CN110361723B - Time-frequency feature extraction method for Doppler radar moving target - Google Patents
Time-frequency feature extraction method for Doppler radar moving target Download PDFInfo
- Publication number
- CN110361723B CN110361723B CN201910669096.3A CN201910669096A CN110361723B CN 110361723 B CN110361723 B CN 110361723B CN 201910669096 A CN201910669096 A CN 201910669096A CN 110361723 B CN110361723 B CN 110361723B
- Authority
- CN
- China
- Prior art keywords
- signal
- cosine wave
- point
- wave
- signals
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/02—Systems using reflection of radio waves, e.g. primary radar systems; Analogous systems
- G01S13/06—Systems determining position data of a target
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/415—Identification of targets based on measurements of movement associated with the target
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a time-frequency characteristic extraction method of a Doppler radar moving target, which comprises the steps of continuously collecting signals of N discrete points of Doppler radar echo signals of the moving target at a preset sampling rate; extracting all single-period cosine wave-like signals from the acquired discrete signals; and calculating the frequency and power of each extracted cosine wave signal, and taking the frequency and power of the peak position of each cosine wave signal as a feature vector of the time-frequency domain of the moving target. By adopting the scheme, the motion characteristic of the moving target can be rapidly obtained, the steps are simple, complex Fourier transform is not required, and the calculation amount is small.
Description
Technical Field
The invention relates to the technical field of radar, in particular to a time-frequency feature extraction method for a moving target of a Doppler radar.
Background
Currently, a doppler radar determines a doppler frequency fd through a frequency difference between a reflected echo of a moving target and a local oscillator, and the doppler frequency fd has a relationship with a target moving speed v and a carrier wavelength λ: fd 2v/λ, so the doppler frequency can reflect the motion characteristics of the target. By performing fast fourier FFT on a finite doppler signal sequence, the spectrum of the echo can be obtained, including frequency components and power of the signal, but the FFT cannot provide a joint distribution of the signal in the time domain and the frequency domain, that is, the local motion characteristics of the target cannot be known.
When the local motion characteristics of the target are obtained, the change situation of the speed and the power of the moving target needs to be obtained. The variation of frequency and power along with time can be described by a short-time Fourier transform (STFT), a Wavelet Transform (WT) and other time-frequency analysis methods, but certain signal detection methods such as Constant False Alarm Rate (CFAR) are also required to be applied to a time-frequency domain to eliminate noise and obtain effective motion information. Comprehensively, the complete signal processing chain for extracting the moving target features has the defects of high redundancy, large operation amount and complex steps, and thus, higher hardware requirements are provided for a real-time system.
Disclosure of Invention
The embodiment of the application provides a method for extracting time-frequency characteristics of a moving target of a Doppler radar, and solves the problems that in the prior art, the extraction computation amount of the time-frequency characteristics in Doppler radar signals is large, and the steps are complex.
The embodiment of the application provides a time-frequency feature extraction method of a Doppler radar moving target, which comprises the following steps:
continuously acquiring signals of N discrete points of a Doppler radar echo signal of a moving target at a preset sampling rate;
extracting all single-period cosine wave-like signals from the acquired discrete signals;
and calculating the frequency and power of each extracted cosine wave signal, and taking the frequency and power of the peak position of each cosine wave signal as a feature vector of the time-frequency domain of the moving target.
Optionally, the step of extracting all the single-period cosine wave-like signals from the acquired discrete signals includes:
sequentially extracting quasi-cosine wave signals of each single period according to the sequence of discrete signal acquisition;
judging whether each cosine wave-like signal is a noise signal;
and extracting the cosine wave-like signals which do not belong to the noise signals, and removing the cosine wave-like signals which belong to the noise signals.
Optionally, the step of sequentially extracting the cosine wave-like signals of each single period according to the sequence of discrete signal acquisition includes:
sequentially extracting a first valley point of the current cosine wave-like signal from the first discrete point signal, wherein the amplitude of the first valley point is smaller than the amplitude of the previous discrete point signal and the amplitude of the next discrete point signal;
sequentially extracting a peak point and a second valley point from the first valley point of the current cosine wave-like signal, wherein the amplitude of the peak point is greater than the amplitude of the signal of the previous discrete point and the amplitude of the signal of the next discrete point, and the amplitude of the second valley point is less than the amplitude of the signal of the previous discrete point and the amplitude of the signal of the next discrete point;
the second valley point of the current cosine wave-like signal is taken as the first valley point of the next cosine wave-like signal.
Optionally, the step of determining whether each cosine wave-like signal is a noise signal includes:
judging whether the amplitude of the peak point of each cosine wave-like signal is smaller than a first preset value or not;
if yes, judging that the cosine wave-like signal is a noise signal;
if not, judging that the cosine wave-like signal is not a noise signal.
