CN106155983A - A kind of Correlation Identification detection system based on FFT and method thereof - Google Patents

Abstract

The present invention relates to a kind of Correlation Identification detection system based on FFT, including: the first m-sequence generation module, the first the earth system, very first time synchronization module, the first signal receiving module, a FFT auto-correlation module, the first frequency domain modular converter, a FFT cross-correlation module, the second frequency domain modular converter and the first system function solve module；Described very first time synchronization module and the first m-sequence generation module and the first signal receiving module are connected, and the first m-sequence generation module and a FFT auto-correlation module are connected, and a FFT auto-correlation module is connected with the first frequency domain modular converter；First signal receiving module and a FFT cross-correlation module are connected, and a FFT cross-correlation module is connected with the second frequency domain modular converter；First frequency domain modular converter and the second frequency domain modular converter solve module with the first system function respectively and are connected.The present invention have identification efficiency high, in high precision, characteristic that noiseproof feature is high.

Description

A kind of Correlation Identification detection system based on FFT and method thereof

Technical field

The present invention relates to a kind of Correlation Identification detection system based on FFT and method thereof.

Background technology

During the system function of detection system, especially when the bigger system of identification (the biggest ground system), often Need to gather substantial amounts of data and carry out computing.Use different schemes just to have different operands, generally calculate relevant distinguishing Know relevant time use auto-correlation and cross-correlation function definition calculate, use summation to calculate.Cumulative During calculating, owing to needs are constantly kept in, are multiplied and are added, computational efficiency is substantially reduced.Calculating relevant when, There is one method fast and effectively, it is simply that by correlation theorem, utilize FFT relevant to calculate.First input and output signal Doing fast Fourier transform, then input signal is divided into two groups, the input signal after one of which conversion takes conjugation, another group Retain transformation results.Obtain the result of circular correlation according to correlation theorem, last another mistake switches back to time domain, obtains the knot being correlated with Really.Outwardly, it is transformed into frequency domain and converts back again the many steps of time domain, but the high efficiency calculated due to FFT so that meter The efficiency calculated is greatly improved.

At present, Correlation Identification technology has been applied to the system function identification field of the earth.Press down including by Correlation Identification Noise processed, thus the method reaching the purpose of identification system function.But, be applied to the earth identification when, often because of Make this system seem to waste time and energy for data volume is excessive.One of them important reason is just because calculating relevant mistake Cheng Zhong, uses cumulative definition and calculates.And inventor discovery FFT can be greatly improved the effect of calculating when calculating relevant Rate, thus get down to the method with within the system.Practice and emulation prove, the method can ensure that and former result phase really On the premise of Tong, improve the efficiency calculated.

Summary of the invention

The technology of the present invention solves problem: solves the inefficient problem of the Correlation Identification under big data quantity, remains phase Close the feature of the original effective suppression noise of identification.

The technical solution of the present invention is: a kind of Correlation Identification detection system based on FFT, including: the first m-sequence is raw Become module, first the earth system, very first time synchronization module, the first signal receiving module, a FFT-auto-correlation module, first Frequency domain modular converter, a FFT-cross-correlation module, the second frequency domain modular converter and the first system function solve module, Qi Zhong One FFT-auto-correlation module, the first frequency domain modular converter, a FFT-cross-correlation module, the first frequency domain modular converter and first are System function solves module and belongs to data processing section；Described very first time synchronization module and the first m-sequence generation module and first Signal receiving module is connected, and the first m-sequence generation module is connected with a FFT-auto-correlation module again, a FFT-auto-correlation mould Block and the first frequency domain modular converter are connected；First signal receiving module and the first cross-correlation module are connected, a FFT-cross-correlation Module is connected with the second frequency domain modular converter；First frequency domain modular converter and the second frequency domain modular converter solve mould with system function Block is connected；

The function of described first m-sequence generation module is to be sent to the earth system according to after setup parameter generation m-sequence；First First very first time synchronization module arranges predetermined instant makes the first m-sequence generation module generate the m of relevant parameter in the moment set Sequence, after generating m-sequence, m-sequence is just sent to the earth system by electrode.So, just complete the generation of m-sequence and send out Send；The voltage data bag x of transmission can be collected simultaneously；

The function of described first signal receiving module has been the reception through current signal too greatly.According to the very first time In the reservation moment that synchronization module is arranged, the first signal receiving module receives through voltage signal too greatly, thus forms packet y；

A described FFT-auto-correlation module includes that a FFT frequency domain converting unit, the first multiplication unit, an IFFT become Change unit and the first data cutout unit.Wherein a FFT frequency domain converting unit is connected with the first multiplication unit, the first multiplication list Unit is connected with an IFFT converter unit, and an IFFT unit and the first data cutout unit are connected.First m-sequence generation module After being connected with a FFT frequency domain converting unit, through a FFT-auto-correlation module, calculate auto-correlation by the method for frequency domain Autocorrelative frequency-domain result, intercept out therein for lap, obtained time domain autocorrelation result accurately；

Described first frequency domain modular converter includes FFT operator unit and a frequency domain conversion results.Oneth FFT operator unit It is connected with the first data cutout unit and the first auto-correlation frequency domain unit respectively.First data cutout unit sends first to The first auto-correlation frequency domain unit is obtained after FFT operator unit.So, autocorrelative frequency-domain result has just been obtained.

