data-driven model based on a data-driven approach.

[0049] ‌Data-driven is a way to implement artificial intelligence. ‌Data-driven artificial intelligence (Data-Driven AI) refers to training and optimizing machine learning models through large amounts of data to enable them to have better performance and accuracy. The core idea of ​​this method is to drive the training and optimization of machine learning models through data, so that they can make more accurate predictions and decisions.

[0050] It is understandable that the data-driven model is used to predict false alarm targets, such as Figure 2 As shown, the data-driven model includes a target state classification model (false alarm model) and a parameter regression model (false alarm target model), wherein the target state classification model and parameter regression model in the data-driven model are both constructed based on an artificial neural network (ANN). An artificial neural network is a mathematical model or computational model that simulates the structure and function of a biological neural network. It is based on the basic principles of neural networks in biology, and by understanding and abstracting the structure of the human brain and the external stimulus response mechanism, a mathematical model capable of processing complex information is constructed. It should be noted that the data-driven model can not only be constructed using an artificial neural network ANN as a framework, but can also be established through other similar algorithms, which is not limited here.

[0051] Specifically, the target state classification model includes an output layer, a hidden layer and an output layer. The target state classification model is used to identify scene information (the speed of the scene target, the angle of the scene target, the distance of the scene target and the heading angle of the scene target), that is, to predict whether a false alarm target (ghost object, that is, a false target that does not exist in the scene but is identified by the radar due to various interferences) detected by the radar exists according to the scene information. If the false alarm target exists, the parameter regression model is called to calculate the parameters of the false alarm target.

[0052] It is understandable that before using the target state classification model, it is necessary to collect a data set, and use the data set to train, test and verify the target state classification model so that the target state classification model can output the classification result (false alarm target) that you want to predict. Specifically, obtain historical data, which includes historical scene information and the classification results corresponding to the historical scene information, and use the historical scene information as a feature (feature, which can also be understood as an independent variable, which is an independent variable used to train the model, usually a matrix, each row represents a sample, and each column represents a feature), and represent the data with the classification result of a false alarm target in the grayscale image with a pixel value of 255, and mark other classification results as background areas in the grayscale image with 0 as a label (label, which can be understood as a target variable, which is a dependent variable that the model needs to predict, usually a vector. For classification problems, it represents a category label; for regression problems, it represents a continuous target value).

[0053] Furthermore, a data set is constructed based on the features and the labels, and the data set is divided into a training set, a test set, and a validation set in a preset ratio (for example, the data ratio of the training set, the test set, and the validation set is 3:1:1), the training set is used to train the target state classification model, the test set is used to evaluate the target state classification model in each round of training, and the validation set is used to evaluate the trained target state classification model.

[0054] Furthermore, if Figure 2 As shown, the parameter regression model includes multiple sub-models: ANN-SGRCSRM (target scattering interface sub-model), ANN-SGVxRM (lateral velocity sub-model), ANN-SGVyRM (longitudinal velocity sub-model), ANN-SGRxRM (lateral distance sub-model), ANN-SGRyRM (longitudinal distance sub-model), these sub-models are used to calculate the variable values ​​of the false alarm target: RCS (Radar Cross Section, target scattering interface. A physical quantity that measures the ability of a target to reflect signals on a radar, usually used to describe the intensity of a target's reflection of radar waves. The larger the RCS value, the stronger the target's reflection ability on the radar, and the easier it is to be detected), the distance of the false alarm target, the angle of the false alarm target and the speed of the false alarm target.

[0055] It is understandable that before using the parameter regression model, it is also necessary to collect a data set and use the data set to train, test and verify the parameter regression model. The process is similar to the training process of the target state classification model and will not be repeated here.

[0056] Step S20: Acquire first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target.

[0057] Specifically, first scene information is obtained, wherein the first scene information includes the speed of the scene target, the angle of the scene target, the distance of the scene target and the heading angle of the scene target; the first scene information is input into the target state classification model for identification, and the identified false alarm target is output; the first scene information is input into the parameter regression model for prediction, and the variable value of the false alarm target is output.