Optionally, the step of determining whether each cosine wave-like signal is a noise signal further includes:
judging whether the first valley point and the second valley point of each cosine wave signal are approximately symmetrical relative to the peak point; if yes, judging that the cosine wave-like signal is not a noise signal; if not, judging that the cosine wave-like signal is a noise signal.
Optionally, the step of determining whether the first valley point and the second valley point of each cosine wave signal are substantially symmetrical with respect to the peak point includes:
acquiring the time interval from the first valley point to the peak point and the time interval from the peak point to the second valley point of each cosine wave-like signal;
judging whether the difference value of the two time intervals is smaller than a second preset value, if so, approximately symmetrical a first valley point and a second valley point of the cosine wave-like signal relative to a peak point; if not, the first wave valley point and the second wave valley point of the cosine wave-like signal are asymmetric relative to the wave peak point.
Optionally, the step of determining whether the first valley point and the second valley point of each cosine wave signal are substantially symmetrical with respect to the peak point further includes:
acquiring a first amplitude difference value between a first valley point and a second valley point of each cosine wave-like signal, an average amplitude value between the first valley point and the second valley point, and a second amplitude difference value between the amplitude value of the peak point and the average amplitude value, and calculating a ratio of the first amplitude difference value to the second amplitude difference value;
judging whether the ratio is smaller than a third preset value or not, if so, approximately symmetrical a first valley point and a second valley point of the cosine wave-like signal relative to a peak point; if not, the first wave valley point and the second wave valley point of the cosine wave-like signal are asymmetric relative to the wave peak point.
Optionally, the step of determining whether each cosine wave-like signal is a noise signal further includes:
judging whether the cosine wave signals are correlated with standard cosine wave signals or not; if yes, judging that the cosine wave-like signal is not a noise signal; if not, judging that the cosine wave-like signal is a noise signal.
Optionally, the step of determining whether there is a correlation between each cosine wave signal and a standard cosine wave signal includes:
constructing a standard cosine wave signal sequence with the same number of discrete points and the same period as the discrete points of each type of cosine wave signal;
calculating a correlation coefficient between each discrete point amplitude of the cosine wave-like signal and each discrete point amplitude of the corresponding standard cosine wave signal sequence;
and judging whether the correlation coefficient is larger than a fourth preset value, if so, determining that the quasi-cosine wave signal is correlated with the standard cosine wave signal, and if not, determining that the quasi-cosine wave signal is not correlated with the standard cosine wave signal.
Optionally, the step of calculating the frequency and power of each extracted cosine wave signal comprises:
acquiring the number of discrete points contained in each kind of cosine wave signals, and calculating the frequency of each kind of cosine wave signals according to the following formula:
obtaining the average value of the amplitudes of all the discrete points and the amplitudes of the discrete points contained in each kind of cosine wave signals, and calculating the power of each kind of cosine wave signals according to the following formula:
One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:
by adopting the scheme, all discrete points which can form the characteristics of the single-period cosine waves in the discrete signals sampled every time are extracted, the frequency and the power of the discrete points at the positions of the wave crests of each type of cosine wave signals are calculated to be used as a characteristic vector of the moving target at the moment corresponding to the discrete points at the positions of the wave crests in the time-frequency domain, the characteristic vector represents one action of the moving target, all the single-period cosine waves are extracted, all the characteristic vectors are obtained, and the motion condition of the moving target in the time duration corresponding to the N sampled discrete points can be obtained equivalently. And according to the method, continuing the next sampling until all Doppler echo signals of the moving target are completely analyzed, and then depicting all motion characteristics of the moving target. By the time-frequency feature extraction of the method, the motion characteristics of the moving target can be rapidly obtained, the steps are simple, complex Fourier transform is not needed, and the computation amount is small.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the structures shown in the drawings without creative efforts.
FIG. 1 is a schematic diagram illustrating steps of a method for extracting time-frequency characteristics of a moving object of a Doppler radar according to an embodiment of the present invention;
FIG. 2 is a schematic diagram illustrating the steps of a time-frequency feature extraction method for a moving object of a Doppler radar according to another embodiment of the present invention;
fig. 3 is a schematic diagram illustrating a specific step of step S210 in fig. 2.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that, if all the directional indicators (such as upper, lower, left, right, front and rear … …) in the embodiment of the present invention are used only for explaining the relative position relationship between the components, the motion situation, etc. in a specific posture (as shown in the drawing), if the specific posture is changed, the directional indicator is changed accordingly.
In addition, if the description of "first", "second", etc. is referred to in this disclosure for descriptive purposes only, it is not to be construed as indicating or implying any relative importance or implicit indication of the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In addition, technical solutions between various embodiments may be combined with each other, but must be realized by a person skilled in the art, and when the technical solutions are contradictory or cannot be realized, such a combination should not be considered to exist, and is not within the protection scope of the present invention.
