KR20230160076A - Target location determine apparatus and method thereof - Google Patents

Abstract

A target positioning device is disclosed. The target positioning device includes a radar device that emits a radar signal and receives a radar signal reflected from the target, and a processor that determines the position of the target based on the received radar signal, and the processor determines the location of the target based on the received radar signal. Apply the Discrete Fourier Transform (DFT), estimate the signal-to-noise ratio (SNR) based on the radar signal to which the Discrete Fourier Transform has been applied, and if the estimated signal-to-noise ratio is less than a preset value, use the RELAX algorithm. The location of the target is determined, and if the estimated signal-to-noise ratio is greater than a preset value, the location of the target is determined using a subspace algorithm.

Description

Target location determination device and method {TARGET LOCATION DETERMINE APPARATUS AND METHOD THEREOF}

The present disclosure relates to a target position determination device and method, and to a target position determination method and method that can determine the target position with optimal performance and low complexity using a high-resolution algorithm corresponding to the target detection environment.

A radar sensor is a sensor that estimates the distance, speed, and angle of a target. Not only does it have a longer estimable distance than other sensors, but it also has characteristics that are robust to the environment.

Among various radar systems, the FMCW (Frequency Modulation Continuous Wave) radar method only needs to sample and store the difference information between the transmitted and received waveforms as a digital signal, thereby reducing complexity and reducing costs, so it is applied in many fields.

In FMCW radar systems, it is important to accurately estimate frequency parameter components. This FMCW radar system uses an FFT-based method to estimate frequency components, but the FFT-based method does not have high resolution, so there is a high probability that it will be recognized as a single target when multiple targets are adjacent.

To overcome this problem, various high-resolution algorithms are used. However, since the complexity and performance of each high-resolution algorithm are different, and in particular, there are performance differences depending on the application environment, a target positioning method with low complexity while having optimal performance in various environments was required.

Therefore, the purpose of the present disclosure is to provide a target position determination method and method that can determine the target position with low complexity while having optimal performance using a high-resolution algorithm corresponding to the target detection environment.

A target positioning device according to the present disclosure for achieving the above object is a radar device that emits a radar signal and receives a radar signal reflected from the target, and determines the position of the target based on the received radar signal. It includes a processor, wherein the processor applies a Discrete Fourier Transform (DFT) to the received radar signal, and estimates a signal-to-noise ratio (SNR) based on the radar signal to which the Discrete Fourier Transform is applied. , if the estimated signal-to-noise ratio is less than a preset value, the location of the target is determined using the RELAX algorithm, and if the estimated signal-to-noise ratio is more than the preset value, the location of the target is determined using the subspace algorithm. decide

In this case, the processor may estimate the signal-to-noise ratio based on the average power and peak value of the radar signal to which the discrete Fourier transform has been applied.

Meanwhile, if the estimated signal-to-noise ratio is less than a preset value, the processor calculates the location and size of the scattering point in the radar signal to which the discrete Fourier transform is applied, and generates a characteristic vector based on the calculated location and size. The location of the target can be determined by comparing the generated characteristic vector with previously stored vector information.

Meanwhile, the subspace algorithm may be a multiple signal classification algorithm (MUSIC) or an estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm.

Meanwhile, the target position determination method according to an embodiment of the present disclosure includes the steps of applying a Discrete Fourier Transform (DFT) to a received radar signal, and a signal-to-noise ratio based on the radar signal to which the Discrete Fourier Transform is applied ( Estimating the SNR), and if the estimated signal-to-noise ratio is less than a preset value, determining the location of the target using the RELAX algorithm, and if the estimated signal-to-noise ratio is more than the preset value, using a subspace algorithm It includes the step of determining a target position to determine the position of the target.

In this case, the step of estimating the signal to noise may estimate the signal to noise ratio based on the average power and peak value of the radar signal to which the discrete Fourier transform has been applied.

