CN104655929A - Measuring method for digital time frequency of time domain signal and corresponding target identification method - Google Patents

Abstract

The invention provides a measuring method for digital time frequency of a time domain signal. The measuring method comprises the following steps: (1) setting time window length delta t according to the sampling rate v of a to-be-detected signal; confirming frequency spectrum range of temporal frequency measurement, wherein the frequency spectrum ranges form 1/delta T to v/2; (2) intercepting the to-be-detected signal through the time window; (3) setting discrete frequency points sequence in the certain frequency spectrum range; equaling frequency value to current frequency points towards each frequency point; taking two sinusoidal signals of which the phase difference is 90 degrees constantly; conducting related calculation on the current to-be-processed signal slices respectively; taking two related calculation results as real part and imaginary part calculation mould and argument respectively to obtain spectrum value and phase position value of a current time point and a current frequency point; (4) actuating the step (2) and the step (3) repeatedly until obtaining corresponding spectrum values and phase position values of the time point and frequency point combination of the to-be-texted signal. The measuring method can accurately measure phase position time-frequency spectrum while accurately measuring spectrum time-frequency spectrum , and is high in noise resisting capability , response speed and time resolving power.

Description

A kind of digital time-frequency measuring method of time-domain signal and corresponding target identification method

Technical field

The present invention relates to electronics and time and frequency measurement technical field, specifically, the present invention relates to a kind of digital time-frequency measuring method of time-domain signal.

Background technology

Frequency spectrograph is a kind of typical spectrum measuring device, and it is widely used in each fields such as electrotechnical, electronic, physical chemistry, biomedicine and national defense safety.Along with the progressively raising of application level and requirement, the requirement of people to frequency spectrograph function is also more and more higher.The application of such as ultrasonic diagnosis, cardiogram and pulsed radar aspect, frequency spectrograph not only needs to have excellent frequency spectrum resolution characteristic, also should possess excellent time resolution.

The popular frequency spectrograph in current domestic and international market is the frequency spectrograph based on Fourier transform, for amplitude spectrum, it possesses good frequency spectrum resolution characteristic and time resolution, can process the complicated signal of transient response and the differentiation of derivative signal thereof to a certain extent.But Fourier transform exists the window truncation effect of time domain, it needs to utilize the signal in limited time window to carry out the signal of approximate representation without limit, and this causes phase measurement and amplitude measurement to there is obvious deviation.Namely the existing frequency spectrograph based on Fourier transform is while the amplitude spectrum accurately measuring signal, its simultaneously measured by phase spectrum there is larger deviation even mistake.And phase spectrum can reflect the large measure feature of measured signal, if can not synchronously measure amplitude spectrum and phase spectrum exactly, be just difficult to information entrained in intactly analytic signal (signal of especially complicated transient response and derivative signal thereof).

For overcoming above-mentioned defect, Chinese patent application CN 101308175A proposes a kind of improvement project on Fourier transform basis, the program introduces some parameters and revises its phase spectrum in Fourier transform process, thus reduces phase deviation to a certain extent.But this scheme fundamentally can not change the time-domain window truncation effect in Fourier's change, and this scheme is likely because artificial parameter of introducing brings extra deviation, and therefore the accuracy of its phase spectrum still Shortcomings, has much room for improvement.

Therefore, currently the amplitude spectrum of time-domain signal and the solution of phase spectrum can synchronously be measured exactly in the urgent need to a kind of.

Summary of the invention

Task of the present invention is to provide a kind ofly can synchronously measure the amplitude spectrum of time-domain signal and the solution of phase spectrum exactly.

For achieving the above object, the invention provides a kind of digital time-frequency measuring method, comprise the following steps:

1) setting-up time length of window Δ T, determine the spectral range that can measure according to the sampling rate v of set time window length Δ T and measured signal, this spectral range is: 1/ Δ T to v/2;

2) intercept measured signal with time window, obtain the pending signal burst that current point in time is corresponding;

3) in step 1) the discrete frequency sequence of setting in the scope determined, for each frequency in frequency sequence, equal current frequency with frequency values, and constant phase difference is two sinusoidal reference signals of 90 degree, carries out correlation computations respectively to current pending signal burst; Using correlation calculation result corresponding for described two reference signals as real part and imaginary part composition plural number, calculate the modulus of complex number and argument that form, using described mould and argument as the range value of current point in time, current frequency and phase value;

4) next time point is set as current point in time, repeated execution of steps 2) to 3), until obtain each time point of measured signal and frequency combination corresponding to range value and phase value.