One embodiment of the invention
Referring to fig. 1, an embodiment of the present invention provides a method for extracting time-frequency features of a moving target of a doppler radar, where the method for extracting time-frequency features of a moving target of a doppler radar includes the following steps:
step S100: continuously acquiring signals of N discrete points of a Doppler radar echo signal of a moving target at a preset sampling rate;
step S200: extracting all single-period cosine wave-like signals from the acquired discrete signals;
step S300: and calculating the frequency and power of each extracted cosine wave signal, and taking the frequency and power of the peak position of each cosine wave signal as a feature vector of the time-frequency domain of the moving target.
In order to know the motion condition of a moving target, a radar signal is transmitted to the moving target through a radar generating device, the radar signal is reflected when encountering the moving target, the reflected radar signal generates a Doppler effect, and a Doppler radar echo is formed and returns to a radar receiving device; by processing the received Doppler radar echo signals in the steps, the characteristics of the frequency and the power of the moving target in a time-frequency domain can be obtained.
In this embodiment, the sampling preset frequency of the doppler radar echo signal of the moving target is set as fsThen, the sequence formed by the amplitudes of the N discrete point signals collected each time is set as x [ N ]1]Wherein n is1N1, 2, 3.; the sequence of sampling timing positions of N discrete point signals is x [ N ]2]Wherein n is2N, 1,2,3. According to x [ n ]1]、x[n2]And classifying and extracting the discrete points according to the characteristics of each sequence unit to obtain the cosine wave-like signal. Of course, all the sine-like wave signals of a single period may also be extracted from the acquired discrete signals for processing, and the process of extracting the sine-like wave signals is performed with reference to the process of extracting cosine-like waves in this embodiment, which is not described herein again. All discrete points which can form the characteristic of the single-period cosine waves in the discrete signal are extracted, the frequency and the power of the discrete point at the position of the wave crest of each type of cosine wave signal are calculated to be used as a characteristic vector of the moving target at the moment corresponding to the discrete point at the position of the wave crest in the time-frequency domain, the characteristic vector represents an action of the moving target, all the single-period cosine waves are extracted, all the characteristic vectors are obtained, and the motion condition of the moving target in the duration corresponding to N sampled discrete points can be obtained equivalently. And according to the method, continuing the next sampling until all Doppler echo signals of the moving target are completely analyzed, and then depicting all motion characteristics of the moving target. By the time-frequency feature extraction of the method, the motion characteristics of the moving target can be rapidly obtained, the steps are simple, complex Fourier transform is not needed, and the computation amount is small.
Further, referring to fig. 2, the step S200 specifically includes:
step S210: sequentially extracting quasi-cosine wave signals of each single period according to the sequence of discrete signal acquisition;
step S220: judging whether each cosine wave-like signal is a noise signal;
step S230: and extracting the cosine wave-like signals which do not belong to the noise signals, and removing the cosine wave-like signals which belong to the noise signals.
In step S210, assuming that the extracted one-cycle cosine-like wave signal is composed of M discrete points of the N discrete points (M ≦ N), the sequence composed of the amplitudes of the M discrete point signals is set to y [ N ]1]Wherein n is11,2,3.. M; the sequence of the sampling timing positions of the M discrete point signals is y [ n ]2]Wherein n is2＝1,2,3...M。
Specifically, referring to fig. 3, step S210 specifically includes the following processes:
step S211: sequentially extracting a first valley point of the current cosine wave-like signal from the first discrete point signal, wherein the amplitude of the first valley point is smaller than the amplitude of the previous discrete point signal and the amplitude of the next discrete point signal;
according to the characteristics of cosine wave signal, the wave valley point is a sequence x [ n ]1]The amplitude in (1) is changed from a gradually smaller inflection point to a gradually larger inflection point, and the peak point is x [ n ]1]According to the characteristic that the amplitude value in (1) is changed from gradually increasing to gradually decreasing inflection point, the amplitude value can be sequentially changed from x [ n ]1]Each valley point and peak point is extracted.
In this embodiment, each cosine wave-like signal at least includes two adjacent valley points and one peak point sandwiched between the two valley points, and there may be several discrete points between the two valley points and the peak point.
Taking the example of extracting the first cosine wave-like signal as an example for explanation, the subsequent extraction process of the cosine wave-like signal refers to the extraction process of the first cosine wave-like signal, which is not repeated herein. Firstly, searching in sequence from a first discrete point, and when an inflection point that the amplitude of the first discrete point is changed from gradually decreasing to gradually increasing occurs, the inflection point is a first valley point of a first kind of cosine wave signal, namely the amplitude of the first valley point is smaller than that of a previous discrete point signal and that of a next discrete point signal.
Then, step S212 is performed to sequentially extract a peak point and a second valley point from the first valley point of the current cosine wave-like signal, where the amplitude of the peak point is greater than the amplitude of the previous discrete point signal and the amplitude of the next discrete point signal, and the amplitude of the second valley point is less than the amplitude of the previous discrete point signal and the amplitude of the next discrete point signal.