Meanwhile, in the step of determining the target position, if the estimated signal-to-noise ratio is less than a preset value, the position and size of the scattering point are calculated from the radar signal to which the discrete Fourier transform is applied, and based on the calculated position and size, A feature vector can be generated, and the location of the target can be determined by comparing the generated feature vector with previously stored vector information.

Meanwhile, the subspace algorithm may be a multiple signal classification algorithm (MUSIC) or an estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm.

As described above, according to various embodiments of the present disclosure, a target can be accurately detected using the RELAX algorithm in a low SNR environment, and a target can be detected with high performance and low complexity by using a subspace-based algorithm in a high SNR environment. You can.

1 is a diagram illustrating a specific configuration of a target positioning device according to an embodiment of the present disclosure;
Figure 2 is a diagram for explaining the application operation of a high-resolution algorithm for each SNR according to an embodiment of the present disclosure;
Figure 3 is a diagram showing an example spectrum for each SNR, and
Figure 4 is a flowchart illustrating a method for determining a target location according to an embodiment of the present disclosure.

Terms used in this specification will be briefly described, and the present disclosure will be described in detail.

The terms used in the embodiments of the present disclosure have selected general terms that are currently widely used as much as possible while considering the functions in the present disclosure, but this may vary depending on the intention or precedent of a technician working in the art, the emergence of new technology, etc. . In addition, in certain cases, there are terms arbitrarily selected by the applicant, and in this case, the meaning will be described in detail in the description part of the relevant disclosure. Therefore, the terms used in this disclosure should be defined based on the meaning of the term and the overall content of this disclosure, rather than simply the name of the term.

Embodiments of the present disclosure may be subject to various changes and may have various embodiments, and specific embodiments will be illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the scope to specific embodiments, and should be understood to include all transformations, equivalents, and substitutes included in the disclosed spirit and technical scope. In describing the embodiments, if it is determined that a detailed description of related known technology may obscure the point, the detailed description will be omitted.

Terms such as first, second, etc. may be used to describe various components, but the components should not be limited by the terms. Terms are used only to distinguish one component from another.

Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, “comprises.” Or “made up.” Terms such as are intended to designate the presence of features, numbers, steps, operations, components, parts, or a combination thereof described in the specification, but are intended to indicate the presence of one or more other features, numbers, steps, operations, components, parts, or It should be understood that the existence or addition possibility of combinations of these is not excluded in advance.

In an embodiment of the present disclosure, a 'module' or 'unit' performs at least one function or operation, and may be implemented as hardware or software, or as a combination of hardware and software. Additionally, a plurality of 'modules' or a plurality of 'units' may be integrated into at least one module and implemented with at least one processor, except for 'modules' or 'units' that need to be implemented with specific hardware.

Below, with reference to the attached drawings, embodiments of the present disclosure will be described in detail so that those skilled in the art can easily practice them. However, the present disclosure may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present disclosure in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

Hereinafter, the present disclosure will be described in more detail with reference to the drawings.

1 is a diagram illustrating a specific configuration of a target positioning device according to an embodiment of the present disclosure.

Referring to FIG. 1, the target location determination device 100 may include a radar device 110 and a processor 120. Here, the target positioning device 100 may be a radar capable of emitting radar signals, a navigation device for autonomous driving, a hazard detection device that detects hazards while driving the vehicle, or a vehicle including these.

The radar device 110 may use a phased array antenna. Here, a phased array antenna (or phased array system) is an antenna that scans a radiation beam by electronically changing the phase fed to each element of the array antenna. The phased array antenna used in the present disclosure may have a structure in which a plurality of antennas are arranged at intervals of λ/2 of the radar signal. At this time, the phased array antenna may be composed of various numbers of antennas (eg, 8, 12, 16, 20, 24, 28, 32). The radar device 110 in the present disclosure may operate in a vehicle FMCW (Frequency Modulation Continuos Wave) radar method.

Here, FMCW is a radar signal that is transmitted by modulating the frequency of a sine wave and reflected from the front. It is a system that can simultaneously measure the distance and speed of the target by using the time-frequency domain simultaneously. The FMCW radar method has very high bandwidth efficiency and low complexity, so it is widely used in automotive radar systems.

