CN106772227B - A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic - Google Patents
A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic Download PDFInfo
- Publication number
- CN106772227B CN106772227B CN201710021562.8A CN201710021562A CN106772227B CN 106772227 B CN106772227 B CN 106772227B CN 201710021562 A CN201710021562 A CN 201710021562A CN 106772227 B CN106772227 B CN 106772227B
- Authority
- CN
- China
- Prior art keywords
- band
- signal
- frequency
- unmanned plane
- feature sub
- 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
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
- G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
- G01S3/80—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic or infrasonic waves
- G01S3/802—Systems for determining direction or deviation from predetermined direction
-
- 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
- G01S3/00—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received
- G01S3/80—Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received using ultrasonic, sonic or infrasonic waves
- G01S3/8003—Diversity systems specially adapted for direction finding
Abstract
The invention discloses a kind of unmanned plane direction determining methods based on the identification of vocal print multiple-harmonic.Traditional incoherent subspace DOA estimation algorithm based on High-Resolution Spectral, is divided into each narrow band signal for broadband signal, carries out angle estimation with multiple signal classification (MUSIC) algorithm on narrowband, takes arithmetic average.Conventional method has ignored the energy unevenness and signal-to-noise ratio difference of sub-band, causes algorithm performance bad.The method of the present invention does time frequency analysis by the voice signal to target information source, obtains characteristic frequency.The sub-band where characteristic frequency is only extracted to broadband signal, narrowband MUSIC operation is done, energy ratio, signal-to-noise ratio in conjunction with each feature sub-band assign each feature sub-band deflection different weighting coefficients, weighted average obtains angle estimation as a result, to reach more accurate angle estimation performance.
Description
Technical field
The present invention relates to unmanned plane sound positioning fields, and in particular to a kind of unmanned plane side based on the identification of vocal print multiple-harmonic
To estimation method.
Background technique
The method of general passive sound direction estimation specifically includes that (1) is based on the location technology of reaching time-difference (TDOA)；
(2) steerable beam based on peak power output forms technology；(3) it is based on High-Resolution Spectral Estimation technology.Relative to first two
Technology can use less array number based on High-Resolution Spectral Estimation technology, higher resolution ratio be obtained, in direction estimation
It is used widely.
But traditional High-Resolution Spectral Estimation algorithm, generally be directed to narrow band signal, most classic is based on matrix
Multiple signal classification (MUSIC) algorithm of Eigenvalues Decomposition.In the Wideband Signal Processing, the subspace of incoherent signal is proposed
(ISM) broadband signal in frequency domain decomposition is narrow-band component, using based on Eigenvalue Decomposition on each narrowband by algorithm
Multiple signal classification MUSIC algorithm carries out angle estimation, then takes arithmetic average.The algorithm is disadvantageous in that, is had ignored
The Energy distribution unevenness and signal-to-noise ratio of sub-band are different, caused by the bad problem of algorithm estimation performance.
Summary of the invention
To solve the above problems, the present invention provides a kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic.
The purpose of the present invention is achieved through the following technical solutions: a kind of unmanned plane based on the identification of vocal print multiple-harmonic
Direction determining method includes the following steps:
(1) voice signal for handling unmanned plane takeoff phase compares the time-frequency figure for front and back of taking off, obtains characteristic frequency section
[f1L,f1R],[f2L,f2R],…,[fNL,fNR], [f1L,f1R] it is fundamental frequency, remaining is the harmonic frequency of integral multiple.
(2) by unmanned plane during flying time domain observation signal when is divided into K subsegment, and every time is Td, every section of sampling number
For J point, the discrete Fourier transform (DFT) of J point is carried out to every segment data, obtains K frequency domain snap.
(3) on each frequency domain snap, according to the characteristic frequency section that step (1) obtains, spectrum peak search is carried out, by K
Frequency domain snap is in characteristic frequency section [fjL,fjR] the corresponding frequency values of spectral peak be averaging, obtain characteristic frequency section [fjL,
fjR] feature sub-band fj, j=1 ..., N.