Referring to the foregoing, starting from the first valley point of the first cosine wave-like signal, when an inflection point occurs where the amplitude of the first discrete point changes from gradually increasing to gradually decreasing, the inflection point is the first peak point of the first cosine wave-like signal; when the inflection point with the amplitude of the discrete point changing from gradually decreasing to gradually increasing appears again, the inflection point is the second valley point of the first kind of cosine wave signal. All discrete points (including two valley points) between the first valley point and the second valley point constitute the first cosine wave-like signal, i.e. the first cosine wave-like extraction is completed.
In this case, the process proceeds to step S213, where the second valley point of the current cosine wave-like signal is used as the first valley point of the next cosine wave-like signal.
According to the above, the second valley point of the first cosine wave-like signal is used as the first valley point of the second cosine wave-like signal, and the peak point and the second valley point are searched again according to the above steps until all cosine wave-like signals are extracted. By this, the process of extracting the cosine wave-like signal of step S210 is completed.
In step S220, since there may be noise interference in the extracted motion target motion, noise determination is required to determine each y [ n ]1]And y [ n ]2]Whether each sequence unit meets the preset condition or not is judged, namely whether the extracted cosine wave signals are noise signals or not is judged. If y [ n ]1]And y [ n ]2]If each sequence unit in the sequence unit meets the preset condition, the extracted corresponding cosine wave-like signal is not a noise signal, and if y [ n ], (n)1]And y [ n ]2]If each sequence unit does not meet the preset condition, the extracted corresponding cosine wave-like signal is a noise signal.
Specifically, the y [ n ] is judged1]And y [ n ]2]Whether the represented cosine wave-like signal is a noise signal specifically includes the following three aspects:
judging whether the amplitude of a peak point of each cosine wave-like signal is smaller than a first preset value, if so, judging that the cosine wave-like signal is a noise signal; if not, judging that the cosine wave-like signal is not a noise signal.
In this embodiment, the intensity of the noise signal is usually not strong, and each y [ n ] is determined1]Whether the sequence unit of the peak point position in the signal is smaller than a first preset value or not is judged, if so, the cosine wave signal is considered as a noise signal, and if not, the cosine wave signal is considered as not the noise signal. In this embodiment, the first preset value may be determined according to the strength of the entire doppler radar echo signal.
Secondly, judging whether a first wave valley point and a second wave valley point of each cosine wave signal are approximately symmetrical relative to a wave peak point; if yes, judging that the cosine wave-like signal is not a noise signal; if not, judging that the cosine wave-like signal is a noise signal.
In a standard cosine wave signal, the positions of two wave valley points are symmetrical relative to the position of a wave peak point in one period, so if the extracted cosine wave-like signal is a noise signal, y [ n ] may exist1]The amplitude variation in the sequence is similar to that of a cosine wave signal, but the waveform actually formed by the sequence is greatly different from that of the cosine wave signal, so that whether the cosine wave is a noise signal is judged by judging whether the first valley point and the second valley point of each cosine wave signal are approximately symmetrical relative to the peak point.
In this embodiment, determining whether the first valley point and the second valley point of each cosine wave signal are substantially symmetrical with respect to the peak point includes the following two aspects:
(1) firstly, judging whether the positions of the two wave valley points are approximately symmetrical relative to the wave peak point or not, and comprising the following steps:
acquiring the time interval from the first valley point to the peak point and the time interval from the peak point to the second valley point of each cosine wave-like signal;
the time interval from the first valley point to the peak point can be obtained by the number of discrete points between the first valley point and the peak point, and in this embodiment, the time interval between every two adjacent discrete points is known as the sampling preset frequency
Assuming that a discrete points exist between the first valley point and the peak point, the time interval between the first valley point and the peak point is
In the same way, the time interval from the second valley point to the peak point can be obtained by the number of discrete points between the second valley point and the peak point, and in this embodiment, assuming that there are b discrete points between the second valley point and the peak point, the time interval between the second valley point and the peak point is
In this embodiment, the difference between the two time intervals is
Judging whether the first valley point and the second valley point of the cosine-like wave signal are approximately symmetrical relative to the peak point by judging whether the difference value of the time interval is smaller than a second preset value, and judging that the first valley point and the second valley point of the cosine-like wave signal are approximately symmetrical relative to the peak point when the difference value of the time interval is smaller than the second preset value; when the time interval difference is greater than or equal to a second preset value, the relative peak points of the first valley point and the second valley point of the cosine-like wave signalIs asymmetric.