A described FFT-cross-correlation module includes that the 2nd FFT frequency domain converting unit, the second multiplication unit, the 2nd IFFT become Change unit and the second data cutout unit.2nd FFT frequency domain converting unit is connected with the second multiplication unit, the second multiplication unit with 2nd IFFT converter unit, the 2nd IFFT converter unit and the second data cutout unit are connected.Second frequency-domain transform unit simultaneously Being connected with the second multiplication unit, the first signal receiving module and the 2nd FFT frequency-domain transform unit are connected.Secondary signal receiver module The reception packet y obtained obtains the result of frequency domain through the 2nd FFT frequency domain converting unit, is then transferred to the second multiplication list Unit, the second frequency-domain transform unit is also transferred to the second multiplication unit result simultaneously, and both are exactly the frequency domain of cross-correlation after being multiplied As a result, it is then passed through the 2nd IFFT converter unit and obtains the transformation results of time domain, finally intercept out with the second data cutout unit The underlapped cross correlation results accurately of time domain.

Described second frequency domain modular converter includes the 2nd FFT operator unit and the first cross-correlation frequency domain unit.2nd FFT calculates Subelement is connected with the 2nd FFT frequency domain converting unit and the first cross-correlation frequency domain unit respectively.2nd FFT frequency domain converting unit warp Crossing the 2nd FFT operator unit and obtain the first cross-correlation frequency domain unit, this has just been calculated the frequency-domain result of cross-correlation.

System function solves module and includes the first divider and the first the earth frequency domain identification unit.First divider is divided It is not connected with the first the earth frequency domain identification unit with the first cross-correlation frequency domain unit, the first auto-correlation frequency domain unit.First is mutual Close frequency domain unit and the first auto-correlation frequency domain unit through the first divider, the first cross-correlation frequency domain unit divided by first from phase After closing frequency domain unit, obtaining the first the earth frequency domain identification unit, result is ultimately stored on the first the earth frequency domain identification unit.

The relevant treatment work of the Correlation Identification of detection system is made by what FFT operator completed at frequency domain, operation efficiency pole Height, is particularly suitable for when the data volume of correlation computations is bigger using.A described FFT-auto-correlation module, a FFT-are mutual It is as follows that pass module implements process solving the system function of detection system when: first, transmitting terminal collects packet x After receiving terminal all collects packet y simultaneously, data are given respectively FFT-auto-correlation module and a FFT-cross-correlation Module, after the FFT frequency-domain transform unit in F the oneth FT-auto-correlation module obtains result X (w) of frequency domain, transmitting terminal Packet x consign to the first multiplication unit, at this moment, multiplication unit can carry out conjugation process to one of which data, then borrows Correlation theorem is helped to calculate the autocorrelative frequency-domain result of circulation.Computing formula is: For following The autocorrelative result of ring.By correlation theorem it is recognised that circulation auto-correlation and the equal condition of cross-correlation areWherein M refer to M dot cycle be correlated with, L_xRefer to counting of x；In like manner, circulation The equal condition of cross-correlation and cross-correlation isWherein M is M dot cycle cross-correlation, L_y For counting of packet y.According to this condition, data cutout unit intercepts non-overlapping correlated results, namely The result of linear correlation.Thus auto-correlation and cross-correlation are calculated with FFT, then according to the frequency domain identification of Correlation Identification is public Formula calculates the system function of the earth.

Very first time synchronization module default setting in the case of no modification is: 12:01, i.e. predetermined instant 12:01 send out Sending end and receiving terminal work simultaneously.

Present invention advantage compared with prior art is:

(1) way that this computational efficiency calculating relevant Correlation Identification based on FFT adds up apparently higher than traditional use Calculating relevant, big data quantity when, operational efficiency is the highest, can be greatly saved the identification time.

(2), the when of identification system, m-sequence suppression noise can be combined by Correlation Identification.

(3) present invention can demonstrate the system function treating examining system at final display module.

Accompanying drawing explanation

Fig. 1 is the composition frame chart of a kind of Correlation Identification detection system based on FFT of the present invention；

Fig. 2 is the method flow diagram being embodied as example of the present invention.