(4) to each feature sub-band fj, angle, θ (f is calculated using narrowband MUSICj): K frequency domain snap is utilized, is calculated
Covariance matrix, then characteristics of decomposition value, obtain noise subspace and signal subspace.Utilize the direction vector and noise of signal source
The orthogonal property of subspace constructs the space MUSIC spectral function, and search spectrum peak, the corresponding angle of spectral peak is deflection θ (fj)。
(5) energy ratio, the signal-to-noise ratio for combining each feature sub-band, assign each feature sub-band deflection different weighting systems
Number, weighted average, obtains the direction of arrival angular estimation θ in broadband, the i.e. deflection of unmanned plane:
Wherein,wjIndicate feature sub-band fjEnergy (j=1 ..., N), wsumIndicate each feature frequency
The gross energy of band, reality take spectrum value when calculating；It is θ (fj) weighting coefficient (j=1 ..., N), indicate feature son frequency
Band fjEnergy account for the ratio of gross energy.The energy of general characteristics of low-frequency sub-band is higher and noise is relatively high, estimates
Deflection confidence level with higher, accounts for the large percentage of the deflection finally estimated.
A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic proposed by the present invention, to specific frequency
Unmanned plane acoustical signal angle estimation accuracy it is high, error is small.Compared with prior art, the present invention has the advantage that
1. traditional ISM algorithm needs broadband to be evenly dividing sub-band, MUSIC fortune in narrowband is carried out to each sub-band
It calculates.The present invention only needs to extract several feature sub-bands, narrowband MUSIC calculating is carried out, when bandwidth is bigger, with several features
The calculated angle Synthesize estimation direction of arrival angle of sub-band can obtain preferably estimating accuracy, while calculation amount is significantly
It reduces.
2. traditional ISM algorithm simply seeks arithmetic average after acquiring deflection to each sub-band, error is big, and performance is poor.
And this method is the energy ratio and signal-to-noise ratio in conjunction with each feature sub-band, assigns each feature sub-band deflection different weighting systems
Number, the angle estimation for being weighted and averaged to the end is as a result, to reach more accurately estimation performance.
Detailed description of the invention
Fig. 1 is a kind of flow chart of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic of the invention；
Fig. 2 is the angle estimation effect picture of the method for the unmanned plane audio direction estimation of conventional method；
Fig. 3 is the angle estimation effect picture of the method for the present invention.
Specific embodiment
The present invention is described in further detail below in conjunction with the drawings and specific embodiments.
As shown in Figure 1, a kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic provided by the invention, including
Following steps:
(1) voice signal for handling unmanned plane takeoff phase compares the time-frequency figure for front and back of taking off, obtains characteristic frequency section
[f1L,f1R],[f2L,f2R],…,[fNL,fNR], [f1L,f1R] it is fundamental frequency, remaining is the harmonic frequency of integral multiple.It is specific as follows:
The frequency of any time t is obtained using Short Time Fourier Transform (STFT) to the voice data of unmanned plane takeoff phase
Spectral power changes with the spectrum power after taking off before watch the takeoff, observes the characteristic frequency section of unmanned plane.
Voice data when to unmanned plane during flying hovering does discrete Fourier transform (DFT), obtains in conjunction with according to time-frequency figure
Characteristic frequency section, obtain more accurate characteristic frequency section.The sound of unmanned plane can be regarded as discrete broadband letter
Number, it is made of a series of narrow band signal.Under certain sample frequency, the section [f of available N number of characteristic frequency1L,
f1R],[f2L,f2R],…,[fNL,fNR]。
(2) by unmanned plane during flying time domain observation signal when is divided into K subsegment, and every time is Td, every section of sampling number
For J point, the discrete Fourier transform (DFT) of J point is carried out to every segment data, obtains K frequency domain snap.It is denoted as Xk(fj), k=1,
2,...,K。
(3) on each frequency domain snap, according to the characteristic frequency section that step (1) obtains, spectrum peak search is carried out respectively, is obtained
To the sub-band (assuming that having N number of) of characteristic frequency.It is specific as follows:
By K frequency domain snap in characteristic frequency section [f1L,f1R] the corresponding frequency values of spectral peak be averaging, obtain frequency
f1, that is, it is characterized sub-band f1.Similarly obtain feature sub-band f1,...,fN。
(4) to each feature sub-band fj, narrowband can be regarded as, calculate angle, θ (f using narrowband MUSICj):
Using K frequency domain snap, covariance matrix is calculated,
Parameter interpretation: K is the number of frequency domain snap, and X is frequency domain snap.Rx(fj) it is feature sub-band fjCovariance square
Battle array.
