Directional radiation source positioning method combined with time difference method and time inversion technologyTechnical Field
The invention relates to the field of electromagnetic radiation source positioning, in particular to a directional radiation source positioning method combining a time difference method and a time reversal technology.
Background
Intentional electromagnetic interference (Intentional Electromagnetic Interference, IEMI) is increasingly receiving attention in the field of electromagnetic compatibility. IEMI is defined in IEC61000-2-13 as an intentionally created electromagnetic energy to be introduced into electrical and electronic systems, which in turn have adverse effects on the system, light-duty signal disturbance, logic confusion, heavy-duty load short-circuiting, dielectric breakdown. The high-power microwave radiation source is generally classified into radiation and conduction interference, the former is usually referred to as a high-power microwave radiation source, and the high-power microwave radiation source consists of a pulse generator and a high-power radiation antenna, and has the characteristics of long acting distance, wide frequency spectrum component, extremely strong directivity and the like, so that the high-power microwave radiation source is more widely focused.
In order to avoid the threat posed by the IEMI radiation source, it is the first viable measure to precisely locate it and then take effective measures to prevent it. Currently, the method for spatial positioning of a radiation source mainly includes an incoming wave direction method (Direction of Arrival, DOA), an energy attenuation method (RECEIVED SIGNAL STRENGTH Indication, RSSI), a Time of Arrival (TOA), a Time difference positioning method (TIME DIFFERENCE of Arrival, TDOA), and the like, and performs positioning by extracting characteristic information such as an angle difference, an energy difference, and a Time difference of Arrival of a signal generated by the radiation source at a measurement point. Generally, the above algorithm is in use, typically treating the radiation source as a non-directional point source, irrespective of the effect of radiation directivity. But when applied in IEMI radiation source positioning, the antenna gain is dependent on the radiation direction, the signal frequency, considering that such radiation sources are typically formed by radiation antennas with a relatively strong directivity.
Referring to fig. 1, a common radiation signal includes a plurality of frequency components, attenuation of each direction and each frequency point is inconsistent, and then a stronger directional radiation characteristic is formed, so that waveform characteristics of time domain response signals of receiving antennas located at different positions are inconsistent, which results in that accurate waveform information is difficult to extract by the method, and positioning errors are generated. Thus, for IEMI radiation sources with extremely strong directivity, a precise positioning method is needed.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a directional radiation source positioning method combined with a time difference method and a time reversal technology.
The invention is realized by the following technical scheme:
a directional radiation source positioning method combining a time difference method and a time inversion technology comprises the following steps:
S1, selecting random observation points rm(xm,ym in a plurality of areas to receive electromagnetic radiation time domain full wave signals, and recording instantaneous electric field signals E (rm, T) generated by a radiation source, wherein xm、ym is a coordinate point of the random observation points, m is a sequence number of the random observation points, T epsilon [ Ts,ts+T],ts ] is a measurement starting moment, and T is a signal sampling recording length;
S2, extracting peak points and corresponding moments tm of the peak points from the instantaneous electric field signals E (rm, t) received by the random observation points rm(xm,ym), and performing pre-positioning by using a TDOA method to obtain pre-positioning pointsCoordinate points that are predetermined points;
S3, at a predetermined siteCentered at a predetermined positionThe vicinity uses EMTR techniques to determine the radiation source location.
Preferably, in S1, the number of the random observation points rm(xm,ym) is greater than or equal to three.
Preferably, in S2, the predetermined siteBy carrying out simultaneous acquisition on any three random observation points rm(xm,ym), the calculation formula is as follows:
Wherein c is the light speed, and m1、m2 and m3 are the sequence numbers of the random observation points respectively.