Detailed description of the invention

As it is shown in figure 1, a kind of Correlation Identification detection system based on FFT of the present invention includes: the first m-sequence generation module 1, First the earth system 2, very first time synchronization module the 3, first signal receiving module the 4, the oneth FFT-auto-correlation module the 5, first frequency Territory modular converter the 6, the oneth FFT-cross-correlation module the 7, second frequency domain modular converter 8 and the first system function solve module 9, wherein Oneth FFT-auto-correlation module the 5, first frequency domain modular converter the 6, the oneth FFT-cross-correlation module the 7, second frequency domain modular converter 8 and The first system function solves module 9 and belongs to data processing section；Described very first time synchronization module 3 and the first m-sequence generate mould Block 1 is connected with the first signal receiving module 2, and the first signal generator module 3 is connected with a FFT-auto-correlation module 5 again, and first FFT-auto-correlation module 5 is connected with the first frequency domain modular converter 6；First signal receiving module 4 is connected with the first cross-correlation module 7, a FFT-cross-correlation module 7 is connected with the second frequency domain modular converter 8；First frequency domain modular converter 6 and the conversion of the second frequency domain Module 8 solves module 9 with the first system function and is connected；

The function of described first m-sequence generation module 1 is according to being sent to the first the earth system after setup parameter generation m-sequence System 2；First very first time synchronization module 3 arrange predetermined instant make the first m-sequence signal generating unit 13 set moment generate phase Answering the m-sequence of parameter, after generating m-sequence, m-sequence is just sent to the first the earth system 2 by electrode.So, m sequence is just completed The generation of row and transmission；The voltage data bag x of transmission can be collected simultaneously；

The function of described signal receiving module 4 has been the reception through current signal too greatly.According to time synchronized mould The predetermined instant that block 3 is arranged, signal receiving module 4 receives through voltage signal too greatly, thus forms packet y；

A described FFT-auto-correlation module 5 includes FFT frequency domain converting unit 51, first multiplication unit 52, first IFFT converter unit 53 and the first data cutout unit 54.Wherein FFT frequency domain converting unit 51 and first multiplication unit 52 Being connected, the first multiplication unit 52 is connected with an IFFT converter unit 53, IFFT unit 53 and a first data cutout unit 54 are connected.After first m-sequence generation module 1 is connected with a FFT frequency domain converting unit 51, through a FFT-auto-correlation module 5, calculate autocorrelative frequency-domain result by the method for frequency domain, intercepted out underlapped part therein, obtain Time domain autocorrelation result accurately；

Described first frequency domain modular converter 6 includes FFT operator unit the 1st and frequency domain conversion results the 1st.First FFT operator unit 61 is connected with the first data cutout unit 54 and the first auto-correlation frequency domain unit 62 respectively.First data are cut Take and obtain the first auto-correlation frequency domain unit 62 after unit 54 sends a FFT operator unit 61 to so, just obtained auto-correlation Frequency-domain result.

A described FFT-cross-correlation module 7 includes the 2nd FFT frequency domain converting unit the 71, second multiplication unit 72, second IFFT converter unit 73 and the second data cutout unit 74.2nd FFT frequency domain converting unit 71 is connected with the second multiplication unit 72, Second multiplication unit 72 and the 2nd IFFT converter unit 73, the 2nd IFFT converter unit 73 and the second data cutout unit 74.With Time the first frequency-domain transform unit 51 be connected with the second multiplication unit 72, the first signal receiving module 4 and the 2nd FFT frequency domain transform list Unit 71 is connected.The reception packet y that first signal receiving module 4 obtains obtains frequency domain through the 2nd FFT frequency domain converting unit 71 Result, be then transferred to the second multiplication unit 72, the first frequency-domain transform unit 51 is also transferred to square law list result simultaneously Unit 72, both are exactly the frequency-domain result of cross-correlation after being multiplied, and are then passed through the 2nd IFFT converter unit 73 and obtain the conversion of time domain As a result, the underlapped cross correlation results accurately of time domain is finally intercepted out with the second data cutout unit 74.

Described second frequency domain modular converter 8 includes the 2nd FFT operator unit 81 and the first cross-correlation frequency domain unit 82.Second FFT operator unit 81 is connected with the 2nd FFT frequency domain converting unit 71 and the first cross-correlation frequency domain unit 82 respectively.2nd FFT frequency Territory converting unit 71 obtains the first cross-correlation frequency domain unit 82 through the 2nd FFT operator unit 81, and this has just been calculated mutually The frequency-domain result closed.

The first system function solves module 9 and includes the first divider 91 and the first the earth frequency domain identification unit 92.First Divider 91 respectively with the first cross-correlation frequency domain unit the 82, first auto-correlation frequency domain unit 62 and the first the earth frequency domain identification list Unit 92 is connected.First cross-correlation frequency domain unit 82 and the first auto-correlation frequency domain unit 62 are through the first divider 91, and first is mutual After relevant frequency domain unit 82 is divided by the first auto-correlation frequency domain unit 62, obtaining the first the earth frequency domain identification unit 92, result is final It is stored in the first the earth frequency domain identification unit 92.