Wherein, described step 1) in, determined described spectral range is: 2/ Δ T to v/5.

Wherein, described step 3) also comprise: for each current frequency f, calculate the current period k/f corresponding to it, the one piece of data at pending signal burst end is cast out, to ensure that the time span of the pending signal participating in correlation computations is the integral multiple of the corresponding Cycle Length 1/f of current frequency from current described pending signal burst; Further, the time span of each described reference signal is all consistent with the time span of the pending signal of described participation correlation computations.

Wherein, described step 3) also comprise: by described step 1) the scope internal linear determined gets a little or a non-linear discrete frequency sequence of getting described in setting.

Wherein, described step 3) in, described non-linear getting a little comprises: logarithm is evenly got a little, and polynomial function is evenly got a little or inverse is evenly got a little.Namely in step 1) in the scope determined, along logarithm, polynomial expression or reciprocal function are evenly got a little on abscissa axis, the ordinate of taken point is exactly got value of frequency point, discrete frequency sequence obtained like this will along logarithm, polynomial expression or reciprocal function arrangement, thus obtain required time-frequency spectrum more neatly.

Wherein, described step 3) in, when the discrete frequency sequence described in setting adopt linearly to get some time, for any one frequency f in discrete frequency sequence, make time window length Δ T be the integral multiple of the Cycle Length 1/f that this frequency f is corresponding.

Compared with prior art, the present invention has following technique effect:

1, phase place time-frequency spectrum can accurately be measured while accurately measuring amplitude time-frequency spectrum.

2, anti-noise ability is strong.

3, fast response time.

4, time sense is high.

Accompanying drawing explanation

Fig. 1 shows the schematic diagram of the technology of the present invention principle;

Fig. 2 shows the process flow diagram of one embodiment of the invention;

Fig. 3 shows the time domain beamformer of a signal to be analyzed in one embodiment of the invention;

Fig. 4 shows the analyzed signal for Fig. 3, the phase and magnitude frequency spectrum (black box line represents) that the phase and magnitude frequency spectrum (gray line represents) surveyed according to one embodiment of the invention is surveyed with traditional Fourier transform scheme contrast schematic diagram;

Fig. 5 shows the time domain beamformer of another signal to be analyzed in one embodiment of the invention;

Fig. 6 shows the analyzed signal for Fig. 5, according to the amplitude spectrum that one embodiment of the invention is surveyed;

Fig. 7 shows the analyzed signal for Fig. 5, analyzes amplitude spectrogram according to the time-frequency combination that one embodiment of the invention is surveyed;

Fig. 8 shows the analyzed signal for Fig. 5, analyzes phase place spectrogram according to the time-frequency combination that one embodiment of the invention is surveyed;

Fig. 9 shows the time domain beamformer of another signal to be analyzed in one embodiment of the invention;

Figure 10 shows the analyzed signal for Fig. 9, according to the amplitude spectrogram that one embodiment of the invention is surveyed;

Figure 11 shows the analyzed signal for Fig. 9, the signal time domain distribution plan of the signal burst between 116 microseconds surveyed according to one embodiment of the invention under compared with very noisy situation to 124 microseconds and frequency spectrum profile.

Wherein a) b) be respectively not containing time domain and the frequency domain spectra of ground unrest during echo signal; C) time domain and the frequency domain spectra of the band noise targets signal of single measurement d) is respectively; E) average time domain and the frequency domain spectra of 10 sub-band noise targets signal measurements f) is respectively; G) average time domain and the frequency domain spectra of 100 sub-band noise targets signal measurements h) is respectively; I) average time domain and the frequency domain spectra of 1000 sub-band noise targets signal measurements j) is respectively.

Embodiment

Below, the present invention will be further described in conjunction with the accompanying drawings and embodiments.