To covariance matrix Rx(fj) carry out Eigenvalues Decomposition (assuming that array element number is M, target information source number is D).
Wherein, biIt is and eigenvalue λiCorresponding feature vector.
Characteristic value is arranged according to sequence from big to small: λ1≥λ2≥...≥λD≥...≥λM, then before the big spies of D
Value indicative to the characteristic value of induction signal, behind M-D small characteristic values correspond to the characteristic value of noise.The corresponding spy of the characteristic value of signal
It levies vector and constitutes signal subspace, the corresponding feature vector of the characteristic value of noise constitutes noise subspace.Noise subspace is denoted as:
En=[vD+1,vD+2,...,vM]。
To arbitrary characteristics sub-band fj, j=1 ..., N utilize the principle of narrowband MUSIC, construct space spectral functionSearch spectrum peak, the corresponding angle of spectral peak are deflection θ (fj).Wherein, a (fj) it is signal
The direction vector in source, EnIt is noise subspace.
To all feature sub-band f in entire broadbandj, j=1 ..., N do same processing, estimate deflection θ
(fj), j=1 ..., N.
(5) energy ratio, the signal-to-noise ratio for combining each feature sub-band, assign each feature sub-band deflection different weighting systems
Number, weighted average, obtains the direction of arrival angular estimation θ in broadband, the i.e. deflection of unmanned plane:
Wherein,wjIndicate feature sub-band fjEnergy (j=1 ..., N), wsumIndicate each feature
The gross energy of frequency band, reality take spectrum value when calculating；It is θ (fj) weighting coefficient (j=1 ..., N), indicate feature son
Frequency band fjEnergy account for the ratio of gross energy.The energy of general characteristics of low-frequency sub-band is higher and noise is relatively high, estimates
Deflection confidence level with higher, account for the large percentage of the deflection finally estimated.
Embodiment
Smart -3 quadrotor drones flight of big boundary is controlled in cross acoustic array region overhead, is run on computer
Labview program, by the microphone collected sound signal of acoustic array, sample frequency 5120HZ, array element spacing is 0.17m,
Array number is 3, and target information source number is 1.The angle for defining the normal plane of sound and acoustic array is deflection, range [- 90,90].
Under current sampling frequency, available characteristic frequency section: [140,200], [300,380], [470,550], [650,
740].(unit: HZ).
Fig. 2 and Fig. 3 is the effect picture of the unmanned plane sound angle estimation of conventional method and the method for the present invention respectively, and comparison can
To find out that the method for the present invention can preferably estimate that the angle of unmanned plane sound, accuracy are substantially better than conventional method.