When operating in the FMCW radar method, the radar device 110 can generate and transmit a regular wave frequency-modulated with time through a waveform generator and a voltage-controlled oscillator. For example, the radar device 110 may create and transmit a transmission signal by modulating the preceding frequency by the frequency BW (Bandwidth) during △t. Here, the frequency may be 24 GHz, 77 GHz, etc., but is not limited thereto.

The propagated signal is reflected by an object in front and can be received by the radar device 110 with a time delay depending on the distance and a Doppler frequency due to the speed difference. Here, the received radar signal can be multiplied by the received signal and the transmitted signal using a mixer, and the distance can be estimated based on the frequency difference between them. The received radar signal described later may be the radar signal itself received through an antenna, or may be a signal obtained by multiplying the received signal and the transmitted signal by a mixer as described above.

Additionally, the radar device 110 may perform analog-to-digital conversion on the received radar signal and provide the digitally converted radar signal to the processor 120. That is, the radar signal described later can be expressed as a digital value.

The processor 120 controls each component within the target position detection device 100. Specifically, the processor 120 controls the radar device 110 to emit a radar signal to detect objects in front of (or around) the device in real time, and uses the radar signal received from the radar device 110 to The target location can be detected through analysis.

Additionally, the processor 120 may estimate the SNR environment in the environment where the target is located or the environment where the target positioning device is currently located. For example, the processor 120 performs a Discrete Fourier Transform (DFT) on the radar signal, calculates a peak to average power ratio (PAPR) using the discrete Fourier transformed radar signal, and uses the calculated PAPR. Thus, it is possible to estimate whether it is a low SNR or a high SNR.

Here, the discrete Fourier transform is a technique for estimating frequency components, and the resolution is not high. Accordingly, when multiple targets are adjacent, there is a high probability that they will be recognized as a single target. Accordingly, high-resolution algorithms such as the RELAX algorithm, MUSIC algorithm, and ESPRIT algorithm are used.

However, these high-resolution algorithms each have different complexity, and their performance also varies depending on the SNR environment. For example, the RELAX algorithm is a type of CLEAN technique, which is a technique for predicting the poles of a signal. It reconstructs the PSF (point spread function) in the order of the highest pole values of the signal and subtracts them from the original signal in order. This is a technique to obtain a clean signal. In this way, the RELAX algorithm can identify targets with a small number of scattering points, but its complexity is high compared to other high-resolution algorithms.

Meanwhile, the subspace-based algorithm is a method of improving angular resolution only through signal processing without increasing the number of antennas. It separates the space consisting of data measured by antennas or sensors into signal space and noise space to determine the locations of signal sources. This is a method of tracking position using the fact that the vector is orthogonal to the noise space. This subspace-based algorithm has relatively low complexity compared to the RELAX algorithm, and performs better than the RELAX algorithm in a high SNR environment.

Here, the high-resolution algorithm (or super-resolution algorithm) may be a multiple signal classification algorithm (MUSIC), and when implemented, ESPRIT (estimation of signal parameters via rotational invariance techniques) or other high-resolution frequency detection algorithms may be used. It may be possible. Hereinafter, for ease of explanation, the description will be made assuming the case of applying a multi-signal classification technique called MUSIC.

As the performance of high-resolution algorithms is different for each SNR environment, and the complexity of each high-resolution algorithm is also different, in this disclosure, the SNR environment is estimated and the target location is detected using a high-resolution algorithm suitable for the estimated SNR environment.

Meanwhile, in the above description, the SNR environment is estimated using PAPR, but the SNR environment can also be estimated using other information in addition to PAPR. Specifically, it is possible to predict the target distance and target angle using only signals obtained through discrete Fourier transform, and to estimate the SNR environment based on the predicted distance and angle. For example, if the target distance is close and the predicted target position is within 30 degrees, it is assumed to be a high SNR environment. If the target distance is greater than a preset value or the angle is predicted to be outside of 30 degrees, it is assumed to be a low SNR environment. It is also possible to do so.