For ease of understanding, first introduce measuring principle of the present invention.Fig. 1 shows the schematic diagram of the technology of the present invention principle, the present invention proposes method that phaselocked loop and correlator combine to the quick and precisely measurement of the amplitude and phase place that realize high density spectrum, the now measurement of phase place is directly carried out relevant to measured signal respectively by the two paths of signals of phaselocked loop and obtains, and without the need to repeatedly scanning phase place, greatly save the processing time.Specifically, the ultimate principle of the phase-locked time and frequency measurement that the present invention uses is: first set frequencies omega=2 π f to be measured, be responsible for producing the two-way sine wave signal that constant phase difference is 90 degree, as cos (ω t) and sin (ω t), i.e. complex signal e by phaselocked loop ^{i ω t}; Then the two paths of signals allowing phase place locked carries out the related operation as formula (1a) or (1b) respectively with measured signal, obtain the plural number of signal under this setpoint frequency (real part and imaginary part, or amplitude and phase place).Principle formula is expressed as follows:

$S\left(\mathrm{\ω}\right)=\frac{2}{T}{\∫}_{t=0}^T[S\left(t\right)\×e^{\mathrm{i\ωt}}]\mathrm{dt}=\frac{2}{T}{\∫}_{t=0}^T\{S\left(t\right)\×[\cos \left(\mathrm{\ωt}\right)+i\×\sin \left(\mathrm{\ωt}\right)\left]\right\}\mathrm{dt}$ (continuous situation) (1a)

$S\left(\mathrm{\ω}\right)=\frac{2}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}(S_k\×e^{\frac{\mathrm{i\ωk}}{\mathrm{fn}}})=\frac{2}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}\{S_k\×[\cos \left(\frac{\mathrm{\ωk}}{\mathrm{fn}}\right)+i\×\sin \left(\frac{\mathrm{\ωk}}{\mathrm{fn}}\right)\left]\right\}$ (discrete situation) (1b) wherein complex function S (ω) is the frequency spectrum of complex signal S (t), S _kfor complex discrete burst, i be imaginary unit namely-1 square root, n is each cycle numeral sample number, and N is the natural several times of n, and signal length T is the natural several times of 1/f.For common signal, the imaginary part of complex function S (t) is zero, now complex function S (t) equals real-number function S (t), and the real part of plural S (ω) and imaginary part represent respectively, namely wherein, according to formula (1a), plural S (ω) is represented respectively by real part and imaginary part, obtain formula (2a) and (3a), according to formula formula (1b), plural S (ω) is represented respectively by real part and imaginary part, obtains formula (2b) and (3b)

$S_{\mathrm{Re}}\left(\mathrm{\ω}\right)=\frac{2}{T}{\∫}_{t=0}^T[S\left(t\right)\×\cos \left(\mathrm{\ωt}\right)]\mathrm{dt}$ (continuous situation) (2a)

$S_{\mathrm{Re}}\left(\mathrm{\ω}\right)=\frac{2}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[S_k\×\cos \left(\frac{\mathrm{\ωk}}{\mathrm{fn}}\right)]$ (discrete situation) (2b)

$S_{\mathrm{Im}}\left(\mathrm{\ω}\right)=\frac{2}{T}{\∫}_{t=0}^T[S\left(t\right)\×\sin \left(\mathrm{\ωt}\right)]\mathrm{dt}$ (continuous situation) (3a)

$S_{\mathrm{Im}}\left(\mathrm{\ω}\right)=\frac{2}{N}\underset{k=1}{\overset{N}{\mathrm{\Σ}}}[S_k\×\sin \left(\frac{\mathrm{\ωk}}{\mathrm{fn}}\right)]$ (discrete situation) (3b)

