CN102445685A - Small spot radar signal decomposition method - Google Patents
Small spot radar signal decomposition method Download PDFInfo
- Publication number
- CN102445685A CN102445685A CN2011102983263A CN201110298326A CN102445685A CN 102445685 A CN102445685 A CN 102445685A CN 2011102983263 A CN2011102983263 A CN 2011102983263A CN 201110298326 A CN201110298326 A CN 201110298326A CN 102445685 A CN102445685 A CN 102445685A
- Authority
- CN
- China
- Prior art keywords
- parameter
- waveform
- radar signal
- value
- initial value
- 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.)
- Granted
Links
Images
Landscapes
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention relates to a small spot radar signal decomposition method. In the method, original radar signal waveform is taken as superposition of Gaussian waves, so as to estimate a Gaussian decomposition parameter initial value and optimize the value. The process of estimating the Gaussian decomposition initial value comprises the following steps of: extracting three parameters of one waveform with a maximum peak value in the current waveforms; subtracting corresponding component of the waveform from the original radar signal waveform; judging whether a termination condition effects, if so, exiting, otherwise, continuing to decompose the signal. The method can effectively decompose hybrid waveforms, particularly for the condition that the peaks are very closer; the method can accurately estimate the number of waveforms and the parameter initial value; in the method, pretreatments, such as filtering and the like, is not needed for the initial data; and the termination condition is easy to control. With the result of decomposition by the method, a more accurate and detailed digital ground model can be provided.
Description
Technical field
The present invention relates to field of information processing, particularly a kind of decomposition method of small light spot radar signal.
Background technology
On the theoretical foundation that laser wave graphic data Gaussian function decomposes, Liu Feng etc. have proposed to utilize Generalized Gaussian pattern function match pulse waveform, and propose the method (Central-South Forestry University science and technology journal 30 is rolled up in August, 2,010 8 phases) of pulse waveform important parameter.Small light spot laser radar (LiDAR) data can provide more accurately detailed digital terrain model.To the small light spot laser radar data, horse flood is superfine then to have proposed a kind of position that improved EM pulse detection method obtains echo-pulse and higher waveform decomposition method (2009 1 phases of remote sensing journal) of dependable performance, precision of width of using.
These technology have all realized the decomposition of LiDAR Wave data preferably; Also obtained certain parameter, the location is inaccurate but these technology exist initial value, the also inaccurate problem of wavelet shape number estimation; So just cause decomposition result often not restrain, can produce error.
Summary of the invention
To the deficiency of prior art, the present invention proposes a kind of small light spot radar signal decomposition method.
Technical scheme of the present invention is a kind of small light spot radar signal decomposition method; Original radar signal waveform
is considered as a plurality of high bass waves
,
... The stack of
Wherein,
expression waveform point reflection interval position; The truth of a matter of
expression natural logarithm; The amplitude of parameter
expression high bass wave
; The intermediate value of parameter
expression high bass wave
, the variance of parameter
expression high bass wave
;
By iterating on Gaussian
number
, and each Gaussian parameter
,
and
for estimation, parameter
,
and
initial value, and then optimized to obtain initial parameters
,
and
determine the value, according to a Gaussian
number
and parameters
,
and
The determined value decomposition raw radar signal waveform
,
Estimate that Gauss decomposes initial value and may further comprise the steps,
Step 2; Extract the maximum point of peak value in the current waveform; The horizontal ordinate of this point
is as the initial value of parameter
; Peak value is as the initial value of parameter
; Search peak is the horizontal ordinate of 0.5 *
point respectively along this peak value left and right sides; Obtain
and
; Get the littler person of
and
middle distance
and be designated as
, then the initial value of parameter
is (
-
)/0.8326;
Step 3; Initial value according to step 2 gained parameter
,
and
; From current waveform, deduct the corresponding component of high bass wave
, obtain a new Wave data;
Step 4; Judge whether to satisfy end condition; Satisfy end condition and then make n=
; Withdraw from then; Otherwise make
=
+1; The Wave data that step 3 gained is new is as current waveform, repeating step 2 and step 3; The combination of one or more that said end condition is the following option,
(2) Gaussian
of the area with raw radar signal waveform
of the area is less than the preset value than the threshold area;
(4) the high bass wave number that has decomposited
reaches the preset decomposition waveform number upper limit.