As can be seen from the above expression of the time interval difference, in practice, the number difference of the discrete points between the two trough points and the peak point is determined to determine whether the cosine wave signal is substantially symmetrical, and in this embodiment, the setting range of the second preset value is usually less than or equal to 5/fs。
(2) Judging whether the amplitudes of the two wave valley points are approximately symmetrical relative to the wave peak point or not, and comprising the following steps: acquiring a first amplitude difference value between a first valley point and a second valley point of each cosine wave-like signal, an average amplitude value between the first valley point and the second valley point, and a second amplitude difference value between the amplitude value of the peak point and the average amplitude value, and calculating a ratio of the first amplitude difference value to the second amplitude difference value;
in this embodiment, it is assumed that the amplitude of the first valley point is x, the amplitude of the peak point is y, and the amplitude of the second valley point is z. The first amplitude difference between the first valley point and the second valley point is | x-z |, and the average amplitude of the first valley point and the second valley point is |
The second amplitude difference between the amplitude of the peak point and the average amplitude is
The ratio between the first amplitude difference and the second amplitude difference is
The ratio represents the ratio of the amplitude difference between two troughs to the amplitude difference of the whole cosine wave-like waveform. Comparing the ratio with a third preset value, and when the ratio is smaller than the third preset value, indicating that the amplitudes of the first valley point and the second valley point of the cosine wave signal relative to the peak point are approximately symmetrical; and when the ratio is greater than or equal to a third preset value, the first valley point and the second valley point of the cosine wave-like signal are asymmetric relative to the peak point. In the present embodiment, the third preset value range is preferably smaller than or equal toEqual to 0.4.
Thirdly, judging whether the cosine wave signals are correlated with standard cosine wave signals or not, and if yes, judging that the cosine wave signals are not noise signals; if not, judging that the cosine wave-like signal is a noise signal.
The first and second aspects are used for judging the rough shape of various types of cosine waves, and whether the signals are noise signals can be further judged through the accurate shapes of various types of cosine waves through correlation.
In this embodiment, the determining whether there is a correlation between each cosine wave signal and a standard cosine wave signal specifically includes:
firstly, a standard cosine wave signal sequence with the same number of discrete points and the same period as the discrete points of each type of cosine wave signal is constructed. In this embodiment, a sequence of amplitudes of discrete points of the standard cosine wave signal sequence is defined as
Wherein n is1M, 1,2,3. The standard cosine wave signal is a single cycle, the cycle starts from one wave trough and ends with the next wave trough. Each discrete point is equivalent to that M sampling points are averagely arranged in the 2 pi period of the standard cosine wave signal.
Calculating a correlation coefficient between each discrete point amplitude of the cosine wave-like signal and each discrete point amplitude of the corresponding standard cosine wave signal sequence; in the present embodiment, the correlation coefficient is corr.
The correlation coefficient of the two sequences can be calculated according to a calculation formula of the correlation coefficient:
In this embodiment, whether the quasi-cosine wave signal has a correlation with the standard cosine wave signal is determined by determining the magnitude of the correlation coefficient and a fourth preset value, and when the correlation coefficient is greater than the fourth preset value, the quasi-cosine wave signal is determined to have a correlation with the standard cosine wave signal; and when the correlation coefficient is smaller than or equal to a fourth preset value, judging that the cosine wave-like signal is not related to the standard cosine wave signal. In the present embodiment, the range of the fourth preset value is preferably greater than or equal to 0.8.
According to the above analysis, if there is a cosine-like wave signal having a low peak value or a poor relative peak-to-peak symmetry between two troughs or a poor correlation between the cosine-like wave signal and the standard cosine wave signal, the cosine-like wave signal is considered as a noise signal.
Therefore, after the noise signal and the non-noise signal are distinguished according to the above steps, the process proceeds to step S230, where the cosine-like wave signal corresponding to the noise signal is removed, and only the cosine-like wave signal not belonging to the noise signal is left for feature vector calculation. The accuracy of extracting the characteristic vector of the moving target in the time-frequency domain is improved, and a reliable data source is provided for the motion condition analysis of the moving target.
Further, in some embodiments, in step 300, the step of calculating the frequency and power of each extracted type of cosine wave signal comprises:
calculating a frequency value: acquiring the number of discrete points contained in each kind of cosine wave signals, and calculating the frequency of each kind of cosine wave signals according to the following formula:
because each cosine wave-like signal is a single-period signal, the time required by the single period of the cosine wave-like signal can be obtained according to the preset sampling frequency
(M-1 sampling intervals between M discrete points, each interval having a duration of
) Then, the frequency f is calculated by reciprocal of the frequency and the period of the quasi-cosine wave.
