CN103018598A - Method for improving radiating electromagnetic interference mixed signal blind source separation on basis of signal difference - Google Patents

Abstract

The invention provides a method for improving blind source separation of radiated electromagnetic interference mixed signals based on signal difference, which is used in the field of electromagnetic compatibility testing. The method of the present invention tests the background noise separately at each test position, and then conducts a radiation emission test on the tested product at each test position, and processes the two test signals through an adaptive filter to obtain a mixture of electromagnetic interference emitted by the tested product itself. signal, and then perform difference processing on the signal matrix composed of the obtained electromagnetic interference mixed signal to obtain a new signal matrix, and use the independent component analysis method to perform blind source separation on the obtained new signal matrix. The invention improves the effect of blind source separation of radiated electromagnetic interference signals, improves the separation accuracy, and improves the separation speed of blind source separation of electromagnetic interference signals.

Description

A Method for Improving Blind Source Separation of Radiated Electromagnetic Interference Mixed Signals Based on Signal Difference

technical field

The invention relates to the field of electromagnetic compatibility testing, in particular to a method for improving blind source separation of radiated electromagnetic interference mixed signals based on signal difference.

Background technique

According to the provisions of GJB151A-97 "Electromagnetic Emission and Sensitivity Requirements for Military Equipment and Subsystems", electromagnetic compatibility testing is divided into two parts: electromagnetic interference (EMI) testing and electromagnetic susceptibility (EMS) testing, of which electromagnetic interference testing is divided into conduction There are two parts of emission test and radiation emission test. Among them, the conduction emission test examines the interference signals that exist on the AC and DC power lines and are generated by the tested product, while the radiated emission test examines the signal emitted by the device under test through space. Radiation emission test refers to the intentional or unintentional electromagnetic energy radiated by a certain component, antenna, cable or connector in the tested object and propagated through space.

As shown in Figure 1, the existing electromagnetic radiation emission test system generally consists of a receiving antenna, a measuring receiver and a data recording device. The receiving antenna in the system is used to receive the electromagnetic radiation interference voltage. On the one hand, the measuring receiver is used to receive the electromagnetic radiation interference voltage received by the antenna. On the other hand, the received electromagnetic radiation interference voltage is corrected by the antenna coefficient to the corresponding electric field strength output. To the data recording device, the data recording device is used to record the data sequence.

For radiated electromagnetic interference sources with unknown signal types, unknown numbers, and unknown mixing methods, the existing radiation tests are facing great difficulties. The blind source separation theory has the characteristics that the source signal can be separated only from the mixed test signal without knowing the source signal and the source signal mixing method, which is very suitable for the electromagnetic interference test problem of this type of source signal. Therefore, it is very appropriate to choose the theory of blind source separation to solve the actual problem of radiated electromagnetic interference test. In the blind source separation theory, the independent component analysis method is the most used method.

When the measuring receiver is used to collect the mixed electromagnetic interference signal, the time information of the mixed electromagnetic interference signal cannot be collected, which makes the blind source separation effect of the existing test method on the mixed electromagnetic interference signal limited, and the accuracy of the electromagnetic interference test is not ideal.

Contents of the invention

The present invention aims at collecting the time information of the electromagnetic interference mixed signal when the measurement receiver is used to collect the electromagnetic interference mixed signal in the current electromagnetic radiation emission test, so that the effect of blind source separation on the electromagnetic interference mixed signal is limited, and the accuracy of the electromagnetic interference test is not good. High problem, providing a method based on signal difference to improve blind source separation of radiated electromagnetic interference mixed signals.

A method for improving blind source separation of radiated electromagnetic interference mixed signals based on signal difference provided by the present invention comprises the following steps:

Step 1: Set M test positions in the test space for placing receiving antennas to receive the electromagnetic interference mixed signals radiated in the space; the distance between the antenna position and the tested product is set according to the corresponding standard requirements;

a_i为被试品本身发射出的电磁干扰混合信号与背景噪声信号的混合信号；Step 2: First, test the background noise at each test position, and the background noise signal tested at the i-th test position is Then, the radiated emission test is performed on the tested product at each test position, and the mixed signal tested at the i-th test position is

a _i is the mixed signal of the electromagnetic interference mixed signal emitted by the tested product itself and the background noise signal;

Step 3: Input the adaptive filter with a _i (i=1,2,...,M) and ni ( _i =1,2,...,M) as references, and use the background noise signal in the mixed signal and the separate The background noise signal obtained by the test has correlation, and the background noise signal in the mixed signal is filtered out to obtain the electromagnetic interference mixed signal x _i (i=1,2,...,M) emitted by the tested product itself;

Step 4: combine the electromagnetic interference mixed signals of each test position after filtering the background noise signal into a joint matrix X=[x ₁ ,x ₂ ...x _M ] ^T ;

对联合矩阵X进行差值处理，得到新矩阵Z=BX。Step 5: Construct the signal difference matrix

Perform difference processing on the joint matrix X to obtain a new matrix Z=BX.