And, the initial value of parameter
,
and
is optimized adopts memory-limited BFGS algorithm or expectation-maximization algorithm or non-linear least square algorithm.
The present invention can effectively decompose hybrid waveform, particularly corrugation pitch from very near situation; Can accurately estimate waveform number and parameter initial value; Need not carry out pre-service such as filtering to raw data; End condition is controlled easily.Decompose the result who obtains like this, more accurately detailed digital terrain model can be provided.
Description of drawings
Fig. 1 is the process flow diagram of the embodiment of the invention.
Fig. 2 is the oscillogram after the embodiment of the invention iteration first time.
Fig. 3 is the oscillogram after the embodiment of the invention iteration second time.
Fig. 4 is the embodiment of the invention oscillogram after the iteration for the third time.
Fig. 5 is the oscillogram after the 4th iteration of the embodiment of the invention.
Fig. 6 is the oscillogram after the 5th iteration of the embodiment of the invention.
Fig. 7 is the oscillogram after the embodiment of the invention parameter optimization.
Embodiment
Specify technical scheme of the present invention below in conjunction with accompanying drawing and embodiment.
Embodiment is considered as a plurality of high bass waves
,
with original radar signal waveform
... The stack of
Wherein,
expression waveform point reflection interval position; The truth of a matter of
expression natural logarithm; The amplitude of parameter
expression high bass wave
; The intermediate value of parameter
expression high bass wave
, the variance of parameter
expression high bass wave
;
By iterating on Gaussian
number
, and each Gaussian parameter
,
and
for estimation, parameter
,
and
initial value, and then optimized to obtain initial parameters
,
and
determine the value, according to a Gaussian
number
and parameters
,
and
The determined value decomposition raw radar signal waveform
,
The Gauss who estimates is decomposed initial value to be optimized and can to adopt L-BFGS (memory-limited BFGS algorithm) or prior aries such as EM (expectation maximization) or LM (non-linear least square) algorithm.
Referring to Fig. 1, among the embodiment, estimate that Gauss decomposes initial value and may further comprise the steps,
Step 2; Extract the maximum point of peak value in the current waveform; The horizontal ordinate of this point
is as the initial value of parameter
; Peak value is as the initial value of parameter
; Search peak is the horizontal ordinate of 0.5 *
point respectively along this peak value left and right sides; Obtain
and
; Get the littler person of
and
middle distance
and be designated as
, then the initial value of parameter
is (
-
)/0.8326;
Step 3; Initial value according to step 2 gained parameter
,
and
; From current waveform, deduct the corresponding component of high bass wave
, obtain a new Wave data;
Step 4; Judge whether to satisfy end condition; Satisfy end condition and then make n=
; Withdraw from then; Otherwise make
=
+1; The Wave data that step 3 gained is new is as current waveform, repeating step 2 and step 3; The combination of one or more that said end condition is the following option,
(1) peak value of high bass wave
; (being the initial value of step 2 gained parameter
) is less than the peak value resolution of radar equipment;
(2) Gaussian
of the area with raw radar signal waveform
of the area is less than the preset value than the threshold area;
(4) the high bass wave number that has decomposited
reaches the preset decomposition waveform number upper limit.
All high bass waves that decomposited before at high bass wave
are designated as
,
...
, all high bass waves that decomposited
,
... The peak value of
is respectively parameter
,
... The initial value of
.Because each target of decomposing all is the highest waveform of contained peak value in the current waveform, the initial value of parameter
is at parameter
,
... Maximum in the initial value of
.The initial value that utilizes parameter
is as denominator, and the present invention has designed end condition (3) and can supply to select for use.The ratio of the peak value of the peak value of high bass wave
and high bass wave
stops during less than preset peakedness ratio threshold value, and the ratio of the peak value of the peak value of high bass wave
and high bass wave
is the continuation iteration during more than or equal to preset peakedness ratio threshold value.