Claims (1)
1. a kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic, which comprises the steps of:
(1) voice signal for handling unmanned plane takeoff phase compares the time-frequency figure for front and back of taking off, obtains characteristic frequency section [f1L,
f1R],[f2L,f2R],…,[fNL,fNR], [f1L,f1R] it is fundamental frequency, remaining is the harmonic frequency of integral multiple；
(2) by unmanned plane during flying time domain observation signal when is divided into K subsegment, and every time is Td, every section of sampling number is J point,
The discrete Fourier transform that J point is carried out to every segment data, obtains K frequency domain snap；
(3) on each frequency domain snap, according to the characteristic frequency section that step (1) obtains, spectrum peak search is carried out, by K frequency domain
Snap is in characteristic frequency section [fjL,fjR] the corresponding frequency values of spectral peak be averaging, obtain characteristic frequency section [fjL,fjR]
Feature sub-band fj, j=1 ..., N；
(4) to each feature sub-band fj, angle, θ (f is calculated using narrowband MUSICj): K frequency domain snap is utilized, association side is calculated
Poor matrix, then characteristics of decomposition value, obtain noise subspace and signal subspace；It is empty using direction vector and noise of signal source
Between orthogonal property, construct the space MUSIC spectral function, search spectrum peak, the corresponding angle of spectral peak is deflection θ (fj)；
(5) energy ratio, the signal-to-noise ratio for combining each feature sub-band, assign each feature sub-band deflection different weighting coefficients,
Weighted average, obtains the direction of arrival angular estimation θ in broadband, the i.e. deflection of unmanned plane:
Wherein,wjIndicate feature sub-band fjEnergy, wsumIndicate the gross energy of each feature sub-band, it is practical
Spectrum value is taken when calculating；It is θ (fj) weighting coefficient, indicate feature sub-band fjEnergy account for the ratio of gross energy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710021562.8A CN106772227B (en) | 2017-01-12 | 2017-01-12 | A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710021562.8A CN106772227B (en) | 2017-01-12 | 2017-01-12 | A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106772227A CN106772227A (en) | 2017-05-31 |
| CN106772227B true CN106772227B (en) | 2019-04-12 |
Family
ID=58947969
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710021562.8A Active CN106772227B (en) | 2017-01-12 | 2017-01-12 | A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106772227B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107576387B (en) * | 2017-08-18 | 2020-05-22 | 浙江大学 | Unmanned aerial vehicle detection method based on voiceprint multi-harmonic recognition |
| CN107564530A (en) * | 2017-08-18 | 2018-01-09 | 浙江大学 | A kind of unmanned plane detection method based on vocal print energy feature |
| CN107800658B (en) * | 2017-11-10 | 2020-05-08 | 九江学院 | Two-dimensional harmonic signal frequency estimation method |
| CN108375763B (en) * | 2018-01-03 | 2021-08-20 | 北京大学 | Frequency division positioning method applied to multi-sound-source environment |
| CN109067478A (en) * | 2018-08-10 | 2018-12-21 | 北京历正科技有限责任公司 | A kind of unmanned plane detection method, device and equipment |
| CN110213010B (en) * | 2019-04-28 | 2020-07-07 | 浙江大学 | Unmanned aerial vehicle detection system and method based on multi-channel radio frequency signals |
| CN111650556B (en) * | 2020-06-15 | 2023-09-01 | 中国人民解放军国防科技大学 | Broadband radiation source parameter estimation method |
| CN111968671B (en) * | 2020-08-24 | 2024-03-01 | 中国电子科技集团公司第三研究所 | Low-altitude sound target comprehensive identification method and device based on multidimensional feature space |
| CN113225441B (en) * | 2021-07-09 | 2021-10-08 | 北京中电慧声科技有限公司 | Conference telephone system |
| CN115061087A (en) * | 2022-05-27 | 2022-09-16 | 上海事凡物联网科技有限公司 | Signal processing method, DOA estimation method and electronic equipment |
| CN114743562B (en) * | 2022-06-09 | 2022-11-01 | 成都凯天电子股份有限公司 | Method and system for recognizing airplane voiceprint, electronic equipment and storage medium |
| CN116224311B (en) * | 2023-02-10 | 2024-02-02 | 广东海洋大学 | Method for identifying underwater multi-target azimuth by utilizing incoherent subspace method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101320091A (en) * | 2008-07-15 | 2008-12-10 | 浙江大学 | Sound source terminal delay time eliminated frequency detection wireless sensor network ranging method |