Step 6: Taking Z as the electromagnetic interference mixed-signal matrix, use the independent component analysis method to perform blind source separation.

Advantage and positive effect of the present invention are:

(1) Without changing the nature of the test signal, mining information between each group of test signals to improve the separation effect;

(2) Weaken the weak information in the signal, highlight the sharply changed information in the signal, and improve the separation accuracy of blind source separation of electromagnetic interference mixed signals;

(3) The separation speed of blind source separation of electromagnetic interference mixed signals is improved.

Description of drawings

Fig. 1 is the structural representation of conventional electromagnetic radiation emission testing system;

Fig. 2 is three radiation source signal waveform diagrams in the embodiment of the present invention;

Fig. 3 is a schematic diagram of the actual test site layout in the embodiment of the present invention;

Fig. 4 is the mixed signal waveform diagram of the next 5 positions actually measured in the embodiment of the present invention;

Fig. 5 is the signal waveform diagram after the background noise is filtered out at five positions actually measured in the embodiment of the present invention;

Fig. 6 is a signal waveform diagram after the actual measurement of the next five position differences in the embodiment of the present invention;

Fig. 7 is a signal waveform diagram of direct processing and difference post-processing under actual measurement in an embodiment of the present invention;

Fig. 8 is a flow chart of the method for improving blind source separation of radiated electromagnetic interference mixed signals according to the present invention.

Detailed ways

The method of the present invention will be described in detail below in conjunction with the accompanying drawings.

The method of the present invention preprocesses the collected multi-position electromagnetic interference mixed signal, so that the effect of blind source separation for the spectrum signal is improved, the accuracy of the electromagnetic interference test is improved, and the further investigation and rectification of the electromagnetic interference exceeding the standard equipment is provided. A more solid foundation has been laid. In the electromagnetic compatibility radiation emission test, when the interference signal is blindly separated, the mixed interference signal is preprocessed to improve the overall separation effect. After the multi-position electromagnetic mixed interference signal is obtained, the background noise is filtered out, and then according to Based on the principle that the row transformation of the mixed signal matrix does not change the nature of the signal, the row transformation of the mixed interference signal matrix is performed to realize the difference processing of the signals at each position, thereby improving the separation effect of the blind source separation of the electromagnetic interference mixed signal.

A method for improving blind source separation of radiated electromagnetic interference mixed signals based on signal difference provided by the present invention, as shown in Figure 8, has the following steps:

Step 1: Place receiving antennas at M points in the test space to receive the electromagnetic interference mixed signal radiated in the space. The distance between the antenna position and the tested product is set according to the requirements of the corresponding standards.

The arrangement of the test object, antenna, measuring receiver and data receiving device is shown in Figure 1. In the actual test, when the test position is determined in advance, a pair of antennas can be used to test in a certain order at different positions.

再对被试品进行辐射发射测试，在第i个测试位置测试的混合信号记为a_i(i＝1,2,…,M)，测试信号a_i为被试品本身发射出的电磁干扰混合信号与背景噪声的混合信号。Step 2: First test the background noise at each test position, and record the background noise at the i-th test position as

Then conduct radiation emission test on the tested product, the mixed signal tested at the i test position is recorded as a _i (i=1,2,...,M), and the test signal a _i is the electromagnetic interference emitted by the tested product itself Mixed signal with background noise.

The above two tests need to be carried out in accordance with the corresponding electromagnetic compatibility test standards.

Step 3: Input the adaptive filter with a _i (i=1,2,...,M) and ni ( _i =1,2,...,M) as references, and use the background noise in the mixed signal and the individual test The obtained background noise has the characteristic of correlation, the background noise in the mixed signal is filtered out, and the electromagnetic interference mixed signal x _i (i=1,2,...,M) emitted by the tested product itself is obtained.

Use this method to filter out various types of background noise in the test environment, eliminate the influence of background noise on the test results, and ensure the accuracy of the test results. Adaptive filtering can meet the requirements of some optimal criterion by adjusting its own parameters when the statistical characteristics of the input process are unknown or changing. In the present invention, the adopted adaptive filtering uses the least mean square criterion, that is, when the mixed signal a _i (i=1,2,...,M) containing background noise and the background noise n _i (i=1 ,2,…,M) when the least mean square value among them is the smallest, the maximum suppression of noise is achieved.

