CN105738878A - Precision measurement radar frequency measurement method based on frequency offset classification - Google Patents

Abstract

The invention discloses a precision measurement radar frequency measurement method based on a frequency offset classification, and relates to the field of precision measurement radar frequency measurement. The method comprises the following steps: obtaining a frequency shift of a radar signal on the basis of a frequency offset classification method, comparing the frequency shift with a reference threshold which is related to a second maximum value and a maximum value of an amplitude of the radar signal and is not a fixed value, and judging and calculating a frequency of a center according to the frequency shift. Measurement precision is improved through the measurement frequency being far away from a main lobe center by utilizing spectrum translation. Under the condition that a frequency offset is small, the frequency is measured directly using the FFT. Compared with existing methods, the method has obvious improvements in measurement precision and stability.

Description

A kind of precision instrumentation radar frequency measuring method based on frequency deviation classification

Technical field

The present invention relates to precision instrumentation radar frequency measurement field.

Background technology

At present, precision instrumentation radar frequency measuring method is very ripe, and including RIFE, power enhanced, FFT refines method.It is higher that traditional ratio method has frequency-measurement accuracy, and amount of calculation is less, it is easy to the advantage of realization, so the maximum precision instrumentation radar frequency measuring method become in engineering.Traditional ratio method step is: echo and transmitted wave are mixed and after filtering, obtain difference frequency signal by step 1.；Step 2. obtains Amplitude maxima X (k) after difference frequency signal is made FFT and second largest value X (k+r), r are ± 1；Step 3. obtains estimating frequency by the position of FFT maximum and the relational expression of secondary X (k) and X (k+r).Be we can see that by step 3, if measuring in noise circumstance, the signal sequence that frequency departure is less there will be the phenomenon of second largest value decision error and causes Frequency Estimation mistake, in order to anti-situation here occurs, this paper presents the precision instrumentation radar frequency measuring method based on frequency deviation classification.

Summary of the invention

The present invention is directed to a kind of precision instrumentation radar frequency measuring method based on frequency deviation classification of the not enough Curve guide impeller of background technology, the method obtains the frequency shift amount of radar signal based on frequency deviation sorting technique, this frequency shift amount is compared with reference threshold, the size of this reference threshold is relevant with the size of the second largest value of radar signal amplitude and maximum, is not fixed value；The method of the frequency at how many judgement calculating centers according to frequency shift amount；Utilize frequency spectrum translation with away from main lobe center to improve certainty of measurement；When frequency deviation is less, directly measure frequency with FFT.

The technical scheme is that a kind of precision instrumentation radar frequency measuring method based on frequency deviation classification, the method includes:

Step 1: through wave filter after being mixed with transmitted wave by echo, obtain difference frequency signal；

Step 2: obtain Amplitude maxima X (k after difference frequency signal is FFT₀) and second largest value X (k₀+r)；

As X (k₀+1)>X(k₀-1), r=1；As X (k₀+1)≤X(k₀-1), r=-1；

Step 3: utilize formula:

$\frac{X(k_0+r)}{X(k_0+r)+X\left(k_0\right)}=\frac{\mathrm{\Δ}k}{1}$

Obtain frequency spectrum shift value Δ k；Utilize formula:

$f_1=\frac{(k_0+\mathrm{\Δ}k)f_s}{N}$

Frequency f is estimated in acquisition₁；Wherein: f_sRepresenting sample frequency, N represents that FFT counts；

Step 4: obtain contrast frequency f₀: f₀=f_s.k/N(k∈[1,N])；

Step 5: if | f₁-f₀| < 0.02 Δ f then assert that mid frequency is f_e=f₀, wherein Δ f=f_s/N；

If | f₁-f₀| > 0.2 Δ f, then assert mid frequency f_e=f₁；

If 0.2 Δ f >=| f₁-f₀| >=0.02 Δ f, difference frequency signal step 1 obtained is multiplied by exp [j2 π nr (1/2-Δ k)/N], and wherein n represents the serial number of discrete series, then the signal after translation is carried out FFT, find out the frequency that Amplitude maxima is corresponding, assert frequency f centered by this frequency_e。

Further, the method obtaining contrast frequency in described step 4 is Amplitude maxima X (k in obtaining step 1₀) corresponding frequency, this frequency is contrast frequency f₀。

A kind of precision instrumentation radar frequency measuring method based on frequency deviation classification of the present invention, the method mutually than existing methods in certainty of measurement, be obviously improved in stability.

Accompanying drawing explanation

The simulated conditions of Fig. 1,2,3 is all FFT and counts as N=512, and sample frequency is set to fs=512khz.Beat signal frequency is set to f_IF∈ [30000,31000], differs 10hz between two.Do 1000 Monte Carlo Experiments.