During practical implementation, the peak value resolution of radar equipment is decided according to the radar equipment that produces original radar signal, and area can be set up on their own by those skilled in the art than threshold value, peakedness ratio threshold value and the decomposition waveform number upper limit as the case may be.For example area gets 0.02 than threshold value, and the peakedness ratio threshold value gets 0.1, decomposes the waveform number upper limit and gets 5.During practical implementation, one of above-mentioned four kinds of conditions be can select separately for use, two kinds or above combination also can be selected for use.Can be set in when satisfying any one condition and stop, also can be set in when satisfying two kinds or above condition simultaneously and stop.
The end condition of embodiment is made as the high bass wave number
that has decomposited and reaches the preset decomposition waveform number upper limit, decomposes the waveform number upper limit and gets 5.For ease of implement with reference to for the purpose of, it is following to provide embodiment to estimate that Gauss decomposes the concrete iterative process of initial value:
Iteration for the first time: operating procedure 1 and step 2, step 3; Obtain the parameter initial value
=52 of the maximum high bass wave
of peak value;
=25;
=7.207207; The result is as shown in Figure 2; Curve 1 is original radar signal waveform
among the figure, curve 2 first high bass wave
for decompositing.Curve 3 is the current waveform (this time for the raw radar signal waveform
) minus
results.
Iteration for the second time: operating procedure 4; Do not run into end condition; Current waveform is the result that original radar signal waveform
deducts
; Continue step 2,3; Obtain the parameter initial value
=32.316383 of second high bass wave
;
=40;
=6.006006; Decomposition result is as shown in Figure 3; Curve 1 is original radar signal waveform
among the figure, and curve 2 is first high bass wave
that decomposites and the stack of second high bass wave
.The result that curve 3 deducts
for current waveform, the result of stack that promptly original radar signal waveform
deducts
with
.
Iteration for the third time: operating procedure 4; Do not run into end condition, current waveform is that original radar signal waveform
deducts the result that
adds
.Continue step 2,3; Obtain the parameter initial value
=8.255985 of the 3rd Gaussian waveform
;
=35;
=2.402402; Decomposition result is as shown in Figure 4; Curve 1 is original radar signal waveform
among the figure, and curve 2 is the high bass wave
that decomposites, the stack of
and
.The result that curve 3 deducts
for current waveform, promptly original radar signal waveform
deducts
, the result of
stack with
.
The 4th iteration: operating procedure 4; Do not run into end condition, current waveform is that original radar signal waveform
deducts
, the result of
stack with
.Continue step 2 and step 3; Obtain the parameter initial value
=6.577729 of the 4th Gaussian waveform
;
=49;
=2.402402; Decomposition result is as shown in Figure 5; Curve 1 is original radar signal waveform
among the figure, and curve 2 is the stack of the high bass wave
that decomposites,
,
and
.The result that curve 3 deducts
for current waveform, promptly original radar signal waveform
deducts the result of
,
,
stack with
.
The 5th iteration: operating procedure 4; Do not run into end condition, current waveform is the result that original radar signal waveform
deducts
,
,
stack with
.Continue step 2 and step 3; Obtain the parameter initial value
=4.772379 of the 5th Gaussian waveform
;
=54;
=3.603604; Decomposition result is as shown in Figure 6; Curve 1 is an original waveform among the figure, and curve 2 is for decompositing the stack of high bass wave
,
,
,
and
.The result that curve 3 deducts
for current waveform, promptly original radar signal waveform
deducts the result of
,
,
,
and
Die Jia.
Operating procedure 4 runs into end condition (decomposition come out waveform number 5 reach the predetermined upper limit), accomplishes parameter estimation.
Embodiment uses the result of L-BFGS algorithm completion parameter optimization as shown in Figure 7.L-BFGS belongs to the plan Newton's algorithm, and it finds the solution the inverse matrix of the matrix of second derivatives of objective function to be optimized through alternative manner, and then the direction of search of definite optimum solution, along direction of search search optimum solution.Parameter optimization result is following:
Curve 1 is original radar signal waveform
among Fig. 7; Curve 2 is the curve of decomposition result
; Curve 3 is a graph of errors, expresses the poor of original radar signal waveform and decomposition result.