Step 4: Combine the electromagnetic interference mixed signals of each test position after filtering out the background noise into a joint matrix X=[x ₁ ,x ₂ ,…,x _M ] ^T .

The measurement receiver collects the electromagnetic interference mixed signal spectrum data at each position by means of frequency scanning, which are denoted as x ₁ , x ₂ ... x _M . Combining the electromagnetic interference mixed signal spectrum data x ₁ , x ₂ ... x _M at each position into a joint matrix X=[x ₁ , x ₂ ... x _M ] ^T , then each row of the joint matrix X is the electromagnetic interference mixture of a position signal spectrum data. Suppose N electromagnetic interference source signals are s ₁ , s ₂ …s _N , and the source signal matrix is S=[s ₁ , s ₂ …s _N ] ^T . Then X=AS, where A is an M×N order mixing matrix of electromagnetic interference source signals, and the mixing matrix represents the unknown mixing mode of each electromagnetic interference source signal in space. Let the separation matrix be W, the purpose of blind source separation is to solve the matrix W, so that the output result matrix Y=WX=WAS, then Y is the approximation to the source signal S, thus achieving the purpose of separation.

对混合信号x₁,x₂…x_M进行差值处理，得到新矩阵Z=BX。设W'为对Z进行盲源分离的分离矩阵，则对Z进行盲源分离的输出结果矩阵Y′=W′Z＝W′BX，可知Y′＝Y，即结果不变。Step 5: Since the row transformation is performed on the test signal X=[x ₁ ,x ₂ . . . x _M ] ^T , the nature of the signal will not be changed. Therefore, the signal difference matrix is constructed

Perform differential processing on the mixed signals x ₁ , x ₂ ... x _M to obtain a new matrix Z=BX. Let W' be the separation matrix for blind source separation of Z, then the output result matrix of blind source separation for Z is Y'=W'Z=W'BX, it can be known that Y'=Y, that is, the result remains unchanged.

Since each row of the matrix is a mixed signal of electromagnetic interference measured at a position, the transformation of its row does not affect the properties of the signals at each position. The matrix after row transformation is denoted as Z=[z ₁ ,z ₂ …z _M ] ^T . At this time, Z is equivalent to performing difference processing on the signals x ₁ , x ₂ ... x _M of each position in the joint matrix of mixed signals, that is, z ₁ = x ₁ -x ₂ , z ₂ = x ₂ -x ₃ , ... , z _M =x _M −x ₁ .

Step 6: Taking Z as the electromagnetic interference mixed-signal matrix, use the independent component analysis method to perform blind source separation.

The Z matrix obtained after difference processing is input as the electromagnetic interference mixed signal matrix, and the independent component analysis method is used for blind source separation to obtain the separated electromagnetic interference source signals.

Example

The method of the present invention is tested and verified with three radiated electromagnetic interference sources. Three radiated electromagnetic interference sources, one is the amplitude modulation signal s ₁ , the other is the frequency modulation signal s ₂ , and the other is the pulse modulation signal s ₃ . Among them, the fundamental carrier frequency of the amplitude modulation signal s ₁ is f ₁ =800MHz, the modulation frequency is f _mod1 =9MHz, and the modulation depth is 80%; the fundamental carrier frequency of the frequency modulation signal s ₂ is f ₂ =620MHz, and the modulation frequency is f _mod2 =7MHz; the fundamental carrier frequency of the pulse modulation signal s ₃ is f ₃ =340MHz, and the modulation frequency is f _mod3 =10MHz. The waveform diagram of the three signals is shown in Figure 2, and the mathematical expression is:

s ₁ ＝[1+0.8sin(2πf _mod1 t)]·sin(2πf ₁ t)

f ₁ =800MHz, f _mod1 =9MHz;

s ₂ =cos[2πf ₂ t+2cos(2πf _{mod 2} t)]

f ₂ =620MHz, f _mod2 =7MHz;

${\mathrm{<span\; class="notranslate">\; the\; s</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; \&infin;</span>}{\overset{<span\; class="notranslate">\; +</span><span\; class="notranslate">\; \&infin;</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; mod</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; u</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 4</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; mod</span>}\mathrm{<span\; class="notranslate">\; 3</span>}}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&CenterDot;</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 3</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>$

f ₃ =340MHz, f _mod3 =10MHz

Adopt the method for improving blind source separation of radiated electromagnetic interference mixed signal based on signal difference of the present invention, carry out following steps:

Step 1: Arrange five receiving antennas at different positions around the tested product, as shown in Figure 3, which is a schematic diagram of the location of the receiving antenna and the layout of the tested product, and the electromagnetic radiation source in Figure 3 is the tested product.