[0058] It is understandable that the model comprises two parts, a target state classification model and a parameter regression model. The target state classification model is used to predict whether the false alarm target exists. The target state classification model is used to predict the specific numerical values ​​of parameters such as the speed and distance of the false alarm target. In the present embodiment, a classification model is set up to predict whether the false alarm target detected by the radar according to the scene information exists. When the false alarm target exists, the variable value of the false alarm target predicted by the radar at this time is predicted using the scene information, including such as distance, angle, speed and RCS etc., to realize the accurate prediction of the false alarm target.

[0059] Step S30: reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate parameters of the multipath rays.

[0060] In this embodiment, the parameters of the multipath rays include: total distance, receiving direction, speed and energy value of the target; specifically, the variable value of the false alarm target is obtained, and the variable value includes: the distance of the false alarm target, the angle of the false alarm target, the speed of the false alarm target and the target scattering interface of the false alarm target.

[0061] The total distance, receiving direction and target speed of the multipath rays generating the false alarm target are calculated according to the distance, the angle and the speed, and the energy value of the multipath rays generating the false alarm target is calculated according to the target scattering interface of the false alarm target.

[0062] Further, calculating the energy value of the multipath ray that generates the false alarm target according to the target scattering interface of the false alarm target specifically includes:

[0063] The energy value of the multipath ray is calculated according to the target scattering interface, ray parameters and radar parameters of the false alarm target:

[0064] ;

[0065] in, represents the energy value of the multipath ray, represents the target scattering interface, represents the emission power of the ray, represents the gain of the radar receiving antenna, represents the gain of the radar transmitting antenna, represents the radar's operating wavelength, Indicates the distance between the radar and the false alarm target.

[0066] In the present invention, it is considered that multipath propagation of radar signals may cause false targets to appear in radar detection results. Since it is relatively easy to predict radar output data of target level according to scene information, the present invention first establishes a target level model based on a data-driven method to predict false alarm targets representing multipath information. Then, according to the association between the false alarm target and the multipath information, the multipath information is inferred from the false alarm target, thereby enhancing the prediction ability of the ray tracing model for multipath and improving the accuracy of the model.

[0067] Step S40: construct a ray tracing model, obtain second scene information, input the second scene information into the ray tracing model for calculation, and output a ray set consisting of a plurality of original rays.

[0068] In this embodiment, a ray tracing model is constructed. The ray tracing model is a radar ray tracing model, which is a technology used to predict and simulate the propagation characteristics of radar signals. It calculates parameters such as the strength, phase, and delay of the received signal by tracking the path and reflection of the radar signal during the propagation process. This model is of great significance in radar system design and performance evaluation.

[0069] It is understandable that in order to achieve the same prediction efficiency, the conventional ray tracing model must increase the density of rays to improve the calculation accuracy. However, due to the long-tail effect of efficiency and accuracy in mechanism calculation, this method will bring a great computational burden to the ray tracing model. Therefore, the present invention outputs a ray set composed of multiple original rays through a conventional ray tracing model based on the second scene information (the second scene information is a scene model more detailed than the first scene information). The ray set includes parameters corresponding to the multiple original rays (total distance, receiving direction, speed and energy value of the target). The calculation amount of the ray tracing model is not increased. Instead, a data-driven model based on data-driven technology is constructed. Using the data-driven model, the data-driven model can quickly and accurately predict key multipath information, thereby improving the calculation accuracy of multiple key features, and hardly reduces the calculation efficiency of the model.

[0070] Step S50: adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set.

[0071] In this embodiment, the multipath rays corresponding to the false alarm target generated inversely from the false alarm target and the parameters of the multipath rays are added to the ray set as additional outputs of the ray tracing model to obtain a target ray set. Thus, the present invention improves the calculation efficiency while adding the key ray multipath propagation information that is easily missed to the final output, thereby greatly reducing the measurement error.

[0072] Step S60: Perform radar simulation modeling according to the target ray set to obtain a simulation system, and use the simulation system to perform a radar ray tracing task.