The relevant treatment work of the Correlation Identification of detection system is made by what FFT operator completed at frequency domain, operation efficiency pole Height, is particularly suitable for when the data volume of correlation computations is bigger using.Described FFT-auto-correlation module the 5, the oneth FFT-is mutual It is as follows that correlation module 7 implements process solving the system function of detection system when: first, and transmitting terminal collects data After receiving terminal all collects packet y while of bag x, data are given respectively a FFT-auto-correlation module 5 and a FFT-mutual Correlation module 7, after the FFT frequency-domain transform unit 51 in a FFT-auto-correlation module 5 obtains result X (w) of frequency domain, The packet x of transmitting terminal is consigned to the first multiplication unit 52, and at this moment, one of which data can be carried out by the first multiplication unit 52 Conjugation processes, and then calculates the autocorrelative frequency-domain result of circulation by correlation theorem.Computing formula is: For circulating autocorrelative result.By correlation theorem it is recognised that circulate auto-correlation The condition equal with cross-correlation isWherein M refer to M dot cycle be correlated with, L_xRefer to It is counting of x；In like manner, the equal condition of circulation cross-correlation and cross-correlation isWherein M is M dot cycle cross-correlation, L_yFor counting of packet y.According to this condition, intercept out non-overlapping in data cutout unit Correlated results come.Thus calculate auto-correlation and cross-correlation with FFT, then according to the frequency domain identification formula of Correlation Identification Calculate the system function of the earth.

Very first time synchronization module 3 default setting in the case of no modification is: 12:01, i.e. predetermined instant 12:01 send out Sending end and receiving terminal work simultaneously.

According to another embodiment of the invention, described very first time synchronization module 3 is used for arranging predetermined instant so that the One m-sequence generation module 1 generates m-sequence in the moment set；And make the first signal receiving module 4 receive warp at predetermined instant Cross the signal of telecommunication of the first the earth system 2；

Described first m-sequence generation module 1 is for generating m-sequence at predetermined instant and being sent to the first the earth via electrode System 2；And gather the voltage data bag x of transmission；

Described first signal receiving module 4, for the predetermined instant arranged according to very first time synchronization module 3, receives and passes through The voltage signal of the first the earth system 2, thus obtain voltage data bag y；

A described FFT-auto-correlation module 5 includes FFT frequency-domain transform unit 51, first multiplication unit 52, first IFFT converter unit 53 and the first data cutout unit 54；

Wherein a FFT frequency domain converting unit 51 is connected with the first multiplication unit 52, the first multiplication unit 52 and first IFFT converter unit 53 is connected, and an IFFT unit 53 is connected with the first data cutout unit 54；

A described FFT frequency domain converting unit 51 carries out FFT for voltage data bag x, is converted to by time-domain signal Frequency-region signal；

Described first multiplication unit 52, for carrying out conjugate operation by the frequency-region signal of voltage data bag x；And conjugation is transported Result after calculation is multiplied with the frequency-region signal of voltage data bag x, and computing formula is

$\stackrel{\‾}{R_{xx}}\left(w\right)=X\left(w\right)X^{*}\left(w\right)$

Wherein X (w) is the frequency domain of input signal, X^*W () is the conjugation of the frequency-domain result of input signal；

A described IFFT converter unit 53, for the signal that the first multiplication unit 52 exports is carried out IFFT conversion, To time-domain signal, transformation for mula is

$\stackrel{\‾}{r_{xx}}\left(t\right)=IFFT\left(X{\left(w\right)}^{*}X\right(w\left)\right)$

For circulating autocorrelative result；

Described first data cutout unit 54, is used for intercepting a described IFFT converter unit 53 and circulates auto-correlation with linear The part that auto-correlation is equal, obtains the autocorrelative result of time domain linear；

Described first frequency domain modular converter 6 includes FFT operator unit 61 and a first auto-correlation frequency domain unit 62；Wherein A described FFT operator unit 61 is connected with the first data cutout unit 54 and the first auto-correlation frequency domain unit 62 respectively；

A described FFT operator unit 61 is for carrying out the time domain autocorrelation result that the first data cutout unit 54 exports FFT computing, thus obtain frequency domain autocorrelation result；And frequency domain autocorrelation result is sent to the first auto-correlation frequency domain unit guarantor Deposit；

A described FFT-cross-correlation module 7 includes the 2nd FFT frequency-domain transform unit the 71, second multiplication unit 72, second IFFT converter unit 73 and the second data cutout unit 74；

Described 2nd FFT frequency-domain transform unit 71 is connected with the second multiplication unit 72, described second multiplication unit 72 and Two IFFT converter units 73 are connected；Described 2nd IFFT converter unit 73 is connected with the second data cutout unit 74；

Described first frequency-domain transform unit 51 is connected with the second multiplication unit 72, the first signal receiving module 4 and the 2nd FFT Frequency-domain transform unit 71 is connected；

Described 2nd FFT frequency domain converting unit 71 is carried out for the voltage data bag y obtained by the first signal receiving module 4 FFT, is converted to frequency-region signal by time-domain signal；

Described second multiplication unit 72, the frequency domain letter of the voltage data bag x for the first frequency-domain transform unit 51 is exported Number it is multiplied with the frequency-region signal of voltage data bag y of the 2nd FFT frequency domain converting unit 71 output, obtains the frequency domain knot of cross-correlation Really；

Described 2nd IFFT converter unit 73, for the signal that the second multiplication unit 72 exports is carried out IFFT conversion, To time-domain signal；