It is to be noted, the ultimate principle used in the present invention, i.e. formula (1a) and (1b), with the integration of Fourier transform or sue for peace closely similar in form, this is because formula of the present invention possesses time domain convert to frequency domain or the real space feature generally had to density space.In fact, the phase-locked principle of time resolution that the present invention uses is different from Fourier transform in itself.Specifically, Fourier transform hypothesis original signal, in positive and negative infinite district, only has the orthogonality guarantee between any two frequencies when meeting this ideal hypothesis, and then frequency spectrum accuracy can be ensured.But this ideal situation does not exist in actual applications, because the time window of signal is not infinitely great in practical application, this departing from causes actual Fourier transform spectrum can not ensure continuously orthogonal, its accuracy on some Frequency point can part even completely lose, and in phase frequency spectrum, this problem is especially outstanding.The present invention for the disposal route of the actual signal of finite time length is: for the arbitrary frequency point in frequency spectrum, cycle according to its correspondence takes fully enough natural several times (such as 2 times from original signal (i.e. measured signal), 3 times etc.) time span, then the orthogonal signal being locked in 90 degree to a pair phase differential of this Frequency point are carried out relevant, as obtained signal respectively by formula (2b) and (3b) in the real part of this Frequency point and imaginary part, the corresponding time slicing of original signal just can be obtained at the range value of corresponding frequency and phase value.Further, the measurement of signal energy to this Frequency point that this scheme largely avoid other frequency brings interference, and thus the continuity of Frequency point and the measurement accuracy of each Frequency point can be guaranteed simultaneously.

Based on above-mentioned analysis, Fig. 2 shows the schematic flow sheet of a kind of digital time-frequency measuring method provided according to one embodiment of present invention, and the method comprises the following steps:

Step 101: obtain time-domain digital signal.This time-domain digital signal both can be the digital signal directly received, and also can be the digital signal obtained analog signal sampling.

Step 102: carry out initialization, time point variable Ti to be scanned is set, start time point is set to T0 (now i=0), setup times length of window Δ T, time point movement pace dt, determine the spectral range that can measure according to the sampling rate v of obtained time-domain digital signal, this spectral range between 1/ Δ T and v/2, wherein, 1/ Δ T is low frequency edge, too low frequency just there will be the obvious decline of measurement accuracy, and according to Shannon's sampling theorem, v/2 is high frequency limit.Preferably, this spectral range, between 2/ Δ T and v/5, can increase certain engineering nargin like this, and such as, when the upper limit gets v/5, upper frequency limit adds the engineering nargin of 2.5 times.

Generate frequency sequence according to determined spectral range, each frequency f in described frequency sequence all satisfies condition: the sampled point number of time window length Δ T is the integral multiple of the sampled point number corresponding to frequency f Cycle Length (i.e. 1/f).Such as: frequency of getting can be: 2/ Δ T, 3/ Δ T ..., n/ Δ T.Wherein n can select as required, such as, can be tens of to hundreds of, when selecting n, should note ensureing that n/ Δ T is less than the upper limit of determined spectral range, such as, being less than v/5.

For ease of describing, hereafter frequency is designated as Fj, j is the natural number of 1 to N, and j initial value is 1, N is the frequency points forming frequency sequence.

Step 103: the time-domain digital signal to be measured extracting Ti to the Ti+ Δ T period, obtains the time-domain digital signal fragment that current point in time Ti is corresponding.

Step 104: produce the phaselocked loop that Fj frequency is corresponding, namely producing constant phase difference is the two-way reference signal cos (ω t) of 90 degree and sin (ω t), and wherein ω is exactly the angular frequency corresponding with Fj frequency, ω=2 π Fj.Step 103 and step 104 can executed in parallel.

Step 105: carry out related operation with the time-domain signal in actual time window respectively with two-way reference signal cos (ω t) and sin (ω t), obtain two correlated results.

Step 106: two of step 105 correlated results are formed a plural number as real part and imaginary part, then mould and the argument of this plural number is calculated, using mould as time point Ti, the range value of frequency Fj, is designated as A (Ti, Fj), using argument as time point Ti, the phase value of frequency Fj, is designated as Ψ (Ti, Fj), then j increases 1 certainly.

Step 107: judge whether j is greater than N, if the judgment is No, get back to step 104, produce next frequency, and the continuation time-domain digital signal fragment corresponding with current point in time Ti carries out relevant treatment, if the judgment is Yes, then enters step 108.

Step 108:i is reset to 1 from increasing 1, j.Due to setup times point movement pace dt during initialization, therefore in this step, when i is in time increasing 1, Ti is from increasing dt.

Step 109: judge whether current Ti+ Δ T surmounts the scope of signal time-domain digital signal to be measured, if so, step 110 is entered, if not, then get back to step 103, produce next time-domain digital signal fragment and the phaselocked loop corresponding with current frequency Fj carries out related operation.