Calculating a power value: obtaining the average value of the amplitudes of all the discrete points and the amplitudes of the discrete points contained in each kind of cosine wave signals, and calculating the power of each kind of cosine wave signals according to the following formula:
Obtaining the frequency and power of each kind of cosine wave signal through the formula of the frequency and power, and generally taking the frequency and power of a discrete point at the peak position of each kind of cosine wave signal as a feature vector of a moment corresponding to the discrete point at the peak position of the moving target in a time-frequency domain; and finally, obtaining the time-frequency characteristics of the echo signals of the whole Doppler radar by referring to the method.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention, and all modifications and equivalents of the present invention, which are made by the contents of the present specification and the accompanying drawings, or directly/indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (7)
1. A time-frequency feature extraction method for a Doppler radar moving target is characterized by comprising the following steps:
continuously acquiring signals of N discrete points of a Doppler radar echo signal of a moving target at a preset sampling rate;
extracting all single-period cosine wave-like signals from the acquired discrete signals, wherein the method comprises the following steps: sequentially extracting quasi-cosine wave signals of each single period according to the sequence of discrete signal acquisition; judging whether each cosine wave-like signal is a noise signal; extracting cosine wave-like signals which do not belong to the noise signals, and removing cosine wave-like signals which belong to the noise signals;
calculating the frequency and power of each extracted cosine wave signal, including: acquiring the number of discrete points contained in each kind of cosine wave signals, and calculating the frequency of each kind of cosine wave signals according to the following formula:
wherein f is the frequency of each said kind of cosine wave signal, fsSetting a preset sampling rate for signal acquisition, wherein M is the number of discrete point signals contained in each cosine wave signal;
obtaining the average value of the amplitudes of all the discrete points and the amplitudes of the discrete points contained in each kind of cosine wave signals, and calculating the power of each kind of cosine wave signals according to the following formula:
where P is the power of each said cosine wave signal, y [ n ]1]For the set of amplitudes of the discrete points included in each of said cosine wave signals, n1M, ave is the average of all discrete point amplitudes;
and taking the frequency and the power of the peak position of each cosine wave-like signal as a feature vector of a moving target time-frequency domain.
2. The method for extracting time-frequency characteristics of a moving object of a doppler radar according to claim 1, wherein the step of sequentially extracting cosine wave-like signals of each single period according to the sequence of discrete signal acquisition comprises:
sequentially extracting a first valley point of the current cosine wave-like signal from the first discrete point signal, wherein the amplitude of the first valley point is smaller than the amplitude of the previous discrete point signal and the amplitude of the next discrete point signal;
sequentially extracting a peak point and a second valley point from the first valley point of the current cosine wave-like signal, wherein the amplitude of the peak point is greater than the amplitude of the signal of the previous discrete point and the amplitude of the signal of the next discrete point, and the amplitude of the second valley point is less than the amplitude of the signal of the previous discrete point and the amplitude of the signal of the next discrete point;
the second valley point of the current cosine wave-like signal is taken as the first valley point of the next cosine wave-like signal.
3. The method for extracting time-frequency characteristics of a moving object according to claim 1, wherein the step of determining whether each cosine wave-like signal is a noise signal comprises:
judging whether the amplitude of the peak point of each cosine wave-like signal is smaller than a first preset value or not;
if yes, judging that the cosine wave-like signal is a noise signal;
if not, judging that the cosine wave-like signal is not a noise signal.
4. The method of claim 3, wherein the step of determining whether each cosine wave-like signal is a noise signal further comprises:
judging whether the first wave valley point and the second wave valley point of each kind of cosine wave signals are approximately symmetrical relative to the wave peak point or not, comprising the following steps: obtaining the distance between the first valley point and the peak point of each cosine wave-like signalAnd the time interval from the peak point to the second valley point; judging whether the difference value of the two time intervals is smaller than a second preset value, if so, approximately symmetrical a first valley point and a second valley point of the cosine wave-like signal relative to a peak point; if not, the first wave valley point and the second wave valley point of the cosine wave-like signal are asymmetric relative to the wave peak point; wherein the second preset value is less than or equal to 5/fs；
If the first wave valley point and the second wave valley point of each cosine wave-like signal are approximately symmetrical relative to the wave peak point, judging that the cosine wave-like signal is not a noise signal; if not, judging that the cosine wave-like signal is a noise signal.
5. The method of claim 4, wherein the step of determining whether the first valley point and the second valley point of each cosine wave signal are substantially symmetrical with respect to the peak point further comprises:
acquiring a first amplitude difference value between a first valley point and a second valley point of each cosine wave-like signal, an average amplitude value between the first valley point and the second valley point, and a second amplitude difference value between the amplitude value of the peak point and the average amplitude value, and calculating a ratio of the first amplitude difference value to the second amplitude difference value;
judging whether the ratio is smaller than a third preset value or not, if so, approximately symmetrical a first valley point and a second valley point of the cosine wave-like signal relative to a peak point; if not, the first wave valley point and the second wave valley point of the cosine wave-like signal are asymmetric relative to the wave peak point; wherein the third preset value is less than or equal to 0.4.