Described second data cutout unit 74, in the time-domain signal intercepting described 2nd IFFT converter unit 73 output Underlapped part, obtains time domain cross correlation results；

Described second frequency domain modular converter 8 includes the 2nd FFT operator unit 81 and the first cross-correlation frequency domain unit 82；

Described 2nd FFT operator unit 81 respectively with the second data cutout unit 74 and the first cross-correlation frequency domain unit 82 phase Even；

Described 2nd FFT operator unit 81 is for carrying out the time domain cross correlation results that the second data cutout unit 74 exports FFT computing, obtains frequency domain cross correlation results；And frequency domain cross correlation results is sent to the first cross-correlation frequency domain unit 82 preserves；

The first system function solves module 9 and includes the first divider 91 and the first the earth frequency domain identification unit 92；Described First divider 91 respectively with the first cross-correlation frequency domain unit the 82, first auto-correlation frequency domain unit 62 and the first the earth frequency domain Identification unit 92 is connected；

Described first divider 91 for frequency domain cross correlation results that the first cross-correlation frequency domain unit 82 is exported divided by The frequency domain cross correlation results of the first auto-correlation frequency domain unit 62 output, obtains the first the earth frequency domain identification result, and is stored in the In one the earth frequency domain identification unit 92.

Below by embodiment Correlation Identification detection system of based on FFT a kind of to the present invention detection method further Describe in detail；

As shown in Figure 2:

Step 201: first, connects the battery of instrument, connects the big voltage electrical generation equipment outside transmitter and sender electrode, After inspection connects, the power on button of transmitter on (X-on)；

Step 202: after connecting modules start, peripheral apparatus to be done initial work by transmitter.Peripheral apparatus Including temperature sensor, AD1274, SD card, buzzer, PCF8563,24C02 and TFLCD display screen etc.；Initial work includes These peripheral hardwares doing the pretreatment operated, sets the communication protocol of peripheral hardware and equipment, inspection peripheral hardware is the most normal；

Step J01: when transmitter initializes, needs to judge whether modules initializes according to the practical situation of peripheral hardware Success, if it is, perform step 204；If it is not, then perform step 203；

Step 203: initialize peripheral hardware failure, then show the error message that this peripheral hardware is corresponding；

Step 204: after initializing peripheral hardware success, then showing the main interface of transmitter, main interface is broadly divided into three pieces, arranges Parameter (sample rate, chip width, m-sequence cycle, m-sequence exponent number, the magnitude of voltage of transmission, delivery time, transmission time interval), Transmitter icon shows, operates brief description, then performs step 205 after instrument stabilizer；

Step 205: the parameter of the m-sequence of transmission, sample rate, delivery time, transmission time interval (how long sending) are set, According to information after setting completed, step 206 is then performed；

Step 206:GPS module obtains the time of GPS, the present moment of the shortest more new instrument in real time；

Step J03: judge whether to arrive delivery time according to the time of real-time update.If it is, perform step 216, If it is not, then perform step 218；

Step 216: send m-sequence according to the parameter of the m-sequence arranged；

Step 218: execution etc. are to be operated, waiting time arrival delivery time；

Step 217: the big voltage signal of m-sequence sent according to the sample rate collection arranged, simultaneously as required collection Data be transferred to receiver；

Step J05: the time interval whether time that judges sets more than or equal to user, if it is not, then perform step 216； If it is, perform step 220；

Step 208: first, connects the battery of instrument and receives electrode, checks that after connecting, the start of transmitter on (X-on) is pressed Button；

Step 209: after connecting modules start, peripheral apparatus to be done initial work by receiver.Peripheral apparatus Including temperature sensor, AD1274, SD card, buzzer, PCF8563,24C02 and TFLCD display screen etc.；Initial work includes These peripheral hardwares doing the pretreatment operated, sets the communication protocol of peripheral hardware and equipment, inspection peripheral hardware is the most normal；

Step J02: judge that modules initializes the most successful, if it is, perform step 211；If it is not, then perform Step 210；

Step 210: certain peripheral hardware initializes unsuccessfully, then show the error message that peripheral hardware is corresponding；

Step 211: after initializing peripheral hardware success, then showing the main interface of receiver, main interface is broadly divided into three pieces, arranges Parameter (sample rate, the time of reception, reception time interval), receiver icon show, operate brief description, after instrument stabilizer then Perform step 212；

Step 212: the sample rate of receiver, the time of reception and reception time interval are set.Noting, these numerical value should be with The numerical value of transmitter is identical；

Step 213:GPS module obtains the time of GPS in real time, constantly updates the present moment of receiver；

Step J04: judge whether to arrive the time of reception, if it is, perform step 215；If it is not, then execution step 214；

Step 214: the present moment obtained according to GPS, waits and arrives the time of reception；

Step 215: gather voltage signal according to the sample rate of user setup；

Step J06: judge that whether acquisition time is more than or equal to receiving time interval.If it is not, then perform step 215；As Fruit is, then perform step 219；