Step 110: export the time-frequency range value A (Ti that all Ti and Fj are corresponding, and time-frequency phase value Ψ (Ti Fj), Fj), the time-frequency amplitude spectrum (the two-dimentional spectrogram such as shown in Fig. 7) that composition is two-dimentional and time-frequency phase spectrum (the two-dimentional spectrogram such as shown in Fig. 8).

Referring again to Fig. 1, therefrom can find out that signal of the present invention flows to, more illustratively show the disposition of digital time domain signal to be measured in each stage.Digital time domain signal to be measured is first cut into time-domain signal fragment, then time-domain signal fragment and digital frequency conversion phaselocked loop are carried out related operation, obtain two-way correlated results, this two-way correlated results, respectively as a real and imaginary part, just can obtain amplitude time-frequency spectrum and phase place time-frequency spectrum by the mould of calculated complex and argument like this.

Further, in described step 102, when setting spectral range, the sampled point number in time window length Δ T is 2 times of the sampled point number corresponding to minimum frequency of determined spectral range.Now, the uncertainty of measurement of frequency, amplitude and phase place can be controlled within 10%.And in another embodiment, the sampled point number of time window length Δ T is the multiple value of the sampled point number corresponding to minimum frequency of determined spectral range also can be 3, now the uncertainty of measurement of frequency, amplitude and phase place can be down within 2%, but computation complexity can increase to some extent.Certainly, above-mentioned multiple also can the integer of value more than 3, and this is that those skilled in the art are understandable.

In previously described step 102, frequency sequence by linearly getting an acquisition in determined spectral range.In another embodiment of the invention, frequency sequence also can be obtained by a non-linear strategy of getting, and such as logarithm is evenly got point, polynomial function and evenly got point, inverse and evenly get a little etc.A this strategy of getting can obtain the frequency spectrum with required resolution more neatly.Simultaneously, adopt non-linear get tactful time, the one piece of data that length is the time window end of Δ T can be cast out, to ensure that the sampled point number of the time window participating in correlation computations is the integral multiple of the sampled point number corresponding to frequency f Cycle Length (i.e. 1/f), thus guarantee the accuracy of phase spectrum.

Be described further below by three the concrete effects of signal measurement to above-described embodiment.

Example 1: the frequency of frequency component each in unknown signaling, amplitude and phase place are measured

Set a signal, it is respectively 2.3kHz, 37.7kHz and 397.3kHz by three frequencies without direct frequency doubling relation, amplitude is respectively 1.1V, 0.2V and 0.7V, phase place is respectively 30 degree, 60 degree and is added with the sinusoidal signal of 120 degree and synthesizes, its expression formula, as shown in formula (4):

$S_1\left(t\right)=1.1\×\sin (2\mathrm{\πt}\×2300+\frac{\mathrm{\π}}{6})+0.2\×\sin (2\mathrm{\πt}\×37700+\frac{\mathrm{\π}}{3})+0.7\×\sin (2\mathrm{\πt}\×397300+\frac{\mathrm{\π}}{6})---\left(4\right)$

The time domain waveform of this signal as shown in Figure 3.Fig. 4 shows the analyzed signal for Fig. 3, the phase and magnitude frequency spectrum (black box line represents) that the phase and magnitude frequency spectrum (gray line represents) surveyed according to the present invention is surveyed with traditional discrete Fourier transformation scheme contrast schematic diagram.With reference to figure 4, can find out under logarithmic coordinate, the coordinate points that Fourier transform obtains is compared when low frequency and is wanted much sparse under high frequency, and no matter the frequency spectrum using the present invention to obtain can evenly be got a little under linear or logarithmic coordinate.In addition inventor also investigates the contrast that the two carries out window effect, and the frequency spectrum that Fourier transform obtains has significant change, and the present invention does not then have significant change.The result of unknown signaling and contrasting of setting value is measured by two kinds of frequency spectrum analysis methods, as shown in table 1, can find out that the frequency spectrum that the present invention obtains and the frequency spectrum that Fourier transform frequency spectrum obtains are at Frequency point, the measurement of amplitude especially phase place all has clear superiority, wherein last three row data are the measured value of comprehensive three Frequency points and the uncertainty of measurement mean value of setting value difference, Fourier pair frequency, the uncertainty of measurement of amplitude and phase place is respectively 3.9%, 5.5% and 98.8%, the uncertainty using the present invention to measure then is only 0.17% respectively, 1.1% and 6.7%.Table 1 shows and uses the present invention to contrast with the unknown signaling measurement result of usual discrete Fourier transformation frequency spectrum to example 1