Step 2: First test the background noise at each test location, and record the data as n _i (i=1,2,...,M), and then conduct the radiation emission test on the tested product, and record the data as a _i (i= 1, 2,..., M), the test signal is the mixed signal of the electromagnetic interference mixed signal emitted by the tested product itself and the mixed signal of the background noise; the waveform diagram of the mixed signal measured at 5 positions is shown in Figure 4.

Step 3: Input the adaptive filter with a _i (i=1,2,...,M) and ni ( _i =1,2,...,M) as references, and use the background noise in the mixed signal and the individual test The obtained background noise is correlated, and the background noise in the mixed signal is filtered out to obtain the electromagnetic interference mixed signal x _i (i=1,2,...,M) emitted by the tested product itself; the mixed signal of 5 positions filters out the noise The subsequent wave form is shown in Figure 5.

Step 4: Combine the electromagnetic interference mixed signals of each test position after filtering out the background noise into a joint matrix X=[x ₁ ,x ₂ . . . x _M ] ^T . Suppose N electromagnetic interference source signals are s ₁ , s ₂ …s _N , then the matrix of source signals is: S=[s ₁ ,s ₂ …s _N ] ^T . Then X=AS, where A is an M×N order electromagnetic interference source-signal mixing matrix. Assuming the separation matrix is W, the output result matrix is Y=WX=WAS.

Step 5: Transform the test signal X=[x ₁ ,x _{2 .} . . x _M ] ^T without changing the nature of the signal. Therefore, the signal difference matrix is constructed If Z=BX, then Y′=W′Z=W′BX, it can be seen that Y′=Y, that is, the result remains unchanged; the signal waveform diagram of each position after the difference in the embodiment of the present invention is shown in FIG. 6 .

Step 6: Taking Z as the electromagnetic interference mixed-signal matrix, use the independent component analysis method to perform blind source separation. At the same time, the independent component analysis method is used to perform blind source separation on the undifferentiated signal matrix X. The comparison diagram of the separation results in the two cases is shown in Figure 7, where the column shown in (A) is the three interference signal sources obtained by using the independent component analysis method to perform blind source separation on the signal matrix X without difference processing, The column shown in (B) is the three interference signal sources obtained by performing blind source separation on the signal matrix Z obtained through difference processing by using the independent component analysis method. It can be clearly seen that in actual measurement, the results obtained by using the method of the present invention are the same as those obtained by using the conventional method, and three signal sources are obtained, and the signal separation effect of the difference preprocessing before separation by the method of the present invention is obviously better than that of Signal without difference preprocessing. Simultaneously, the number of signal separation iterative steps for difference preprocessing before separation by the method of the present invention is 9-12 steps, while the number of iterative steps for signal separation without difference processing is 12-15 steps, it can be seen that by adopting the method of the present invention The speed of blind source separation of radiated electromagnetic interference mixed signals is also improved.

Claims (2)

1. a method of improving the blind source separation of Radiative EMI mixed signal based on signal difference is characterized in that, comprises the steps:

Step 1: M test position is set in the test space, is used for placing receiving antenna;

Step 2: at first at each test position ground unrest is tested, the ambient noise signal of testing at i test position is

Then at each test position test specimen is carried out radiation emission test, the mixed signal of testing at i test position is a _i(i=1,2 ..., M), a _iThe electromagnetic interference (EMI) mixed signal of launching for test specimen self and the mixed signal of ambient noise signal;

Step 3: with a _i(i=1,2 ..., M) and n _i(i=1,2 ..., M) being reference quantity input adaptive wave filter, the ambient noise signal in the filtering mixed signal obtains the electromagnetic interference (EMI) mixed signal x that test specimen self is launched _i(i=1,2 ..., M);

Step 4: the electromagnetic interference (EMI) mixed signal x that each test position is obtained _i(i=1,2 ..., M) be combined into confederate matrix X=[x ₁, x ₂X _M] ^T

Step 5: make up the signal difference matrix Confederate matrix X is carried out difference process, obtain new matrix Z=BX, matrix Z=[z ₁, z ₂Z _M] ^T, z ₁=x ₁-x ₂, z ₂=x ₂-x ₃..., z _M=x _M-x ₁

Step 6: take new matrix Z as electromagnetic interference (EMI) mixed signal matrix, utilize independent component analysis method to do blind source and separate.

2. the method for improving the blind source separation of Radiative EMI mixed signal based on signal difference according to claim 1 is characterized in that, the described sef-adapting filter of step 3 uses the lowest mean square criterion to carry out auto adapted filtering.

Images