Preferably, in S3, the following steps are included:
s31, at a predetermined siteSelecting a guessed radiation source position rn(xn,yn), wherein N is the number of the guessed radiation source, n=1, 2,3,..N, N is a constant, and obtaining the distance d (rm,rn) between the random observation point rm(xm,ym) and the guessed radiation source position rn(xn,yn);
S32) returning the time-reversal signals ETR(rm, T-T) to the guessed radiation source position rn(xn,yn) at the random observation points rm(xm,ym), and calculating the superimposed signals Esum(rn, T) for each random observation point rm(xm,ym) to the guessed radiation source position rn(xn,yn);
S33, calculating the energy Pt(rn of the superimposed signal at each guessed radiation source position rn(xn,yn) by using the superimposed signal Esum(rn, t);
S34, comparing the energy of the superimposed signals at each guessed radiation source position rn(xn,yn), taking the energy maximum point as the radiation source position, i.e
Preferably, in S31, the step of guessing the radiation source position rn(xn,yn) is selected by selecting a predetermined positionThe target area S is selected as the center, the target area S is divided into grids with the interval d, and the end points of the grids are guessed radiation source positions rn(xn,yn).
Preferably, in S32, the time-reversal signals ETR(rm, T-T) are obtained by time-reversal of the signals E (rm, T) received at the random observation point rm(xm,ym for an acquisition duration T.
Preferably, in S32, the signal amplitude is kept unchanged during the return of the time-reversal signals ETR(rm, T-T).
Preferably, in S32, the calculation formula of the superimposed signal Esum(rn, t) at rn(xn,yn) is:
Where d (rm,rn) represents the distance between the observation point rm(xm,ym) and rn(xn,yn) at the position of the speculative radiation source, c represents the speed of light, and M represents the number of speculative radiation sources, m=n.
Preferably, in S33, the calculation formula of the energy Pt(rn) of the superimposed signal at the guessed radiation source position rn(xn,yn) is:
compared with the prior art, the invention has the following beneficial effects:
The invention performs the pre-positioning by using the TDOA technology and the accurate positioning by using the EMTR technology, effectively combines the advantages of the pre-positioning and the accurate positioning, overcomes the adverse effect of the directional radiation characteristic on the TDOA, and overcomes the defects of longer time consumption and larger memory requirement of the traditional EMTR technology.
The time information corresponding to the peak value is extracted, the TDOA method is utilized for pre-positioning, rough information of the target source position is obtained, the range of the position of the radiation source is narrowed, the accurate positioning of the next step is facilitated, and the calculation efficiency is improved.
In the final positioning, the time and space self-adaptive focusing principle of the time inversion technology is utilized, which is different from the judgment and analysis of the details of the radiation signals by the traditional method, and inversion calculation is carried out by injecting in an integral mode, so that the characteristics of full-wave signals are effectively utilized, and the problems of misjudgment of the arrival time of the signals, difficult extraction of the characteristics of the signals and the like are avoided.
The invention can effectively utilize the directional radiation characteristic brought by the radiation source, and avoid the problem of precision error brought by extracting information such as time difference, energy difference and the like among the multi-measuring-point received signals.
Further, at least 3 observation points are used for receiving electromagnetic radiation time domain full wave signals generated by the radiation source.
Further, the signal amplitude is kept constant during the return of the time-reversal signals ETR(rm, T-T) in order to reduce the positioning errors and energy attenuation caused by the distance.
Furthermore, the maximum energy point of the returned superimposed signal is used as the corresponding radiation source position, so that the radiation signals in different directions are synchronously analyzed, the difference of time domain waveforms is reduced, the influence of the maximum energy point on the directivity of the radiation source is reduced, and the positioning error is reduced.
Drawings
Fig. 1 is a radiation pattern of a directional radiation antenna at various frequencies;
FIG. 2 is a computational flow diagram of a directional radiation source positioning method incorporating time difference and time inversion techniques in accordance with the present invention;
FIG. 3 is a schematic diagram of the positions of random observation points;
FIG. 4 is a radiation model;
FIG. 5 is a schematic illustration of a predetermined site;
FIG. 6 is a schematic diagram of a time reversal operation;
FIG. 7 is a time domain waveform diagram of a radiation field;
FIG. 8 is a graph of a spectrum in a radiation field;
FIG. 9 is a diagram of an experimental configuration;
FIG. 10 is a waveform diagram of a received electric field;
fig. 11 is a normalized energy plot.
Detailed Description
The invention will now be described in further detail with reference to specific examples, which are intended to illustrate, but not to limit, the invention.