The frequency measurement performance of tradition RIFE method that Fig. 1 is SNR when being 5dB and inclined factorization method

The frequency measurement performance of tradition RIFE method that Fig. 2 is SNR when being-3dB and inclined factorization method

The frequency measurement performance of tradition M-RIFE method that Fig. 3 is SNR when being 5dB and frequency offset factor method

Detailed description of the invention

A kind of precision instrumentation radar frequency measuring method based on frequency deviation classification, the method includes:

Step 1: through wave filter after being mixed with transmitted wave by echo, obtain difference frequency signal；

Step 2: obtain Amplitude maxima X (k after difference frequency signal is FFT₀) and second largest value X (k₀+r)；

As X (k₀+1)>X(k₀-1), r=1；As X (k₀+1)≤X(k₀-1), r=-1；

Step 3: utilize formula:

$\frac{X(k_0+r)}{X(k_0+r)+X\left(k_0\right)}=\frac{\mathrm{\&Delta;}k}{1}$

Obtain frequency spectrum shift value Δ k；Utilize formula:

$f_1=\frac{(k_0+\mathrm{\&Delta;}k)f_s}{N}$

Frequency f is estimated in acquisition₁；Wherein: f_sRepresenting sample frequency, N represents that FFT counts；

Step 4: obtain contrast frequency f₀: f₀=f_s.k/N(k∈[1,N])；

If step 5. | f₁-f₀| < 0.02 Δ f, f_e=k Δ f；

If | f₁-f₀| > 0.2 Δ f, f_e=f_IF,

If 0.2 Δ f >=| f₁-f₀| >=0.02 Δ f, makes FFT by y (n)=x (n) .exp [j2 π nr (1/2-Δ k)/N], obtains Amplitude maxima Y (k₁) and second largest value Y (k₁+r₁)；Wherein r₁=± 1, it is determined that method is consistent with r, k₁∈[1,N]；Obtain the frequency f after frequency spectrum shift₂=(k₁+r₁c₁) Δ f wherein c₁=Y (k₁+r₁)/[Y(k₁+r₁)+Y(k₁)]；Use f₂Deduct shifting frequency to deduct r (1/2-Δ k) Δ f and obtain final frequency f_e=[k₁+r₁c₁-r(1/2-Δk)].Δf。

For accompanying drawing, the present invention will be described:

Bias based on the precision instrumentation radar frequency measuring method of frequency deviation classified adaptive factor can ensure that son after frequency displacement between 0.2 Δ f to 0.5 Δ f, ensure after FFT maximum and second largest value amplitude very close to, namely second largest value is obvious with the third-largest value amplitude difference, under certain state of signal-to-noise, do not have and cause, because of the positive and negative decision error of r, the situation that estimated frequency error is very big.Fig. 1,2 is be respectively with tradition RIFE (traditional ratio method) and the business of inclined factorization method gained frequency measurement error and frequency measurement error in 5db situation in signal to noise ratio, as frequency deviation Δ k ∈ [-0.5,-0.2] ∪ [0.2,0.5], time, two kinds of method essence are the same, and measured result is consistent.The precision instrumentation radar Frequency Measurement Algorithm classified based on frequency offset factor when frequency deviation Δ k ∈ (-0.2,0.2) has all had than RIFE method in certainty of measurement, in stability and has been obviously improved.Frequency deviation is at 1.35 times to 13.23 times that Δ k ∈ (-0.2,0.2) frequency-measurement accuracy is tradition interpolation Frequency Measurement Algorithm, and stability is 1.2 times to 6.3 times of tradition interpolation Frequency Measurement Algorithm.And noiseproof feature is good.

Fig. 3 is under the situation tradition of SNR=5dB, the mean value error absolute value of tradition M-RIFE and frequency offset factor method.Bias tradition method partially is definite value due to son, in certain noise environment, the corresponding signal measurement precision for some frequency is still very low.Bias sub-method thresholding more accurately with careful, have employed optimal signal processing method for specific region, so certainty of measurement is high.

Claims (2)

1., based on a precision instrumentation radar frequency measuring method for frequency deviation classification, the method includes:

Step 1: through wave filter after being mixed with transmitted wave by echo, obtain difference frequency signal；

Step 2: obtain Amplitude maxima X (k after difference frequency signal is FFT₀) and second largest value X (k₀+r)；

As X (k₀+ 1) > X (k₀-1), r=1；As X (k₀+1)≤X(k₀-1), r=-1；

Step 3: utilize formula:

$\frac{X(k_0+r)}{X(k_0+r)+X\left(k_0\right)}=\frac{\mathrm{\&Delta;}k}{1}$

Obtain frequency spectrum shift value Δ k；Utilize formula:

$f_1=\frac{(k_0+\mathrm{\&Delta;}k)f_s}{N}$

Frequency f is estimated in acquisition₁；Wherein: f_sRepresenting sample frequency, N represents that FFT counts；

Step 4: obtain contrast frequency f₀: f₀=f_s.k/N(k∈[1,N])；

Step 5: if | f₁-f₀| < 0.02 Δ f, then assert that mid frequency is f_e=f₀, wherein Δ f=f_s/N；

If | f₁-f₀| > 0.2 Δ f, then assert mid frequency f_e=f₁；

If 0.2 Δ f >=| f₁-f₀| >=0.02 Δ f, difference frequency signal step 1 obtained is multiplied by exp [j2 π nr (1/2-Δ k)/N], and wherein n represents the serial number of discrete series, then the signal after translation is carried out FFT, find out the frequency that Amplitude maxima is corresponding, assert frequency f centered by this frequency_e。

2. as claimed in claim 1 a kind of based on frequency deviation classification precision instrumentation radar frequency measuring method, it is characterised in that in described step 4 obtain contrast frequency method be Amplitude maxima X (k in obtaining step 1₀) corresponding frequency, this frequency is contrast frequency f₀。