[0073] It can be understood that the target ray set and the target energy value set are used in radar simulation modeling (in tasks using ray tracing, such as radar, lidar and camera modeling). Through radar simulation modeling, a variety of radar simulation systems can be designed, and the simulation systems can be used to perform radar ray tracing tasks; wherein, the radar ray tracing tasks include autonomous driving tasks and drone tasks.

[0074] It can be seen that in order to achieve the same prediction efficiency, conventional methods must increase the density of rays to improve the calculation accuracy. However, due to the long-tail effect of efficiency and accuracy in mechanism calculation, this method will bring a great computational burden. Taking this defect into consideration, the present invention uses the easy-to-predict characteristics of target-level results to reversely infer the original ray information from the target-level results. This effectively improves the model's prediction ability for key multipath information without increasing the amount of calculation, and almost does not reduce the model's calculation efficiency while improving the calculation accuracy of multiple key features, effectively improving the ray tracing model's prediction ability for key multipath information.

[0075] Furthermore, the radar data set established by the present invention is used for training and testing. Multiple sub-models of the target state classification model and the parameter regression model are established respectively: ANN-SGRCSRM, ANN-SGVxRM, ANN-SGVyRM, ANN-SGRxRM, ANN-SGRyRM. The results show that the model established by the present invention can predict the false alarm state of the target with an accuracy of 97.53%. The multiple sub-models of the parameter regression model achieve high-precision fitting of the false alarm target variables measured by the radar. Moreover, the present invention also counts the operation time of each sub-model in the radar model in which the target state is decoupled from the variable. The computer configuration for running the model is GTX2080, 16G memory, i7-13700, and the model is built based on pytorch (an open source deep learning framework for machine learning and deep learning). Among them, the radar model operation time is about 0.2 milliseconds, and the total serial operation time of all radar variable models is about 1.5 milliseconds; the parallel operation time is less, and the computational burden is hardly increased.

[0076] Furthermore, in the present invention, the output data of the signal-level radar model after multipath ray enhancement based on the target-level radar model is compared with the unenhanced model output data and test data. The result obtained is: after adding the echo signal of the false alarm target in the echo map using this method, an obvious multipath echo signal appears in the echo map, and the MSE of the model is improved by 31.36%. This result shows that the multipath ray enhancement method of the present invention can ensure that the radar model can accurately predict the key features in the real radar echo map.

[0077] It can be seen that the radar ray tracing method based on data-driven provided by the present invention builds a data-driven model based on the data-driven method, and the data-driven model includes a target state classification model and a parameter regression model; the first scene information is input into the data-driven model to output a false alarm target; the multipath rays of the false alarm target are reversely generated according to the false alarm target, and the energy value of the multipath rays is calculated; a ray tracing model is constructed, and the second scene information ray tracing model is used to output a ray set and a ray energy value set composed of multiple original rays; the energy values ​​of the multipath rays and the multipath rays are added to the ray set and the energy value set to obtain a target energy value set; radar simulation modeling is performed according to the target ray set and the target energy value set to obtain a simulation system, and the simulation system is used to perform the radar ray tracing task. The present invention is based on data-driven inference of multipath information, and effectively improves the accuracy of radar ray tracing.

[0078] Furthermore, if Figure 3 As shown, based on the above data-driven radar ray tracing method, the present invention also provides a data-driven radar ray tracing system, wherein the data-driven radar ray tracing system includes:

[0079] A driving model building module 51, used to build a data-driven model based on a data-driven method;

[0080] A false alarm target prediction module 52 is used to obtain first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target;

[0081] A multipath information inversion module 53 is used to reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate the parameters of the multipath rays;

[0082] The original ray calculation module 54 is used to construct a ray tracing model, obtain second scene information, input the second scene information into the ray tracing model for calculation, and output a ray set consisting of a plurality of original rays;

[0083] A multipath information fusion module 55, used for adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set;

[0084] The simulation application module 56 is used to perform radar simulation modeling according to the target ray set to obtain a simulation system, and use the simulation system to perform the radar ray tracing task.