A directional radiation source positioning method combining time difference method and time inversion technology, referring to fig. 2, comprising the steps of:
S1, referring to FIG. 3, random observation points rm(xm,ym in a plurality of areas are selected to receive electromagnetic radiation time domain full wave signals, and instant electric field signals E (rm, T) generated by a radiation source are recorded, wherein xm、ym is a coordinate point of the random observation points, m is a sequence number of the random observation points, T epsilon [ Ts,ts+T],ts ] is a measurement starting moment, T is a signal sampling recording length, and the number of the random observation points rm(xm,ym) is more than or equal to three.
Referring to fig. 4, since a convolution operation is required for a radiation expression or the like in the time domain, a derivation process becomes complicated, and thus the derivation is performed in the frequency domain for the method. The signal received by the m-th measuring point is E (rm, t), and the corresponding frequency domain expression E (rm, omega) is:
Wherein: -unit vector of main polarization direction of electric field, -I (ω) -amplitude of excitation signal, wherein ω represents angular frequency, - ψ (ω) -phase of excitation signal, -F (θm,φm, ω) -directivity factor, -d (rs,rm) -distance of radiation source from mth measurement point; time delay, where k represents the wavenumber of free space.
The directional radiation characteristics affect the amplitude and phase of the radiation field signal. The directional gain in terms of amplitude is related to the frequency of the excitation signal and the angle the radiation direction makes with the main axis of the radiating antenna. Referring to fig. 1, when a pulse having a wide frequency band is used as excitation, the gain of each frequency point is changed with the change of the radiation direction after fourier transformation of an electric field signal generated by antenna radiation, and the gain of each frequency point is usually the largest on the principal axis. When the difference of signals in the frequency domain is reflected in the time domain, the difference comprising rising edges, amplitude values and the like can appear on the characteristics of the time domain waveforms in different radiation directions, and the difficulty of positioning the radiation source is increased. On the other hand, the directivity also affects the phase of the field signal. However, considering that in free space, the detection array is typically located in a far field region further from the radiation source, the size of the radiation source itself is negligible, i.e. it is regarded as a point source, and the equiphase plane generated by its radiation can be regarded approximately as a sphere. Therefore, the influence on the phase due to the change in radiation direction can be ignored, and only the influence on the amplitude due to the directional radiation characteristic, i.e., F (θm,φm, ω) contains only the real part, is considered.
S2, extracting a time tm corresponding to a peak point from a signal E (rm, t) received by a random observation point rm(xm,ym), and performing pre-positioning by using a TDOA method to obtain a pre-positioning pointWherein,A coordinate point that is a predetermined site.
Predetermined siteBy carrying out simultaneous acquisition on any three random observation points rm(xm,ym), the calculation formula is as follows:
Wherein c is the speed of light.
S3, referring to FIG. 5, at a predetermined siteCentered at a predetermined positionNearby uses EMTR technology to find the radiation source location.
S31, at a predetermined siteSelect guess radiation source position rn(xn,yn) to a predetermined pointThe target area S is selected as the center, the target area S is subdivided into grids with an interval d, the end points of the grids are the guessed radiation source positions rn(xn,yn), N is the number of the guessed radiation sources, n=1, 2,3,...
S32, referring to fig. 6, the time-reversed signals ETR(rm, T-T) are returned to the guessed radiation source position rn(xn,yn) at the random observation point rm(xm,ym) in a manner that the signal amplitude is unchanged, and the time-reversed signals ETR(rm, T-T) are obtained by time-reversing the signals E (rm, T) received at the random observation point rm(xm,ym) within the acquisition period T.
The time-reversal transformation is a time-series inverse operation on the measured signal, which corresponds to a phase conjugate transformation in the frequency domain. Since the directivity factor includes only the real part, it is not affected by the time-reversal transformation. Thus, the frequency domain expression E*(rm, ω) for the time-reversed signal can be expressed as:
Calculating the superimposed signal Esum(rn, t) of each random observation point rm(xm,ym) transmitted back to the guessed radiation source position rn(xn,yn), the calculation formula of the superimposed signal Esum(rn, t) is:
Where d (rm,rn) represents the distance between the observation point rm(xm,ym) and rn(xn,yn) at the position of the speculative radiation source, c represents the speed of light, and M represents the number of speculative radiation sources, m=n.