Accompanying drawing 2~7th, the match oscillogram, the overlapping more expression precision of curve is high more.
Specific embodiment described herein only is that the present invention's spirit is illustrated.Person of ordinary skill in the field of the present invention can make various modifications or replenishes or adopt similar mode to substitute described specific embodiment, but can't depart from spirit of the present invention or surmount the defined scope of appended claims.
Claims (2)
1. small light spot radar signal decomposition method; It is characterized in that: original radar signal waveform
is considered as a plurality of high bass waves
,
... The stack of
Wherein,
expression waveform point reflection interval position; The truth of a matter of
expression natural logarithm; The amplitude of parameter
expression high bass wave
; The intermediate value of parameter
expression high bass wave
, the variance of parameter
expression high bass wave
;
By iterating on Gaussian
number
, and each Gaussian parameter
,
and
for estimation, parameter
,
and
initial value, then the initial value for the optimized parameters
,
and
The determined value, according to the Gaussian
number
and the parameter
,
and
The determined value decomposition of raw radar signal waveform
,
Estimate that Gauss decomposes initial value and may further comprise the steps,
Step 2; Extract the maximum point of peak value in the current waveform; The horizontal ordinate of this point
is as the initial value of parameter
; Peak value is as the initial value of parameter
; Search peak is the horizontal ordinate of 0.5 *
point respectively along this peak value left and right sides; Obtain
and
; Get the littler person of
and
middle distance
and be designated as
, then the initial value of parameter
is (
-
)/0.8326;
Step 3; Initial value according to step 2 gained parameter
,
and
; From current waveform, deduct the corresponding component of high bass wave
, obtain a new Wave data;
Step 4; Judge whether to satisfy end condition; Satisfy end condition and then make n=
; Withdraw from then; Otherwise make
=
+1; The Wave data that step 3 gained is new is as current waveform, repeating step 2 and step 3; The combination of one or more that said end condition is the following option,
(2) Gaussian
to the area of the original radar signal waveform
The area is less than a preset threshold area ratio;
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110298326 CN102445685B (en) | 2011-09-28 | 2011-09-28 | Small spot radar signal decomposition method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN 201110298326 CN102445685B (en) | 2011-09-28 | 2011-09-28 | Small spot radar signal decomposition method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102445685A true CN102445685A (en) | 2012-05-09 |
CN102445685B CN102445685B (en) | 2013-05-08 |
Family
ID=46008368
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN 201110298326 Expired - Fee Related CN102445685B (en) | 2011-09-28 | 2011-09-28 | Small spot radar signal decomposition method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102445685B (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105044702A (en) * | 2015-09-18 | 2015-11-11 | 宁波华仪宁创智能科技有限公司 | Fitting method for pulse waveforms |
CN104408018B (en) * | 2014-11-19 | 2018-02-09 | 武汉大学 | A kind of conformal LiDAR waveforms denoising method and system |
CN111077532A (en) * | 2019-11-22 | 2020-04-28 | 同济大学 | Surface feature space information acquisition method based on deconvolution and Gaussian decomposition |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2914432A1 (en) * | 2007-03-27 | 2008-10-03 | Thales Sa | Stationary signal i.e. synthetic aperture radar video signal, extrapolating method for remote or Doppler high resolution analysis of signal, involves performing spectral decomposition of sub-signals and concatenation of spectrums |
CN101833087A (en) * | 2010-05-13 | 2010-09-15 | 王成 | Satellite-bone laser radar waveform data resolving method based on wavelet analysis |
CN101923158A (en) * | 2009-06-08 | 2010-12-22 | 霍尼韦尔国际公司 | Be used for Gauss's decomposition of weather radar data for communication for the system and method that transmits |
US8032319B1 (en) * | 2007-06-29 | 2011-10-04 | Hal Enterprises, LLC | Methods for analyzing streaming composite waveforms |
-
2011