Step 219: receive voltage signal and the voltage signal of receiver collection that transmitter transmits, calculate big The system function of ground system；

Described calculate the earth system function mode particularly as follows:

Relation according to input with output signal:

(1) u (t)=m (t) * h (t)+n (t)

Wherein, m (t) is the pseudorandom m signal of input the earth, has rhythmic pseudo-randomness；H (t) is to need to calculate The ssystem transfer function come, n (t) is random noise signal, and y (t) is the signal that instrument collects；

(2) after then FFT being at equation two ends, then it is multiplied by Y^*W () obtains

$\stackrel{\‾}{R_{xy}}\left(w\right)=Y\left(w\right)X^{*}\left(w\right)=X\left(w\right)X^{*}\left(w\right)H\left(w\right)+N\left(w\right)X^{*}\left(w\right)$

Wherein,For the frequency-domain result of input signal Yu the cyclic convolution of the cross-correlation function of output signal, X (w) For the frequency domain of input signal, X^*W () is the conjugation of the frequency-domain result of input signal, N (w) is the frequency-domain result of noise, owing to making an uproar Sound and input signal are uncorrelated, last N (w) X^*W () product is 0；

(3) after, equation both sides are carried out fast Fourier transform IFFT, obtain the Wiener Hopf equation of time domain

$\stackrel{\‾}{r_{xy}}\left(t\right)=\stackrel{\‾}{r_{xx}}\left(t\right)*h\left(t\right)$

WhereinFor the result of input signal Yu the circulation cross-correlation of output signal,Circulate from phase for input signal The result closed；

Then according to the condition that circular correlation and linear correlation are equal, circular correlation is converted into linear correlation, conversion stripes Part isWherein M is that M dot cycle is correlated with, L_yFor counting of packet y,Wherein M is that M dot cycle is correlated with, L_xFor counting of packet y, then to equation two End does FFT, is then calculated

$\left(4\right)---H\left(w\right)=\frac{FFT\left(r_{xy}\right(t))}{FFT\left(r_{xx}\right(t\left)\right)}$

Can be calculated by above formula and treat the system function of examining system and show.

Step 221: terminate identification work；

According to one embodiment of present invention, when utilizing detection system of the present invention, first select and treat examining system.Then arrange The time sent and the parameter sending signal.Treating that examining system is a second-order system, its system function is: sys_fun= 11.5e^-5tSin (8.6t), the parameter of m-sequence be amplitude be 1V, sample rate is 1000Hz, and exponent number is 12～14, and the cycle is 3, nothing Interpolation, noise is 30V random noise.The hardware of core calculations FFT realizes being mainly 32 by STM32F407ZGT6 single essences Degree hardware FPU unit (FPU Float Point Unit), when traditional fixed-point CPU calculates floating point unit, relatively time consuming, and for by For the CPU of FPU, the most only needing several instructions, budget speed is the fastest.The FPU used is relative to traditional CPU calculating side Formula, exceeds the decades of times even operational performance of hundreds of times.Meanwhile, in written in code control, compare traditional CPU also the most very Many.

The description of the present invention is not done prior art known to the technical staff of this detailed professional field.

Schemes described above is a kind of application scenario of this system, and this invention is mainly used in and is correlated with big data quantity The system identification of identification, is used in this invention other field or does corresponding modify and some improvement should also be as being considered as this invention Protection domain.

Claims (6)

1. a Correlation Identification detection system based on FFT, it is characterised in that including: the first m-sequence generation module (1), first Greatly system (2), very first time synchronization module (3), the first signal receiving module (4), FFT-auto-correlation module (5), One frequency domain modular converter (6), a FFT-cross-correlation module (7), the second frequency domain modular converter (8) and the first system function solve Module (9)；Described very first time synchronization module (3) and the first m-sequence generation module (1) and the first signal receiving module (4) phase Even, the first m-sequence generation module (1) is connected with FFT-auto-correlation module (5), FFT-auto-correlation module (5) and the One frequency domain modular converter (6) is connected；First signal receiving module (4) is connected (7) with a FFT-cross-correlation module, a FFT- Cross-correlation module (7) is connected with the second frequency domain modular converter (8)；First frequency domain modular converter (6) and the second frequency domain modular converter (8) solve module (9) respectively with the first system function to be connected.