| CN102013911A (en) * | 2010-12-02 | 2011-04-13 | 哈尔滨工程大学 | Broadband signal direction of arrival (DOA) estimation method based on threshold detection |
| CN102938897A (en) * | 2012-11-14 | 2013-02-20 | 浙江大学 | Wireless charging node capable of monitoring signal strength in real time |
| CN105160934A (en) * | 2015-10-14 | 2015-12-16 | 浙江大学 | Intelligent parking system suitable for large-scale parking lot |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9689965B2 (en) * | 2011-09-23 | 2017-06-27 | Bitwave Pte Ltd. | Hostile fire detection for an airborne platform |
-
2017
- 2017-01-12 CN CN201710021562.8A patent/CN106772227B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101320091A (en) * | 2008-07-15 | 2008-12-10 | 浙江大学 | Sound source terminal delay time eliminated frequency detection wireless sensor network ranging method |
| CN102013911A (en) * | 2010-12-02 | 2011-04-13 | 哈尔滨工程大学 | Broadband signal direction of arrival (DOA) estimation method based on threshold detection |
| CN102938897A (en) * | 2012-11-14 | 2013-02-20 | 浙江大学 | Wireless charging node capable of monitoring signal strength in real time |
| CN105160934A (en) * | 2015-10-14 | 2015-12-16 | 浙江大学 | Intelligent parking system suitable for large-scale parking lot |
Non-Patent Citations (1)
| Title |
|---|
| 《无人植保机施药雾滴空间质量平衡测试方法》;王昌陵 等;《农业工程学报》;20160731;第32卷(第11期);第54-61页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106772227A (en) | 2017-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106772227B (en) | A kind of unmanned plane direction determining method based on the identification of vocal print multiple-harmonic | |
| Dougherty | Functional beamforming | |
| US9420368B2 (en) | Time-frequency directional processing of audio signals | |
| CN108375763B (en) | Frequency division positioning method applied to multi-sound-source environment | |
| EP2530484B1 (en) | Sound source localization apparatus and method | |
| CN106504763A (en) | Based on blind source separating and the microphone array multiple target sound enhancement method of spectrum-subtraction | |
| CN111123192B (en) | Two-dimensional DOA positioning method based on circular array and virtual extension | |
| US20060204019A1 (en) | Acoustic signal processing apparatus, acoustic signal processing method, acoustic signal processing program, and computer-readable recording medium recording acoustic signal processing program | |
| CN105589066B (en) | A kind of method that underwater uniform motion ROV parameter is estimated using vertical vector battle array | |
| CN105580074B (en) | Signal processing system and method | |
| Li et al. | Estimation of relative transfer function in the presence of stationary noise based on segmental power spectral density matrix subtraction | |
| CN109324322A (en) | A kind of direction finding and target identification method based on passive phased array antenna | |
| CN108089146B (en) | High-resolution broadband direction-of-arrival estimation method for pre-estimated angle error robustness | |
| Orovi et al. | Time-frequency analysis and singular value decomposition applied to the highly multicomponent musical signals | |
| CN105158751A (en) | Acoustic vector array fast DOA (Direction of Arrival) estimation method | |
| Huang et al. | A practical fundamental frequency extraction algorithm for motion parameters estimation of moving targets | |
| Battista et al. | Inverse methods for three-dimensional acoustic mapping with a single planar array | |
| CN107564530A (en) | A kind of unmanned plane detection method based on vocal print energy feature | |
| Chouragade et al. | Continuous mapping of broadband VHF lightning sources by real-valued MUSIC | |
| Bahr et al. | Acoustic data processing and transient signal analysis for the hybrid wing body 14-by 22-foot subsonic wind tunnel test | |
| CN107390198B (en) | Sub-band related registration method under high-speed moving target | |
| KR102235982B1 (en) | Apparatus and method for estimating a biological heart rate parameter for a human fmcw radar | |
| CN109658944B (en) | Helicopter acoustic signal enhancement method and device | |
| JP2001337148A (en) | Extrapolation device for electromagnetic wave arrival direction | |
| CN113109760B (en) | Multi-line spectrum combined DOA estimation and clustering method and system based on group sparsity |
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 |