[0085] Furthermore, if Figure 4 As shown, based on the above data-driven radar ray tracing method and system, the present invention also provides a terminal accordingly, and the terminal includes a processor 10, a memory 20 and a display 30. Figure 4 Only some components of the terminal are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0086] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal. In other embodiments, the memory 20 may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a data-driven radar ray tracing program 40 is stored on the memory 20, and the data-driven radar ray tracing program 40 can be executed by the processor 10, thereby realizing the data-driven radar ray tracing method in the present application.

[0087] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 20, such as executing the data-driven radar ray tracing method.

[0088] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface. The components of the terminal communicate with each other via a system bus.

[0089] In one embodiment, when the processor 10 executes the data-driven radar ray tracing program 40 in the memory 20, the following steps are implemented:

[0090] Build data-driven models based on data-driven methods;

[0091] Acquire first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target;

[0092] Reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate parameters of the multipath rays;

[0093] Constructing a ray tracing model, and acquiring second scene information, inputting the second scene information into the ray tracing model for calculation, and outputting a ray set consisting of a plurality of original rays;

[0094] Adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set;

[0095] Radar simulation modeling is performed according to the target ray set to obtain a simulation system, and the radar ray tracing task is performed using the simulation system.

[0096] Wherein, the data driven model includes a target state classification model and a parameter regression model.

[0097] The data-driven model is used to predict false alarm targets, and the target state classification model and the parameter regression model are both constructed based on artificial neural networks.

[0098] The acquiring of the first scene information, inputting the first scene information into the data-driven model for identification and prediction, and outputting a false alarm target specifically includes:

[0099] Acquire first scene information, wherein the first scene information includes a speed of a scene target, an angle of the scene target, a distance of the scene target, and a heading angle of the scene target;

[0100] Inputting the first scene information into the target state classification model for identification, and outputting the identified false alarm target;

[0101] The first scene information is input into the parameter regression model for prediction, and the variable value of the false alarm target is output.

[0102] The parameters of the multipath ray include: total distance, receiving direction, speed and energy value of the target;

[0103] The reversely generating multipath rays that generate the false alarm target according to the false alarm target, and calculating the parameters of the multipath rays specifically includes:

[0104] Obtaining variable values ​​of the false alarm target, the variable values ​​including: the distance of the false alarm target, the angle of the false alarm target, the speed of the false alarm target and the target scattering interface of the false alarm target;

[0105] The total distance, receiving direction and target speed of the multipath rays generating the false alarm target are calculated according to the distance, the angle and the speed, and the energy value of the multipath rays generating the false alarm target is calculated according to the target scattering interface of the false alarm target.

[0106] The step of calculating the energy value of the multipath ray that generates the false alarm target according to the target scattering interface of the false alarm target specifically includes:

[0107] The energy value of the multipath ray is calculated according to the target scattering interface, ray parameters and radar parameters of the false alarm target:

[0108] ;

[0109] in, represents the energy value of the multipath ray, represents the target scattering interface, represents the emission power of the ray, represents the gain of the radar receiving antenna, represents the gain of the radar transmitting antenna, represents the radar's operating wavelength, Indicates the distance between the radar and the false alarm target.

[0110] The radar ray tracing tasks include autonomous driving tasks and drone tasks.

[0111] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a data-driven radar ray tracing program, and when the data-driven radar ray tracing program is executed by a processor, the steps of the data-driven radar ray tracing method as described above are implemented.

[0112] In summary, the present invention proposes a radar ray tracing method, system and terminal based on data-driven, the method comprising: constructing a data-driven model based on a data-driven method; obtaining first scene information, inputting the first scene information into the data-driven model for identification and prediction, and outputting a false alarm target; reversely generating multipath rays that produce the false alarm target according to the false alarm target, and calculating the parameters of the multipath rays; constructing a ray tracing model, and obtaining second scene information, inputting the second scene information into the ray tracing model for calculation, and outputting a ray set composed of multiple original rays; adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set; performing radar simulation modeling according to the target ray set to obtain a simulation system, and using the simulation system to perform radar ray tracing tasks. The present invention utilizes the characteristic that the target-level results are easy to predict, and reversely infers the original ray information from the target-level results, effectively improving the model's prediction ability for key multipath information without increasing the amount of calculation, and effectively improving the radar ray tracing accuracy.