A kind of Correlation Identification detection system based on FFT the most according to claim 1, it is characterised in that:

Described very first time synchronization module (3) is used for arranging predetermined instant so that the first m-sequence generation module (1) is in setting Moment generates m-sequence；And make the first signal receiving module (4) receive the electricity through the first the earth system (2) at predetermined instant Signal；

Described first m-sequence generation module (1) is for generating m-sequence at predetermined instant and being sent to the first the earth system via electrode System (2)；And gather the voltage data bag x of transmission；

Described first signal receiving module (4), for the predetermined instant arranged according to very first time synchronization module (3), receives and passes through The voltage signal of the first the earth system (2), thus obtain voltage data bag y；

Described FFT-auto-correlation module (5) include a FFT frequency-domain transform unit (51), the first multiplication unit (52), One IFFT converter unit (53) and the first data cutout unit (54)；

Wherein a FFT frequency domain converting unit (51) is connected with the first multiplication unit (52), the first multiplication unit (52) and first IFFT converter unit (53) is connected, and an IFFT unit (53) is connected with the first data cutout unit (54)；

A described FFT frequency domain converting unit (51) carries out FFT for voltage data bag x, and time-domain signal is converted to frequency Territory signal；

Described first multiplication unit (52), for carrying out conjugate operation by the frequency-region signal of voltage data bag x；And by conjugate operation After result be multiplied with the frequency-region signal of voltage data bag x, computing formula is

$\stackrel{\‾}{R_{xx}}\left(w\right)=X\left(w\right)X^{*}\left(w\right)$

Wherein X (w) is the frequency domain of input signal, X^*W () is the conjugation of the frequency-domain result of input signal；

A described IFFT converter unit (53), for the signal that the first multiplication unit (52) exports is carried out IFFT conversion, To time-domain signal, transformation for mula is:

$\stackrel{\‾}{r_{xx}}\left(t\right)=IFFT\left(X{\left(w\right)}^{*}X\right(w\left)\right)$

For circulating autocorrelative result；

Described first data cutout unit (54), is used for intercepting described IFFT converter unit (53) circulation auto-correlation with linear The part that auto-correlation is equal, obtains the autocorrelative result of time domain linear；

Described first frequency domain modular converter (6) includes FFT operator unit (61) and the first auto-correlation frequency domain unit (62)；Its Described in FFT operator unit (61) respectively with the first data cutout unit (54) and the first auto-correlation frequency domain unit (62) It is connected；

Described FFT operator unit (61) is for carrying out the time domain autocorrelation result that the first data cutout unit (54) exports FFT computing, thus obtain frequency domain autocorrelation result；And frequency domain autocorrelation result is sent to the first auto-correlation frequency domain unit (62) Preserve；

A described FFT-cross-correlation module (7) include the 2nd FFT frequency-domain transform unit (71), the second multiplication unit (72), Two IFFT converter units (73) and the second data cutout unit (74)；

Described 2nd FFT frequency-domain transform unit (71) is connected with the second multiplication unit (72), described second multiplication unit (72) with 2nd IFFT converter unit (73) is connected；Described 2nd IFFT converter unit (73) is connected with the second data cutout unit (74)；

Described first frequency-domain transform unit (51) is connected with the second multiplication unit (72), the first signal receiving module (4) and second FFT frequency-domain transform unit (71) is connected；

Described 2nd FFT frequency domain converting unit (71) is carried out for the voltage data bag y obtained by the first signal receiving module (4) FFT, is converted to frequency-region signal by time-domain signal；

Described second multiplication unit (72), the frequency domain letter of the voltage data bag x for the first frequency-domain transform unit (51) is exported The frequency-region signal of number voltage data bag y exported with the 2nd FFT frequency domain converting unit (71) is multiplied, and obtains the frequency domain knot of cross-correlation Really；

Described 2nd IFFT converter unit (73), for the signal that the second multiplication unit (72) exports is carried out IFFT conversion, To time-domain signal；

Described second data cutout unit (74), is used for intercepting in the time-domain signal that described 2nd IFFT converter unit (73) exports Underlapped part, obtains time domain cross correlation results；

Described second frequency domain modular converter (8) includes the 2nd FFT operator unit (81) and the first cross-correlation frequency domain unit (82)；

Described 2nd FFT operator unit (81) respectively with the 2nd F data cutout unit (74) and the first cross-correlation frequency domain unit (82) it is connected；

Described 2nd FFT operator unit (81) is for carrying out the time domain cross correlation results that the second data cutout unit (74) exports FFT computing, obtains frequency domain cross correlation results；And frequency domain cross correlation results is sent to the first cross-correlation frequency domain unit (82) guarantor Deposit；

The first system function solves module (9) and includes the first divider (91) and the first the earth frequency domain identification unit (92)；Institute State the first divider (91) respectively with the first cross-correlation frequency domain unit (82), the first auto-correlation frequency domain unit (62) and first Greatly frequency domain identification unit (92) is connected；

Described first divider (91) for frequency domain cross correlation results that the first cross-correlation frequency domain unit (82) is exported divided by The frequency domain cross correlation results that first auto-correlation frequency domain unit (62) exports, obtains the first the earth frequency domain identification result, and is stored in In first the earth frequency domain identification unit (92).

A kind of Correlation Identification detection system based on FFT the most according to claim 2, it is characterised in that:

Very first time synchronization module (3) generates m-sequence according to parameter preset.

4. according to arbitrary described a kind of based on FFT the Correlation Identification detection system of claim 1-3, it is characterised in that:

Very first time synchronization module (3) default setting in the case of no modification is: 12:01, i.e. sends out at predetermined instant 12:01 Sending end and receiving terminal work simultaneously.