Table 1

Example 2: paired pulses and derivative signal thereof carry out joint time frequency analysis

Sinc function combining simulation pulse waveform P (t) according to formula (5), designs three pulse P ₁(t), P ₂(t), P ₃(t), and white noise interference W (t) of voltage random variation between-1 and 1, according to formula (6) by four weighted arrays, the simulation waveform of final acquisition experiment pulse and derivative signal thereof, its time domain beamformer as shown in Figure 5, Fig. 6 shows the analyzed signal for Fig. 5, according to the amplitude spectrum that one embodiment of the invention is surveyed.Wherein,

P(t)＝sinc(2πt×300000)-2×sinc(2πt×600000)

$P_1\left(t\right)=\left\{\begin{array}{cc}0& (t<12.5\mathrm{\&mu;s})\\ P\left(t\right)& (t\&GreaterEqual;12.5\mathrm{\&mu;s})\end{array}\right.$

$P_2\left(t\right)=\left\{\begin{array}{cc}0& (t<37.5\mathrm{\&mu;s})\\ P\left(t\right)& (t\&GreaterEqual;37.5\mathrm{\&mu;s})\end{array}\right.$ (5)

$P_3\left(t\right)=\left\{\begin{array}{cc}0& (t<117\mathrm{\&mu;s})\\ P\left(t\right)& (t\&GreaterEqual;117\mathrm{\&mu;s})\end{array}\right.$

W(t)＝Random(-1,1)

S ₂(t)＝P ₁(t)+0.5P ₂(t)+0.2P ₃(t)+0.005W(t) (6)

Measured signal energy is mainly distributed between 100kHz and 1MHz as can be seen from Figure 6, in order to Integrative expression goes out the time-frequency characteristic that this signal comprises, the present invention is used to obtain time-frequency combination amplitude spectrogram and the phase place spectrogram of this signal, respectively as shown in Figure 7, Figure 8, the frequency spectrum differentiation in time of pulse signal and derivative signal thereof can therefrom be seen clearly.

Example 3: carried out anti-noise analysis by the echo signal that noise floods in special time period

This signal production process and example 2 similar, the practical application that just simulation is harsher, namely for a certain reason, P ₃t () pulse signal is highly suppressed, be reduced to original 5%, and ground unrest is strengthened to original 100 times, and building-up process is as formula 7:

S ₃(t)＝P ₁(t)+0.5P ₂(t)+0.01P ₃(t)+0.5W(t) (7)

Time domain waveform corresponding to this example signal as shown in Figure 9, Figure 10 shows the analyzed signal for Fig. 9, although it seems that Fig. 6 of the relative example 2 of its overall spectrum does not have notable difference from Figure 10, namely spectrum energy is still mainly distributed between 100kHz and 1MHz, but three feel the pulse and rush P3 (t) and almost cannot distinguish compared with Fig. 5.In order to accurately judge having that it's too late frequency spectrum being strong and weak of this time period signal, focus analysis is carried out to this time period.Figure 11 shows the analyzed signal for Fig. 9, the signal time domain distribution plan of the signal burst between 116 microseconds surveyed according to one embodiment of the invention under compared with very noisy situation to 124 microseconds and frequency spectrum profile.Wherein a) b) be respectively not containing time domain and the frequency domain spectra of ground unrest during echo signal; C) time domain and the frequency domain spectra of the band noise targets signal of single measurement d) is respectively; E) average time domain and the frequency domain spectra of 10 sub-band noise targets signal measurements f) is respectively; G) average time domain and the frequency domain spectra of 100 sub-band noise targets signal measurements h) is respectively; I) average time domain and the frequency domain spectra of 1000 sub-band noise targets signal measurements j) is respectively.As shown in figure 11, by background during Integrated comparative no pulse, and utilize repeatedly the method for synchronous cumulative mean, can't see P although can find out from time domain ₃(t) pulse, but by technology of the present invention, can judge that this signal exists really from the one-shot measurement of this time period frequency spectrum, and through to measure for 10 times and cumulative mean namely can its frequency spectrum of Accurate Determining at 500-700kHz, amplitude is at about 0.01V.