- 2011-09-28 CN CN 201110298326 patent/CN102445685B/en not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2914432A1 (en) * | 2007-03-27 | 2008-10-03 | Thales Sa | Stationary signal i.e. synthetic aperture radar video signal, extrapolating method for remote or Doppler high resolution analysis of signal, involves performing spectral decomposition of sub-signals and concatenation of spectrums |
US8032319B1 (en) * | 2007-06-29 | 2011-10-04 | Hal Enterprises, LLC | Methods for analyzing streaming composite waveforms |
CN101923158A (en) * | 2009-06-08 | 2010-12-22 | 霍尼韦尔国际公司 | Be used for Gauss's decomposition of weather radar data for communication for the system and method that transmits |
CN101833087A (en) * | 2010-05-13 | 2010-09-15 | 王成 | Satellite-bone laser radar waveform data resolving method based on wavelet analysis |
Non-Patent Citations (1)
Title |
---|
黄涛 等: "激光雷达反遮蔽探测技术研究", 《光电工程》, vol. 38, no. 3, 31 March 2011 (2011-03-31) * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104408018B (en) * | 2014-11-19 | 2018-02-09 | 武汉大学 | A kind of conformal LiDAR waveforms denoising method and system |
CN105044702A (en) * | 2015-09-18 | 2015-11-11 | 宁波华仪宁创智能科技有限公司 | Fitting method for pulse waveforms |
CN111077532A (en) * | 2019-11-22 | 2020-04-28 | 同济大学 | Surface feature space information acquisition method based on deconvolution and Gaussian decomposition |
Also Published As
Publication number | Publication date |
---|---|
CN102445685B (en) | 2013-05-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | Wind power forecasting based on singular spectrum analysis and a new hybrid Laguerre neural network | |
Hu et al. | High resolution wind speed forecasting based on wavelet decomposed phase space reconstruction and self-organizing map | |
CN102967854B (en) | Multi-fractal detection method of targets in FRFT (Fractional Fourier Transformation) domain sea clutter | |
CN102608620B (en) | Laser scanning point cloud vegetation filtering method on basis of reflection strength and terrain | |
CN103729637A (en) | Extended target probability hypothesis density filtering method based on cubature Kalman filtering | |
CN107942308A (en) | Adaptive Rao detection methods based on gamma texture under complex Gaussian environment | |
CN104331583B (en) | A kind of Multifractal Modeling method based on Observed sea clutter | |
CN103927556A (en) | ECG signal classifying method based on wavelet packet and approximate entropy | |
CN102445685A (en) | Small spot radar signal decomposition method | |
CN105527617A (en) | Ground penetrating radar data background removing method based on robust principal component analysis | |
CN108804736A (en) | A kind of method and apparatus multiple degrees of freedom FM signal design and optimized | |
CN105676205A (en) | Airborne LiDAR waveform data Gaussian decomposition method | |
CN104268567A (en) | Extended target tracking method using observation data clustering and dividing | |
Alexopoulos et al. | Fractional order Pareto distributions with application to X‐band maritime radar clutter | |
CN110954860B (en) | DOA and polarization parameter estimation method | |
CN104215939A (en) | Knowledge assisted space-time adaptive processing method integrating generalized symmetrical structure information | |
CN106019257A (en) | Interpolation method based on high-frequency ground-wave radar ocean current observation result space-time characteristics | |
CN101739718A (en) | Parameter template-based corn leaf virtual simulation modeling method | |
CN103308829A (en) | GIS (gas insulated switchgear) single-time partial discharge signal extraction and trigger moment adjusting method | |
CN106125039B (en) | Improvement space-time adaptive Monopulse estimation method based on local Combined Treatment | |
CN103049653A (en) | G0 distribution parameter maximum likelihood estimation method based on EM (Expectation-Maximization) algorithm | |
CN103218807A (en) | Multispectral remote-sensing image change detection method | |
CN110658517B (en) | Dimensionality reduction sparse STAP method and device based on uncertain priori knowledge | |
DE102020107429A1 (en) | Determination of the temporal position of a signal | |
CN110516806A (en) | The rarefaction method and apparatus of neural network parameter matrix |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130508 Termination date: 20190928 |
|
CF01 | Termination of patent right due to non-payment of annual fee |