[0113] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or terminal including the element.

[0114] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0115] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A data-driven radar ray tracing method, characterized in that: The data-driven radar ray tracing method includes: Build data-driven models based on data-driven methods; Acquire first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target; Reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate parameters of the multipath rays; Constructing a ray tracing model, and acquiring second scene information, inputting the second scene information into the ray tracing model for calculation, and outputting a ray set consisting of a plurality of original rays; Adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set; Radar simulation modeling is performed according to the target ray set to obtain a simulation system, and the radar ray tracing task is performed using the simulation system.

2. The data-driven radar ray tracing method according to claim 1, characterized in that: The data driven model includes a target state classification model and a parameter regression model.

3. The data-driven radar ray tracing method according to claim 2, characterized in that: The data-driven model is used to predict false alarm targets, and the target state classification model and the parameter regression model are both constructed based on artificial neural networks.

4. The data-driven radar ray tracing method according to claim 3, characterized in that: The obtaining of first scene information, inputting the first scene information into the data-driven model for identification and prediction, and outputting a false alarm target specifically includes: Acquire first scene information, wherein the first scene information includes a speed of a scene target, an angle of the scene target, a distance of the scene target, and a heading angle of the scene target; Inputting the first scene information into the target state classification model for identification, and outputting the identified false alarm target; The first scene information is input into the parameter regression model for prediction, and the variable value of the false alarm target is output.

5. The data-driven radar ray tracing method according to claim 4, characterized in that: The parameters of the multipath ray include: total distance, receiving direction, speed and energy value of the target; The reversely generating multipath rays that generate the false alarm target according to the false alarm target, and calculating the parameters of the multipath rays specifically includes: Obtaining variable values ​​of the false alarm target, the variable values ​​including: the distance of the false alarm target, the angle of the false alarm target, the speed of the false alarm target and the target scattering interface of the false alarm target; The total distance, receiving direction and target speed of the multipath rays generating the false alarm target are calculated according to the distance, the angle and the speed, and the energy value of the multipath rays generating the false alarm target is calculated according to the target scattering interface of the false alarm target.

6. The data-driven radar ray tracing method according to claim 5, characterized in that: The step of calculating the energy value of the multipath ray that generates the false alarm target according to the target scattering interface of the false alarm target specifically includes: The energy value of the multipath ray is calculated according to the target scattering interface, ray parameters and radar parameters of the false alarm target: ； in, represents the energy value of the multipath ray, represents the target scattering interface, represents the emission power of the ray, represents the gain of the radar receiving antenna, represents the gain of the radar transmitting antenna, represents the radar's operating wavelength, Indicates the distance between the radar and the false alarm target.

7. The data-driven radar ray tracing method according to claim 1, characterized in that: The radar ray tracing tasks include autonomous driving tasks and drone tasks.

8. A data-driven radar ray tracing system, characterized in that: The data-driven radar ray tracing system includes: A driving model building module is used to build a data-driven model based on a data-driven approach; A false alarm target prediction module, used to obtain first scene information, input the first scene information into the data-driven model for identification and prediction, and output a false alarm target; A multipath information inversion module is used to reversely generate multipath rays that produce the false alarm target according to the false alarm target, and calculate the parameters of the multipath rays; An original ray calculation module, used to construct a ray tracing model, obtain second scene information, input the second scene information into the ray tracing model for calculation, and output a ray set consisting of a plurality of original rays; A multipath information fusion module, used for adding the multipath rays and the parameters of the multipath rays to the ray set to obtain a target ray set; The simulation application module is used to perform radar simulation modeling according to the target ray set to obtain a simulation system, and use the simulation system to perform radar ray tracing tasks.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a data-driven radar ray tracing program stored in the memory and executable on the processor. When the data-driven radar ray tracing program is executed by the processor, the steps of the data-driven radar ray tracing method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a data-driven radar ray tracing program, and when the data-driven radar ray tracing program is executed by a processor, the steps of the data-driven radar ray tracing method according to any one of claims 1 to 7 are implemented.

Images