5., according to the detection method of the arbitrary described Correlation Identification detection system based on FFT of claim 1-4, its feature exists In:

Step 201: power on；

Step 202: peripheral apparatus initial work；Described peripheral apparatus include temperature sensor, AD1274, SD card, buzzer, PCF8563,24C02 and TFLCD display screen；Described initial work includes these peripheral hardwares do the pretreatment operated, and sets Peripheral hardware and the communication protocol of equipment, inspection peripheral hardware is the most normal；

Step J01: judge whether modules initializes successfully, if it is, perform step 204；If it is not, then execution step 203；

Step 203: initialize peripheral apparatus failure, then show the error message that this peripheral apparatus is corresponding；

Step 204: after initializing peripheral hardware success, then show main interface, then perform step 205；

Step 205: arrange the parameter of the m-sequence of transmission, sample rate, delivery time, transmission time interval, performs step 206；

Step 206: obtain gps time, carry out time synchronized,；

Step J03: judge whether to arrive delivery time according to the time obtained in real time.If it is, perform step 216, if No, then perform step 218；

Step 216: send m-sequence according to the parameter of the m-sequence arranged, perform step 217；

Step 218: execution etc. are to be operated, waiting time arrival delivery time；

Step 217: according to the signal of the m-sequence that the sample rate collection arranged sends, the most as required the data gathered are passed It is defeated by receiving terminal；

Step J05: the time interval whether time that judges sets more than or equal to user, if it is not, then perform step 216；If It is then to perform step 220；

Step 220: stop sending m-sequence；

Step 208: power on；

Step 209: peripheral apparatus initial work；Described peripheral apparatus include temperature sensor, AD1274, SD card, buzzer, PCF8563,24C02 and TFLCD display screen；Initial work includes these peripheral hardwares do the pretreatment operated, and sets peripheral hardware With the communication protocol of equipment, inspection peripheral hardware is the most normal；

Step J02: judge that modules initializes the most successful, if it is, perform step 211；If it is not, then execution step 210；

Step 210: certain peripheral hardware initializes unsuccessfully, then show the error message that peripheral hardware is corresponding；

Step 211: after initializing peripheral hardware success, then show the main interface of receiver, then perform step 212 after instrument stabilizer；

Step 212: the sample rate identical with the first signal generator module, the time of reception and reception time interval are set；

Step 213:GPS module obtains the time of GPS in real time；

Step J04: judge whether to arrive the time of reception according to the time obtained in real time, if it is, perform step 215；If No, then perform step 214；

Step 214: the present moment obtained according to GPS, waits and arrives the time of reception；

Step 215: gather voltage signal according to the sample rate arranged；

Step J06: judge that whether acquisition time is more than or equal to receiving time interval；If it is not, then perform step 215；If it is, Then perform step 219；

Step 219: according to the transmission voltage data bag x gathered and the signal of telecommunication of the reception voltage data bag y of collection, calculate The greatly system function of system showing；

Step 221: terminate identification.

The detection method of Correlation Identification detection system based on FFT the most according to claim 5, it is characterised in that:

Described calculate the earth system function mode particularly as follows:

Relation according to input with output signal:

(1) y (t)=x (t) * h (t)+n (t)

Wherein, x (t) is the pseudorandom m signal of input the earth, has rhythmic pseudo-randomness；H (t) needs to calculate Ssystem transfer function, n (t) is random noise signal, and y (t) is the signal that instrument collects；

(2) after then FFT being at equation two ends, then it is multiplied by Y^*W () obtains

$\stackrel{\‾}{R_{xy}}\left(w\right)=Y\left(w\right)X^{*}\left(w\right)=X\left(w\right)X^{*}\left(w\right)H\left(w\right)+N\left(w\right)X^{*}\left(w\right)$

Wherein,For the frequency-domain result of input signal Yu the cyclic convolution of the cross-correlation function of output signal, X (w) is input The frequency domain of signal, X^*W () is the conjugation of the frequency-domain result of input signal, N (w) is the frequency-domain result of noise, due to noise and defeated Enter signal uncorrelated, last N (w) X^*W () product is 0；

(3) after, equation both sides are carried out fast Fourier transform IFFT, obtain the Wiener Hopf equation of time domain

$\stackrel{\‾}{r_{xy}}\left(t\right)=\stackrel{\‾}{r_{xx}}\left(t\right)*h\left(t\right)$

WhereinFor the result of input signal Yu the circulation cross-correlation of output signal,Autocorrelative for input signal circulation Result；

Then according to the condition that circular correlation and linear correlation are equal, circular correlation being converted into linear correlation, switch condition is

Wherein M is that M dot cycle is correlated with, L_yFor counting of packet y,

Wherein M is that M dot cycle is correlated with, L_xFor counting of packet y,

Then FFT is done at equation two ends, be then calculated

$\left(4\right)---H\left(w\right)=\frac{FFT\left(r_{xy}\right(t\left)\right)}{FFT\left(r_{xx}\right(t\left)\right)}$

The system function treating examining system can be calculated by above formula.