In the above, it was explained that the SNR environment is estimated by considering both distance and angle, but when implementing, it is also possible to estimate the SNR environment using only one of the two degrees, and at least two of the distance, angle, and PAPR can be combined to estimate the SNR environment. It is also possible to estimate the SNR environment or use other information that can be obtained through a discrete Fourier transformed signal.

When the SNR environment is estimated in this way, the processor 120 can determine the location of the target using a high-resolution algorithm corresponding to the estimated environment. For example, if a low SNR environment is estimated, the processor 120 may determine the location of the target by applying the radar signal to the RELAX algorithm.

Specifically, the position and size of the scattering point can be calculated for each of a plurality of channels, and a characteristic vector can be generated based on the position and size of the scattering point for each channel. For example, the size of the scattering points can be arranged in order from the largest value, and the relative distances between different scattering points are calculated based on the largest scattering point, respectively, to generate a characteristic vector. Meanwhile, during implementation, detection of the target can be completed only by generating the corresponding feature vector, and it is also possible to identify which object the feature vector is for through comparison with the feature vector and the previously stored vector. Only feature vectors corresponding to the object can be used for object detection.

Meanwhile, in a high SNR environment, the processor 120 can determine the location of the target by inputting radar signals received through a plurality of channels into a subspace-based high-resolution algorithm. Hereinafter, in order to facilitate the operation of the multi-signal classification technique, the description will be made assuming that a linear array antenna consisting of M antennas with uniform spacing d is used.

Radar signals transmitted from M antennas are reflected by D different targets, and the returning plane waves are incident at different angles depending on the angles at which the targets are located.

At this time, the signal x[n] received by the array antenna can be expressed by Equation 1 below.

<Equation 1>

Here, [n] refers to the nth time sample, and a(θ _i ) is a vector representing the phase difference (dsin(θ)) caused by the incident angle of the plane wave (θ _i ) and the spacing d of the array antenna, s. _i [n] is the nth time surface of the incident signal reflected and returned to the ith target.

When a radar signal is received using the above reception signal, the angle of arrival in the time domain can be estimated using the MUSIC algorithm.

Additionally, the angle of arrival in the frequency domain can be estimated by applying the MUSIC algorithm to the frequency domain. Specifically, only the desired angle can be estimated by applying the spectrum values corresponding to the beat frequency detected in advance through the CFAR (Constant false alarm rate) algorithm.

And the final location of the target can be calculated using the angle of arrival calculated in the time domain and the angle of arrival calculated in the frequency domain.

As described above, the target positioning device according to the present disclosure predicts the SNR environment and detects the target's position using the RELAX algorithm when it is estimated to be a low SNR environment, so that the target can be detected with high precision. Additionally, if it is assumed to be a high SNR environment, the target can be detected using a subspace-based algorithm that can detect the target with low complexity and high performance in a high SNR environment. In this way, by changing the high-resolution algorithm for each SNR environment to detect the target, adaptive target detection is possible for various SNR environments.

In showing and describing FIG. 1, the target detection device is shown and described as including only two components, but may further include other components when implemented. For example, it may further include a communication device for transmitting the detected target location to another device or a display for displaying the detected target location.

In addition, in showing and describing FIG. 1, the processor 120 is shown and described as calculating the target location, but when implemented, the radar device 110 performs an operation of calculating the target location, and the processor 120 may only perform control operations for the radar device 110. Additionally, the target detection device 100 may be mounted on a vehicle and operate as an autonomous driving or danger detection device.

Meanwhile, in describing FIG. 1 , the target detection device 100 is described as detecting one target, but the target detection device 100 can determine the positions of a plurality of targets in the above-described process. In addition, in the above-described process, it was explained that the processor 120 applies the signals of the plurality of channels to the MUSIC algorithm as is, but in the implementation, preprocessing, etc. for the radar signals for the plurality of channels is performed in advance and then the MUSIC algorithm is applied. It is also possible to apply it to an algorithm.