In reality test, the reflected signal of known target is received with certain radar, then the time-frequency spectrum (namely this reflected signal is at the range value of each frequency of set each time window and phase value) of this signal is obtained according to this reflected signal of methods analyst of the present invention, and then according to the amplitude spectrum of time-frequency spectrum gained and phase spectrum information, calculate the orientation and velocity of target, the true bearing of acquired results and this known target and speed are coincide.

Finally it should be noted that, above embodiment is only in order to describe technical scheme of the present invention instead of to limit this technical method, the present invention can extend in application other amendment, change, application and embodiment, and therefore think that all such amendments, change, application, embodiment are all in spirit of the present invention and teachings.

Claims (9)

1. a digital time-frequency measuring method for time-domain signal, comprises the following steps:

1) receive digital signal to be measured, according to the sampling rate v setting-up time length of window Δ T of measured signal, determine the spectral range of time and frequency measurement, described spectral range between 1/ Δ T to v/2, wherein 1/ Δ T<v/2;

2) intercept measured signal with time window, obtain the pending signal burst that current point in time is corresponding;

3) in step 1) the discrete frequency sequence of setting in the scope determined, for each frequency in discrete frequency sequence, current frequency is equaled with frequency values, and constant phase difference be two sinusoidal signals of 90 degree as with reference to signal, respectively correlation computations is carried out to current pending signal burst; Using correlation calculation result corresponding for two described reference signals as real and imaginary part, then calculate mould and the argument of described plural number, and using described mould and argument as the range value of current point in time, current frequency and phase value;

4) next time point is set as current point in time, repeated execution of steps 2) to 3), until obtain each time point of measured signal and frequency combination corresponding to range value and phase value.

2. the digital time-frequency measuring method of time-domain signal according to claim 1, is characterized in that, described step 1) in, determined described spectral range is: 2/ Δ T to v/5.

3. the digital time-frequency measuring method of time-domain signal according to claim 2, it is characterized in that, described step 3) also comprise: for each current frequency f, calculate the current period k/f corresponding to it, the one piece of data at pending signal burst end is cast out, to ensure that the time span of the pending signal participating in correlation computations is the integral multiple of the corresponding Cycle Length 1/f of current frequency from current described pending signal burst.

4. the digital time-frequency measuring method of time-domain signal according to claim 3, is characterized in that, described step 3) also comprise: the time span of each described reference signal is all consistent with the time span of the pending signal of described participation correlation computations.

5. the digital time-frequency measuring method of time-domain signal according to claim 3, it is characterized in that, described step 3) also comprise: by described step 1) the spectral range internal linear determined gets a little or a non-linear discrete frequency sequence of getting described in setting.

6. the digital time-frequency measuring method of time-domain signal according to claim 5, is characterized in that, described step 3) in, described non-linear getting a little comprises: logarithm is evenly got a little, and polynomial function is evenly got a little or inverse is evenly got a little.

7. the digital time-frequency measuring method of time-domain signal according to claim 3, it is characterized in that, described step 3) in, when the discrete frequency sequence described in setting adopt linearly to get some time, for any one frequency f in discrete frequency sequence, time window length Δ T is made to be the integral multiple of the Cycle Length 1/f that this frequency f is corresponding.

8. the digital time-frequency measuring method of time-domain signal according to claim 1, it is characterized in that, described step 3) in, two described reference signals are: cos (ω t) and sin (ω t), t represents the time, and ω represents the angular frequency corresponding to current frequency.

9. a target identification method, is characterized in that, comprises the following steps:

1) detection of a target obtains the reflected signal of target;

2) amplitude time-frequency spectrum and the phase place time-frequency spectrum of described reflected signal is obtained by the digital time-frequency measuring method of the time-domain signal according to any one of claim 1 ~ 6;

3) according to step 2) the amplitude time-frequency spectrum of gained and phase place time-frequency spectrum calculate the orientation and velocity of target.