The frequency domain expression of the superimposed signal Esum(rn, t) is Esum(rn, ω):
S33, calculating the energy Pt(rn) of the superimposed signal at each guessed radiation source position rn(xn,yn) using the superimposed signal Esum(rn, t), the calculation formula of the energy Pt(rn) of the superimposed signal is:
the energy Pt(rn) of the superimposed signal is calculated as follows:
From the equation, Pω(rn) is determined by the ratio of the amplitude to the amplitude ofAndTwo are related. Wherein only the second term is related to the position of the guessed source point. When (when)When, i.e(N is an integer), Pω(rn exhibits a maximum value. Considering that the radiated signal is typically a wideband signal and that Pω(rn) and the wavenumber k are both related to the frequency range, if and only if n=0, i.eAt this time(D is a constant value), the maximum value of the occurrence of Pω(rn) can be obtained. At the position corresponding to this value, all the return signals are superimposed in phase, so that a maximum value occurs.
According to the Paswald's theorem, the total energy of a signal may be calculated as an integral of energy per unit time over the entire time, or may be obtained as an integral of energy per unit frequency over the entire frequency range. Therefore, the energy can be directly solved through the superposition signals in the time domain, fourier transformation is not needed, and the calculation difficulty is effectively reduced.
S34, comparing the energy of the superimposed signals at each guessed radiation source position rn(xn,yn), taking the energy maximum point as the radiation source position, i.e
Examples
(1) The experimental configuration is carried out, and the experimental environment is arranged in an electromagnetic compatibility dark room with the length of 11m, the width of 5.9m and the height of 6.3 m. The radiation system comprises an ultra-wideband pulse source and an IRA radiation antenna, wherein the IRA radiation antenna is a typical IEMI directional radiation antenna, the radiation band of the IRA radiation antenna is selected to be 10MHz-2 GHz, a radiation electric field with a potential of about 3kV and an upper limit frequency of 3dB of 1GHz can be generated under far field conditions, and fig. 7 and 8 are a radiation field time domain waveform diagram and a frequency spectrum diagram at a distance of 4m from the radiation system respectively.
The measuring system consists of a TEM horn receiving antenna, a transmission cable, an attenuator and an oscilloscope, can monitor the electric field waveform of the irradiation space in real time, measures the electric field peak value to exceed 100kV/m, has the bandwidth of several GHz, and meets the test requirement on the radiation signals.
(2) Signal reception referring to fig. 9, in the experiment, the radiating antenna focus was set at (0, 0), and the antenna main axis direction was set to be 3 ° from the x axis. The receive array consists of three TEM electric field antennas D1, D2 and D3, located at (3.6,1.5), (4.6,0) and (3.6, -1.5), respectively. Since the polarization direction of the electric field generated by the radiation system is perpendicular to the ground under far field conditions, the polarization direction of the receiving antenna is set to be consistent with the electric field direction.
Fig. 10 shows waveforms of the electric field of radiation measured by the array. It can be seen that due to the influence of the directivity of the radiation antenna and the distance between the measuring point and the radiation source, the characteristics of the amplitude, the front edge and the like of the received waveform are different, which is also the reason for positioning errors in the traditional TOA and other methods.
(3) The method comprises the steps of presetting, and obtaining peak time of each measuring signal, wherein the peak time of signals received by D1, D2 and D3 is respectively 1.25ns, 3.8ns and 1ns. Based on the measurement signals and the measuring point positions, the coordinates of the preset position point are obtained by solving the TDOA method to be (-2, -0.15).
(4) And (3) accurately positioning, taking a preset site (-2, -0.15) as a center, selecting an area x epsilon (-5, 1), y epsilon (-3, 3) as a target area, taking d=0.02 m as an interval to perform grid subdivision in consideration of the limitation of a computer memory, and taking all points as guessed radiation source positions to perform calculation. Referring to FIG. 11, an accurate positioning model was built using MATLAB, based on normalized inversion of the EMTR method to find the energy maximum location, where the energy maximum point was located (-0.28,0.1), with a positioning error of 0.3m. It can be seen from 9 that the radiation system is less than 5m from the measurement point and its radiation does not fully meet far field conditions compared to the size of the antenna itself. Thus, the error is within an acceptable range.