Figure 2 is a diagram for explaining the operation of applying a high-resolution algorithm for each SNR according to an embodiment of the present disclosure.

Referring to FIG. 2, first, a digital signal for the received radar signal may be received, and Discrete Fourier Transform (DFT) may be applied to the received digitized radar signal (S210).

And, PAPR (Peak to average power ratio) can be calculated based on the radar signal to which the discrete Fourier transform has been applied (S220).

Then, estimate the signal-to-noise ratio (SNR) based on the calculated PAPR (S420. For example, by comparing the calculated PAPR with a preset value, if it is less than the preset value (e.g., 100), low SNR environment It is assumed that it is, and if it exceeds the preset value, it can be assumed that it is a high SNR environment. Here, the preset value can be a value optimized by experiment, and depending on the application environment, select one of various values that suits the current environment. A specific value may be used. As explained above, when implementing, it is also possible to use information other than the PAPR value or to estimate the SNR environment by combining the PAPR with other information.

If it is a low SNR environment (240-Low), the location of the target is determined using the RELAX algorithm (250). Meanwhile, during implementation, distance information and angle information with the target can be proactively calculated based on a radar signal to which a discrete Fourier transform has been applied, and the calculated distance and angle information can be applied to the RELAX algorithm to determine the target's location.

If it is a high SNR environment (240-High), the target location is determined using a subspace algorithm (260). Meanwhile, when implementing, before applying the subspace algorithm, the angle of the target is predicted based on the radar signal to which the discrete Fourier transform is applied, and an appropriate subspace algorithm is used according to the predicted angle of the target, or the predicted angle is used. It is also possible to detect the target with lower complexity by detecting the target using only some of the plurality of channels depending on the angle of the target.

Figure 3 is a diagram showing an example spectrum for each SNR.

Referring to FIG. 3, the first spectrum 310 is a spectrum in a low SNR environment, and the second spectrum 320 is a spectrum in a high SNR environment. In both cases, the peak values are located between 150 and 200, but the PAPR values at those peaks are different. For example, the PAPR in the first spectrum 310 is 12.25, while the PAPR in the second spectrum is 200.64.

Since the PAPR values detected in different SNR environments are different, the PAPR value can be used to distinguish whether the current target detection environment is a high SNR environment or a low SNR environment. For example, if it is an open area environment where a specific object exists nearby, it will be a high SNR environment, and the value of the radar signal reflected from the specific object, as shown in the second spectrum 320, will be the highest. In this environment, a subspace-based algorithm such as the MUSIC algorithm can be used to perform the target with relatively lower complexity than the RELAX algorithm.

Meanwhile, in an environment where several objects exist around, it will be a low SNR environment, and in addition to the value of the radar signal reflected from a specific object, such as the first spectrum 310, there are also radar signal lights caused by various objects. In this environment, targets can be detected at high resolution by using the high-performance RELAX algorithm.

4 is a flowchart illustrating a method for determining a target location according to an embodiment of the present disclosure.

Referring to FIG. 4, first, Discrete Fourier Transform (DFT) is applied to the received radar signal.

Then, the signal-to-noise ratio (SNR) is estimated based on the radar signal to which the discrete Fourier transform has been applied (S420. Specifically, the signal-to-noise ratio can be estimated based on the average power and peak value of the radar signal to which the discrete Fourier transform has been applied. .

If the estimated signal-to-noise ratio is less than the preset value (S430-Y), the location of the target is determined using the RELAX algorithm (S440). Specifically, the location and size of the scattering point are calculated from the radar signal to which the discrete Fourier transform is applied, a feature vector is generated based on the calculated position and size, and the target location is compared with the generated feature vector and previously stored vector information. can be decided.

If the estimated signal-to-noise ratio is greater than the preset value (S430-N), the target location is determined using a subspace algorithm (S450). Specifically, if it is determined that the signal-to-noise ratio environment is high, the target is identified using the Multiple signal classification algorithm (MUSIC) or ESPRIT (estimation of signal parameters via rotational invariance techniques) algorithm, which has lower complexity than the RELAX algorithm. You can decide the location.