6. The method of claim 4, wherein the step of determining whether each cosine wave-like signal is a noise signal further comprises:
judging whether the cosine wave signals are correlated with standard cosine wave signals or not; if yes, judging that the cosine wave-like signal is not a noise signal; if not, judging that the cosine wave-like signal is a noise signal.
7. The method for extracting time-frequency characteristics of a moving target according to claim 6, wherein the step of determining whether there is a correlation between each cosine wave signal and a standard cosine wave signal comprises:
constructing a standard cosine wave signal sequence with the same number of discrete points and the same period as the discrete points of each type of cosine wave signal;
calculating a correlation coefficient between each discrete point amplitude of the cosine wave-like signal and each discrete point amplitude of the corresponding standard cosine wave signal sequence;
and judging whether the correlation coefficient is larger than a fourth preset value, if so, determining that the quasi-cosine wave signal is correlated with the standard cosine wave signal, and if not, determining that the quasi-cosine wave signal is not correlated with the standard cosine wave signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669096.3A CN110361723B (en) | 2019-07-22 | 2019-07-22 | Time-frequency feature extraction method for Doppler radar moving target |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910669096.3A CN110361723B (en) | 2019-07-22 | 2019-07-22 | Time-frequency feature extraction method for Doppler radar moving target |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110361723A CN110361723A (en) | 2019-10-22 |
| CN110361723B true CN110361723B (en) | 2021-11-30 |
Family
ID=68220096
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910669096.3A Active CN110361723B (en) | 2019-07-22 | 2019-07-22 | Time-frequency feature extraction method for Doppler radar moving target |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110361723B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111812623B (en) * | 2020-07-29 | 2021-03-16 | 中国科学院声学研究所 | Target signal detection method and system based on time-frequency characteristic diagram |
| CN114324974B (en) * | 2021-12-16 | 2023-09-12 | 成都戎星科技有限公司 | Single-star radiation source passive monitoring target motion attribute distinguishing method |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7966156B1 (en) * | 2002-10-11 | 2011-06-21 | Flint Hills Scientific Llc | Method, computer program, and system for intrinsic timescale decomposition, filtering, and automated analysis of signals of arbitrary origin or timescale |
| CN102243302A (en) * | 2011-04-15 | 2011-11-16 | 东南大学 | Method for extracting line spectrum time accumulation characteristics of hydro-acoustic target radiation noise |
| CN104122539A (en) * | 2014-07-01 | 2014-10-29 | 四川大学 | Undersampling based high-speed movement target micro-doppler parameter estimation method |
| CN104792364A (en) * | 2015-04-10 | 2015-07-22 | 中铁大桥局集团武汉桥梁科学研究院有限公司 | Dynamic bridge parameter extracting system and dynamic bridge parameter extracting method based on laser Doppler |
| CN105044691A (en) * | 2015-06-03 | 2015-11-11 | 西安电子科技大学 | Rapid radar performance assessment method in sea cluster background |
| WO2016057781A1 (en) * | 2014-10-08 | 2016-04-14 | The University Of Florida Research Foundation, Inc. | Method and apparatus for non-contact fast vital sign acquisition based on radar signal |
| CN106377250A (en) * | 2016-09-21 | 2017-02-08 | 广州视源电子科技股份有限公司 | Marking equipment for data type of electroencephalogram signals in sleeping state |
| CN106580301A (en) * | 2016-12-21 | 2017-04-26 | 广州心与潮信息科技有限公司 | Physiological parameter monitoring method, device and hand-held device |
| EP3084465B1 (en) * | 2013-12-16 | 2017-09-13 | Elmos Semiconductor Aktiengesellschaft | Method for processing an echo signal of an ultrasonic transducer |
| CN107367640A (en) * | 2017-07-06 | 2017-11-21 | 北京理工大学 | Nonuniform envelope unknown pulse edge method of estimation and equipment based on phase information |
| CN108919249A (en) * | 2018-09-18 | 2018-11-30 | 湖北晧天智能科技有限公司 | A kind of radar target Joint estimation method based on two-dimentional local interpolation |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8127138B1 (en) * | 2008-09-29 | 2012-02-28 | The United States Of America As Represented By The Secretary Of The Navy | Method for embedding information in sonar |
| JP2011064558A (en) * | 2009-09-17 | 2011-03-31 | Toto Ltd | Doppler sensor system |
| CN102628941A (en) * | 2012-03-12 | 2012-08-08 | 陈明霞 | Reconstruction method of broadband ultrasonic echo signal |
| CN104935541B (en) * | 2014-03-19 | 2018-05-22 | 日本电气株式会社 | Doppler frequency shift searching method and device |
| CN105871765A (en) * | 2016-06-22 | 2016-08-17 | 南京索拉亚科技有限公司 | Wireless communication carrier wave tracking method based on FFT assistant S-PLL |