Therefore, the target location determination method according to this embodiment predicts the SNR environment using PAPR, and when it is estimated to be a low SNR environment, the target position is detected using the RELAX algorithm, so the target can be detected with high precision. there is. Additionally, if it is assumed to be a high SNR environment, the target is detected using a subspace-based algorithm that can detect the target with low complexity and high performance in a high SNR environment. In this way, by changing the high-resolution algorithm for each SNR environment to detect the target, adaptive target detection is possible for various SNR environments. The target positioning method as shown in FIG. 4 can be executed on a target positioning device having the configuration of FIG. 1 and can also be executed on a target positioning device having another configuration.

In addition, the target position determination method as described above can be implemented as a program including an executable algorithm that can be executed on a computer, and the above-described program is stored and provided in a non-transitory computer readable medium. It can be.

A non-transitory readable medium refers to a medium that stores data semi-permanently and can be read by a device, rather than a medium that stores data for a short period of time, such as registers, caches, and memories. Specifically, programs for performing the various methods described above include magnetic media, CD (Compact Disk), DVD (Digital Video Disk), hard disk, Blu-ray disk, USB, memory card, ROM (Read Only Memory), etc. It may be stored and provided in a non-transitory readable medium.

In addition, although the preferred embodiments of the present disclosure have been shown and described above, the present disclosure is not limited to the specific embodiments described above, and the technical field to which the disclosure pertains without departing from the gist of the present disclosure as claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be understood individually from the technical ideas or perspectives of the present disclosure.

100: target positioning device 110: radar device
120: processor

Claims (8)

In the target positioning device,
A radar device that emits a radar signal and receives a radar signal reflected from the target; and
It includes a processor that determines the location of the target based on the received radar signal,
The processor,
Applying Discrete Fourier Transform (DFT) to the received radar signal,
Estimate the signal-to-noise ratio (SNR) based on the radar signal to which the discrete Fourier transform is applied,
If the estimated signal-to-noise ratio is less than a preset value, the location of the target is determined using the RELAX algorithm,
A target positioning device that determines the position of the target using a subspace algorithm if the estimated signal-to-noise ratio is greater than a preset value. According to paragraph 1,
The processor,
A target positioning device that estimates the signal-to-noise ratio based on the average power and peak value of the radar signal to which the discrete Fourier transform is applied. According to paragraph 1,
The processor,
If the estimated signal-to-noise ratio is less than the preset value, the position and size of the scattering point are calculated from the radar signal to which the discrete Fourier transform is applied, a characteristic vector is generated based on the calculated position and size, and the generated A target positioning device that determines the position of the target by comparing characteristic vectors with previously stored vector information. According to paragraph 1,
The subspace algorithm is a multiple signal classification algorithm (MUSIC) or an estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm. In the target location determination method,
Applying Discrete Fourier Transform (DFT) to the received radar signal;
estimating a signal-to-noise ratio (SNR) based on the radar signal to which the discrete Fourier transform is applied; and
If the estimated signal-to-noise ratio is less than a preset value, the location of the target is determined using the RELAX algorithm, and if the estimated signal-to-noise ratio is more than the preset value, the target is determined using a subspace algorithm. A method for determining a target position, including determining a position. According to clause 5,
The step of estimating the signal to noise is,
A target position determination method for estimating a signal-to-noise ratio based on the average power and peak value of the radar signal to which the discrete Fourier transform is applied. According to clause 5,
The step of determining the target location is,
If the estimated signal-to-noise ratio is less than the preset value, the position and size of the scattering point are calculated from the radar signal to which the discrete Fourier transform is applied, a characteristic vector is generated based on the calculated position and size, and the generated A target location determination method that determines the location of the target by comparing characteristic vectors with previously stored vector information. According to clause 5,
The subspace algorithm is a multiple signal classification algorithm (MUSIC) or an estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm.

Images