| CN108132396B (en) * | 2017-12-11 | 2020-11-03 | 北京经纬恒润科技有限公司 | Method and device for determining motion position |
-
2019
- 2019-07-22 CN CN201910669096.3A patent/CN110361723B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7966156B1 (en) * | 2002-10-11 | 2011-06-21 | Flint Hills Scientific Llc | Method, computer program, and system for intrinsic timescale decomposition, filtering, and automated analysis of signals of arbitrary origin or timescale |
| CN102243302A (en) * | 2011-04-15 | 2011-11-16 | 东南大学 | Method for extracting line spectrum time accumulation characteristics of hydro-acoustic target radiation noise |
| EP3084465B1 (en) * | 2013-12-16 | 2017-09-13 | Elmos Semiconductor Aktiengesellschaft | Method for processing an echo signal of an ultrasonic transducer |
| CN104122539A (en) * | 2014-07-01 | 2014-10-29 | 四川大学 | Undersampling based high-speed movement target micro-doppler parameter estimation method |
| WO2016057781A1 (en) * | 2014-10-08 | 2016-04-14 | The University Of Florida Research Foundation, Inc. | Method and apparatus for non-contact fast vital sign acquisition based on radar signal |
| CN104792364A (en) * | 2015-04-10 | 2015-07-22 | 中铁大桥局集团武汉桥梁科学研究院有限公司 | Dynamic bridge parameter extracting system and dynamic bridge parameter extracting method based on laser Doppler |
| CN105044691A (en) * | 2015-06-03 | 2015-11-11 | 西安电子科技大学 | Rapid radar performance assessment method in sea cluster background |
| CN106377250A (en) * | 2016-09-21 | 2017-02-08 | 广州视源电子科技股份有限公司 | Marking equipment for data type of electroencephalogram signals in sleeping state |
| CN106580301A (en) * | 2016-12-21 | 2017-04-26 | 广州心与潮信息科技有限公司 | Physiological parameter monitoring method, device and hand-held device |
| CN107367640A (en) * | 2017-07-06 | 2017-11-21 | 北京理工大学 | Nonuniform envelope unknown pulse edge method of estimation and equipment based on phase information |
| CN108919249A (en) * | 2018-09-18 | 2018-11-30 | 湖北晧天智能科技有限公司 | A kind of radar target Joint estimation method based on two-dimentional local interpolation |
Non-Patent Citations (1)
| Title |
|---|
| 弹丸转速的多普勒雷达信号分析;汪烊;《中国优秀博硕士学位论文全文数据库 (硕士） 工程科技二辑》;20070115;第1-75页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110361723A (en) | 2019-10-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105137498B (en) | A kind of the underground objects detection identifying system and method for feature based fusion | |
| CN111352102B (en) | Multi-target number detection method and device based on frequency modulation continuous wave radar | |
| Guan et al. | Adaptive fractional Fourier transform-based detection algorithm for moving target in heavy sea clutter | |
| CN104569948B (en) | Sub-band adaptive GLRT LTD detection methods under sea clutter background | |
| US20080106460A1 (en) | Radar microsensor for detection, tracking, and classification | |
| CN110361723B (en) | Time-frequency feature extraction method for Doppler radar moving target | |
| CN110988834B (en) | Pulse arrival time measuring method based on self-adaptive threshold value double thresholds | |
| CN109001708A (en) | The quick process of refinement method of radar maneuvering target based on classification integration detection | |
| CN109932695A (en) | A kind of method and device improving object recognition speed | |
| CN111999726B (en) | Personnel positioning method based on millimeter wave radar | |
| CN106534014A (en) | Accurate detection and separation method for multi-component LFM signal | |
| Abratkiewicz et al. | Radar signal parameters estimation using phase accelerogram in the time-frequency domain | |
| Yang et al. | Automatic algorithm for estimating the jet engine blade number from the radar target signature of aircraft targets | |
| CN114859297A (en) | Bridge vibration target detection method and device | |
| Shcherbakova et al. | Determination of characteristic points of electrocardiograms using multi-start optimization with a wavelet transform | |
| Vizinho et al. | Evaluation of the use of the modified-covariance method in HF radar ocean measurement | |
| CN105242272B (en) | Vehicle-mounted millimeter wave Anticollision Radar CFAR detection method based on auto-regressive time series model | |
| CN115345216A (en) | FMCW radar interference elimination method fusing prior information | |
| CN114355329A (en) | Method for detecting vital sign distance of frequency modulated continuous wave radar | |
| Sathe et al. | Late-time estimation methods to achieve resonance based radar target recognition | |
| CN108872954B (en) | Constant false alarm detection method based on correlation processing in same period | |
| EP1254379B1 (en) | Method for measurement of the radar target cross section of an object with both moving and fixed parts | |
| CN112363133A (en) | Target detection method, medium and device based on Doppler radar | |
| CN111337897A (en) | LFMCW radar rapid target identification method | |
| CN115166648B (en) | Low signal-to-noise ratio radar signal processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |