RU2626284C1 - Passive method of detecting vehicles by its own acoustic noise - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: synchronously receive a signal for two antennas, digitize and store information arrays. Split the signal into intervals, carry out identification by maxima in the amplitude spectrum. Then carry out the scanning directional characteristics of receiving antenna for each nth time interval according to the total signal from the output of each of the two receiving antennas, build fan chart, determine the availability of transport by the presence of unexpected fan charts, then determine the azimuth direction relative to the center of the antenna as a weighted average of maximum angles peaks in fan pattern diagram, the sign of the difference between the angles, decide on direction, calculate the current y coordinate using the coordinates for the neighbouring define interval speed. The second implementation, determine the direction of the movement and location of the sign of the slope dependence of delay time correlation of maximum convolution arrays and on the dynamics of its inclination, calculate the autocorrelation function of total data then sum array and determine the speed of the vehicle as the ratio of the distance between the centers of the antennae to the time between the maximums of the autocorrelation function.

EFFECT: enabling simultaneous identification, calculate the speed and direction of movement.

2 cl, dwg 9 1 tbl

Description

The present invention relates to methods for monitoring the movement of vehicles and, in particular, their detection.

There are various ways to detect objects of transport equipment. The main tasks that are usually solved when a vehicle is detected are: localization of the vehicle in space, its identification, determination of its speed and direction of movement. To solve these problems, in combination or separately, very often they simultaneously use measuring instruments working with different types of physical fields: acoustic (sound range), seismic, optical, infrared, etc., since this combination makes it easier to achieve greater information content and greater independence from conditions observations.

Measuring instruments can carry out their work both actively and passively, the latter having a significant advantage, since it does not allow detecting the presence of measuring instruments in a controlled area of space. Here we can mention such microwave devices of active location, which are widely used in traffic control, as police traffic police radars, which, as you know, are successfully counteracted by radar detectors.

Purely acoustic passive methods of detecting vehicles are known very little, and this is most likely because the known methods still have not allowed to solve all the detection tasks in the aggregate.

So, for example, the method by which the device known from RF patent No. 2509372 “A device for detecting moving land vehicles by acoustic signals” (IPC G08G 1/01, G08G 1/04; priority date 06.22.2012; Authors: Dudkin V.A., Pankov A.A., Akimova Yu.S., Patentee: Federal State Budgetary Educational Institution of Higher Professional Education "Penza State University" (Penza State University)), and in which they receive acoustic signal on one single passive sensor (microphone), amplified, and digitized signal, and then produce various manipulations with them from the treatment area to the principles of so-called "fuzzy logic" accept or not accept the decision on the presence of the vehicle. The mentioned method, according to the authors of the patent, can be used for noise-immune detection of ground vehicles by their acoustic signals while protecting territories and approaches to various objects, but from the description it follows that with its help it is only possible to distinguish traffic noise from a natural acoustic background and fix the fact of violation of the boundaries of the protected area by an undefined vehicle.

From US 7071841, “Truck acoustic data analyzer system” (IPC: G08G 1/04; G08G 1/052; G08G 1/054; G08G 1/056; priority date: 08/19/2004; authors: Haynes; Howard D. (Knoxville , TN), Akerman; Alfred (Knoxville, TN), Ayers; Curtis W. (Kingston, TN); patentee: UT-Battelle, LLC (Oak Ridge, TN)) there is a passive acoustic method for determining the speed of a large freight transport, which carry out the reception and analysis of the signal from one or two microphones and determine the speed of the vehicle by the Doppler shift discrete in the private spectrum of the acoustic noise emitted by it. The method is intended to work with only one of the types of existing vehicles and to solve only one of the problems of detecting a vehicle.

US 5444443 “Sound source determining system” (IPC G08G 1/01; Authors: Umeda Misao, Ishikawa Keiichi; Patentee: Ishikawa Manufacturing Co LTD [JP]; Priority Application JP 19930223671 of 09.09.1993), in which describes the operation of the system for determining the location of the aircraft on the runway at the airport. Signal reception is carried out on at least one multi-element antenna — microphone array, from the outputs of which the signal is fed to delay lines that are individual for each selected channel, with a tunable delay time, and in the channels corresponding to microphones increasingly distant from the middle of the receiving antenna, signal delays by times that are multiples of the period allocated by the selective filters used to filter the received noise (..., -3τ, -2τ, -τ, 0, τ, 2τ, 3τ, ...), where τ is the period of the filtered spectral component Options. Then produce a summation of all signals from the outputs of the delay lines.

The total signal is fed to the input of a tunable (adaptive) filter, the center of the filter band corresponds to a certain frequency (ƒ ₀ = 1 / τ) in the signal spectrum (aircraft turbine noise). The delay tuning in each receiving element of the antenna is performed in conjunction with the frequency tuning of the filter. They achieve maximization of the total total signal at the output of the receiving antenna (selected spectral component), which corresponds to a certain angular direction from which the plane wave front arrives at the selected frequency. The azimuthal direction to the vehicle is automatically calculated by the delay in the angle determination device by plotting the radius line from the center of the receiving antenna toward the source. The indicated line gives direction to a noisy vehicle. When using simultaneously two or more microphone arrays, they are able to calculate the position of a noisy vehicle.

The prototype is aimed at solving a very narrow and specific problem - identifying the location of a stationary vehicle - an airplane with turbine engines turned on, landing or preparing to take off, in the open space of the airport field. It should be noted that aircraft turbines are sources of very powerful air-acoustic incoherent broadband noise. In the prototype, it is necessary to use a bulky well-developed receiving antenna (in the prototype about 15 m in length, i.e. with a wide aperture (many wavelengths) and with a very dense filling with sensors (step - less than half the wavelength)).

Each of the sensors should have a sufficiently wide working frequency band due to the fact that the noise of the proposed source does not have spectral components whose radiation would have mutual spatial coherence, capable of interfering with each other and providing bursts characteristic of the vehicle in its angular wave spectrum, which would allow to solve the detection problem only due to the angular resolution of the receiving antenna described in the prototype, even in the absence of an open floor Reverb airport, even at the lowest possible noise level (ie in the absence of a number of other fixed aircraft turbines included or taking off or coming in to land planes).

The use of a single multi-element linear array of acoustic sensors allows you to determine the azimuthal direction to the vehicle, using several - to determine the coordinates of the vehicle (to identify its location). If the interference is significant, for example, reverberation, then the proposed method ceases to be effective, since it does not allow either to allow two vehicles standing close to each other with working turbines, or to isolate the vehicle against the background of noise from another closely moving similar vehicle. It can be stated that the bulk of the detection tasks disappears in the case of the prototype by itself, since it is not solvable in principle due to the features of the vehicle selected for detection and the observation conditions.

Thus, the disadvantages of the prototype are: the complexity and bulkiness of the recording equipment, low noise immunity, the ability to work only with stationary vehicles and solve only one of the detection tasks - localization of the vehicle in space.

The problem to which the invention is directed is to create a passive method for detecting a vehicle by its own acoustic noise, which would simultaneously detect the vehicle's localization in space, carry out its identification, determine its speed and direction of movement.

The technical effect is achieved by registering the useful signal emitted into the surrounding atmosphere by the vehicle using at least one receiving antenna, consisting of acoustic sensors placed in a line with a constant pitch, and determining the azimuthal direction to the vehicle relative to the center of the receiving antenna.

, где

,

, ƒ_j(j→j+1) - шаг перестройки частоты, осуществляют перестройку в характерной для данного транспортного средства частотной полосе ΔF и в ней же производят обратное БПФ-преобразование для получения суммарного сигнала с выхода каждой из двух приемных антенн на n-м временном интервале. Для каждого n-го временного интервала по суммарному сигналу с выхода каждой из двух приемных антенн в полярных координатах строят веерные диаграммы, характеризующие азимутальное распределение максимумов модуля амплитудного отклика приемной антенны на собственный акустический шум транспортного средства, и убеждаются в наличии транспортного средства на трассе по присутствию в веерных диаграммах пиков с перепадом от максимума до ближайшего минимума не менее 15 дБ. Затем определяют азимутальное направление на транспортное средство относительно центра каждой приемной антенны в текущем n-м временном интервале как средневзвешенное значение углов максимальных пиков в веерной диаграмме. По знаку разности средневзвешенных значений углов максимальных пиков в веерной диаграмме для двух следующих друг за другом n-го и n+1-го временных интервалов любой из приемных антенн принимают решение о направлении движения транспортного средства по трассе, текущую координату x_n транспортного средства на трассе определяют как точку пересечения линий

и

, идущих из центров приемных антенн по направлениям средневзвешенных значений углов максимальных пиков в веерных диаграммах обеих приемных антенн для n-го временного интервала, используя координаты транспортного средства в соседние n-й и n+1-й временные интервалы x_n и x_n+1 определяют скорость движения V=(x_n+1 - x_n)/τ.New, in the case of implementing the invention according to claim 1 of the formula, is that they begin to register a useful signal if it exceeds the threshold level of the acoustic background of the surrounding space in calm weather, P = 60 dB, and if the useful signal exceeds the threshold level of the acoustic background 6 dB during the duration of the full implementation of the signal 10-20 s synchronously receive a useful signal to two receiving antennas, each of which consists of K≥16 acoustic sensors placed in a line along the path izhnii at a known distance H from it with a step d of not more than 1/4 of the sound wavelength λ of the portion used by the receiving antennas in the mid-frequency sound range of 50-1000 Hz, with Kd >> λ, and the distance L between the centers of the receiving antennas is 50-200 m , digitize the received signal from each receive antenna at a sampling frequency ƒ _i and placed in a computing device signals received by each k-th transducer of each receiving antenna, a k-th digital array for each sensor break duration full realization signal n N equal time intervals the duration τ, is carried out the identification of the vehicle from the observed in its amplitude spectrum behavior maxima characteristic of the vehicle ΔF frequency band, provide two quadrature channels for each k-th transducer of each receiving antenna by constructing cos (2πƒ _j t) and sin (2πƒ _j t) FFT spectrum signal at each n-th time slot and form a supplement to the spectral phase of the signal component to be scanned over the angle of the maximum main lobe characteristics receive antenna

where

,

, ƒ _j (j → j + 1) is the frequency tuning step, tuning is performed in the frequency band ΔF characteristic of the vehicle and the inverse FFT conversion is performed in it to obtain the total signal from the output of each of the two receiving antennas on the nth time interval. For each n-th time interval, fan diagrams are constructed using the total signal from the output of each of the two receiving antennas in polar coordinates, characterizing the azimuthal distribution of the maxima of the amplitude modulus of the receiving antenna to the vehicle’s own acoustic noise, and make sure that the vehicle is present on the track by presence in fan diagrams of peaks with a difference from the maximum to the nearest minimum of at least 15 dB. Then, the azimuthal direction to the vehicle relative to the center of each receiving antenna in the current n-th time interval is determined as the weighted average of the angles of the maximum peaks in the fan diagram. According to the sign of the difference between the weighted average values of the angles of the maximum peaks in the fan diagram for two successive n-th and n + 1-th time intervals of any of the receiving antennas, they decide on the direction of the vehicle along the highway, the current coordinate x _n of the vehicle on the highway defined as the point of intersection of lines

and

coming from the centers of the receiving antennas along the directions of the weighted average values of the angles of the maximum peaks in the fan diagrams of both receiving antennas for the n-th time interval, using the vehicle coordinates in the adjacent n-th and n + 1-th time intervals x _n and x _{n + 1} determine the speed of movement V = (x _{n + 1} - x _n ) / τ.

New in the case of implementing the invention according to claim 2 of the formula is that they begin to register a useful signal if it exceeds the threshold level of the acoustic background of the surrounding space in calm weather, P = 60 dB, and if the useful signal exceeds the threshold level of the acoustic background by 6 dB during the duration of the full implementation of the signal 10-20 s synchronously receive a useful signal to two receiving antennas, each of which consists of K≥16 acoustic sensors placed in a line along the path living at a known distance H from it with a step d of no more than 1/4 of the sound wavelength λ of the portion used by the receiving antennas in the mid-frequency sound range of 50-1000 Hz, and the distance L between the centers of the receiving antennas is 10-50 m, the received signal from each receive antenna with a sampling frequency ƒ _i and place the signals recorded by each k-th sensor of each receive antenna in the form of the k-th digital array in the memory of the computing device. For each sensor, the duration of the full signal implementation is divided into N equal time intervals of duration τ, the vehicle is identified by the behavior of the maxima observed in its amplitude spectrum in the frequency band ΔF characteristic of the vehicle, the direction of the vehicle and its location relative to the receiving antenna are determined by the sign of the slope of the dependence of the delay time of the correlation maximum of convolution of arrays corresponding to individual sections olnoy implementation signal _{g kn = g kn (t)} , with more wind- pairs of sensors in a single reception antenna, and the dynamics of its inclination on passing from the current n-th to the subsequent n + 1-th time slot is determined azimuthal direction of transport means relative to the center of one of the receiving antennas as equal to 90 ° and fix the location of the vehicle on the beam of the receiving antenna at the moment the angle of inclination of the dependence of the delay time of the correlation maximum convolution vanishes, take the autocorrelation function ummarnogo dataset both receive antennas with subsequent summation over all elements of each of the receiving antennas, and determines the vehicle speed as a ratio of the distance between the centers of the receiving antennas to a time interval between the characteristic peaks in the autocorrelation function.

The invention is illustrated by the following drawings.

и

- линии, идущие из центров соответственно первой и второй приемных антенн по направлению к транспортному средству под азимутальными углами соответственно θ_1n и θ_2n в n-й интервал времени.In FIG. 1 shows the measurement scheme: L is the distance between the centers of the receiving antennas, N is the distance to the track,

and

- lines coming from the centers of the first and second receiving antennas, respectively, towards the vehicle at azimuthal angles θ _1n and θ _2n, respectively, in the n-th time interval.

FIG. 2 illustrates the intrinsic acoustic noise of a tram car adopted over the duration of the full implementation of the signal, exceeding the background sound level by more than 6 dB. The abscissa axis represents time in seconds, the ordinate axis represents dimensionless units.

In FIG. Figure 3 shows the experimental amplitude spectra of signals received over the course of the full implementation of the signals for a tram car (left) and a passenger car (right). The abscissa is the frequency in hertz, the ordinate is the dimensionless unit.

In FIG. Figure 4 shows the experimentally obtained fan diagrams illustrating the azimuthal distribution of the maxima of the amplitude modulus of the receiving antenna to the acoustic noise of the tram car for different time periods: n = 1, the vehicle is approaching the beam of the receiving antenna (case a ); n = 4, the moment the vehicle passed the traverse of the receiving antenna (case b); n = 6 (case c), n = 8 (case d), the vehicle moves away from the traverse of the receiving antenna. The arrow indicates the direction of the vehicle: the car begins to move at θ = 180 ° and then follows the arrow to θ = 0 °.

In FIG. Figure 5 shows the diagrams obtained from experimental data in the same polar coordinates as the fan diagrams, in which the petal maximum indicates the azimuthal direction to the vehicle for a given time interval, calculated by the formula for the weighted average values of the angles: a - the vehicle approaches the receiving beam antennas; b, c - the moment the vehicle passes the traverse of the receiving antenna; g, d - the vehicle moves away from the traverse of the receiving antenna.

FIG. 6 illustrates the current position of the vehicle relative to the receiving antennas.

In FIG. Figure 7 shows some of the characteristic dependences of the delay time of correlation maxima obtained from experimental data on the number of sensors for different time intervals, n increases from top to bottom, in the case of rail transport (left) and passenger vehicles (right): the vehicle approaches the traverse of the receiving antenna on the left to the right (cases a , b, c, d); n = 4, the moment the vehicle traverses the receiving antenna (case d); the vehicle moves away from the traverse of the receiving antenna (cases e, g).

In FIG. Figure 8 shows some of the characteristic dependences of the delay time of the correlation maxima obtained from experimental data on the number of the pair of sensors for different time intervals, n increases from top to bottom, in the case of passenger cars moving from right to left: the vehicle approaches the traverse of the receiving antenna from left to right (cases a , b) n = 4, the moment the vehicle passes the traverse of the receiving antenna (case c); the vehicle moves away from the traverse of the receiving antenna (cases d, e, f).

In FIG. Figure 9 shows an example of modeling autocorrelation functions for the first four sensors of both receiving antennas 1-1, 2-2, 3-3, 4-4. The abscissa axis represents the delay time in dimensionless units, the ordinate axis represents dimensionless units.

For both cases of the implementation of the proposed passive method for detecting a vehicle by its own acoustic noise, the measurement scheme is the same and is carried out as follows.

Two receiving antennas, each of which consists of K≥16 acoustic sensors placed in a line along the route with a step d of no more than 1, are placed parallel to the route along which the traffic is moving, at a known distance H from it (see Fig. 1) / 4 sound wavelength λ of the portion used by the receiving antennas in the mid-frequency sound range of 50-1000 Hz. Since the acoustic noise produced by vehicles is characterized by a wide frequency band, acoustic sensors are used for measurements, the working frequency band of which covers the mid-frequency sound range. The centers of both receiving antennas are spaced relative to each other by a distance L.

The registration of the useful signal by both receiving antennas begins if it exceeds the threshold level of the acoustic background of the surrounding space in calm weather, P = 60 dB. Reception of a useful signal for its subsequent processing begins when the useful signal at both receiving antennas exceeds the threshold acoustic background level by at least 6 dB. The useful signal is received by both receiving antennas synchronously for a period of 10-20 s, called the duration of the full implementation of the signal, provided that during this period of time the signal level continues to exceed the threshold level of the acoustic background by at least 6 dB (see Fig. 2 ) Then, the received signals from each receiving antenna are digitized with a certain sampling frequency when digitizing ƒ _j and the signal recorded by each k-th acoustic sensor of the receiving antenna is placed in the memory of the computing device as a digital array (each sensor has its own k-th array). After that, for each acoustic sensor, the duration of the full implementation of the signal is divided into N time intervals of equal duration τ = 1-2 s, in order to further be able to analyze the nature of the change in time of the parameters of the received useful signal. On each k-th acoustic sensor there are sections of the full signal realizations g _kn = g _kn (t) for n⋅τ≤t≤ (n + 1) ⋅τ, which are realizations of the signal in the nth time interval.

In all the figures given below as an example, the number of time intervals N = 10, the sampling frequency ƒ _j = 1 Hz, the number of acoustic sensors in each receiving antenna K = 32, and they are placed with a step d = 0.1 m, which does not exclude the general case of implementing the method of other values of the mentioned parameters.

Then, the vehicle is identified by the behavior of the maximums of the amplitude spectrum observed in its amplitude spectrum in a certain frequency band ΔF.

It is known that different types of vehicles possess spectral "portraits" characteristic of them only. In some methods of identifying vehicles, they are recognized by their characteristic “pictures”. For the example of FIG. Figure 3 shows the amplitude spectra for a tram (solid) and a passenger car (discrete). But the main distinguishing feature of each type of vehicle, as the authors of the invention found out as a result of research, is the presence of individual maxima for a given vehicle in the amplitude spectrum of the received signal in the characteristic frequency band ΔF lying in the mid-frequency range. The presence of such maxima is due to the fact that, unlike the noise source in the prototype, in which the noise radiation in the entire wide band is out of phase along the wavefront, the considered vehicles in the designated frequency interval on a scale of the order of the aperture of the receiving antenna have certain components of acoustic noise having in the region the vehicle’s localization is of significant amplitude, they retain the spatial distribution of the phase along the front, and during propagation these acoustic waves interfere, which and manifests itself characteristic of only this vehicle “bursts” in the amplitude spectrum. In other words, the intrinsic acoustic noise of the vehicles in question is partially coherent, and for each particular vehicle it is partially coherent in its own way depending on the type of vehicle. Of course, to carry out the identification, a certain “base” of features characteristic of various vehicles is stored in the memory of the computing device.

Next, the operation of the method according to PP. Formulas 1 and 2 begin to differ.

To implement the invention according to claim 1, formulas create two quadrature channels for each k-th acoustic sensor of each receiving antenna by constructing cos (2πƒ _j t) and sin (2πƒ _j t) FFT spectra of the signal at each n-th time interval, t .e. build cos and sin Fourier spectra of each of n separate sections of the complete signal realization g _kn = g _kn (t):

,

,

,

,

и модуль -

.in addition, for each k, n and ƒ, their phases

and module -

.

Then, additives are added to the phase of the spectral components of the signal for subsequent scanning along the maximum angle of the main lobe of the directivity of the receiving antenna.

и соответствующую ей переменную величину - угол сканирования

. Формирование сканированной (веерной) диаграммы направленности осуществляют на основе введения дополнительного набега фазы сигнала. С учетом дополнительно введенной переменной

, массив значений которой составляет

, теперь имеем четыре переменные

, k, n, ƒ, где

, k=1…K, n=1…N и ƒ_i - шаг перестройки от минимальной до максимальной частоты в характерной для данного вида транспортного средства частотной полосе ΔF (перестройку частоты осуществляют с тем же шагом, с которым осуществляли дискретизацию принимаемого сигнала). Дополнительный набег фазы определяется выражением:To do this, introduce a new discrete variable

and its corresponding variable - scan angle

. The formation of a scanned (fan) radiation pattern is carried out on the basis of the introduction of an additional phase incursion of the signal. Given the additionally entered variable

whose array of values is

, now we have four variables

, k, n, ƒ, where

, k = 1 ... K, n = 1 ... N and ƒ _i is the tuning step from minimum to maximum frequency in the frequency band ΔF characteristic of a given type of vehicle (frequency tuning is performed with the same step with which the received signal was sampled). The additional phase incursion is determined by the expression:

,

,

where the figured numerical constant 2πd / s, the _sound has the dimension of a unit of time (at a step of the receiving antenna d = 0.1 m, it is 0.001848 s).

For each discrete value ƒ _{j, a} new complex spectrum is constructed:

Next, a summation of the newly formed spectral components is performed for all acoustic sensors of each receiving antenna k = 1 ... K, taking into account phase additions:

Then, the signal is reproduced by the inverse Fourier transform of the spectral components (the interval of values of j) in the frequency band ΔF from ƒ _min to ƒ _{max of} interest to us, which is typical for this type of vehicle, and obtaining on this basis the total amplitude response at the output of the receiving antenna to n th time interval:

и добавления

в фазу

выполняют смену угла сканирования и на основе перебора всех направлений

,

создают набор откликов

, соответствующий всем углам азимута. Для каждого

берут максимум модуля

, достигаемый при каком-то t. Строят веерную диаграмму, т.е. зависимость

от

, т.е. от

для

. В следующий n+1 временной интервал, сдвинутый на τ относительно n-го, получим

и, соответственно

, и т.д. Для каждого временного интервала n и соответственно каждого значения

на данном временном интервале находят максимум модуля амплитудного отклика, который откладывают на веерной диаграмме вдоль по радиусу. Таким образом, на основе формирования фазовых добавок для каждого временного интервала n строят две веерные диаграммы - визуализируют результат сканирования главного лепестка характеристики направленности каждой из приемных антенн.For each n-th time interval, fan diagrams in polar coordinates are constructed at the output of each of the two receiving antennas using the total amplitude response. To do this, calculate the maximum values of the modules of the amplitude response at the output of the receiving antenna at the nth time interval during step-by-step phase restructuring as follows. For each n-th time interval for each of the receiving antennas due to the transition

and additions

in phase

carry out a change in the scanning angle and based on enumeration of all directions

,

create a set of responses

corresponding to all azimuth angles. For everybody

take the maximum module

achieved at some t. Build a fan diagram, i.e. dependence

from

, i.e. from

for

. In the next n + 1 time interval shifted by τ relative to the nth, we obtain

and correspondingly

, etc. For each time interval n and, accordingly, each value

in this time interval, the maximum modulus of the amplitude response is found, which is laid on the fan diagram along the radius. Thus, based on the formation of phase additives for each time interval n, two fan diagrams are constructed — the result of scanning the main lobe of the directivity characteristics of each of the receiving antennas is visualized.

The interfering components of the acoustic signal received by the antenna, consisting of many acoustic sensors, appear on the fan diagrams as periodic modulation of the level of the amplitude response module - characteristic peaks form on the fan diagrams (see Fig. 4).

The set of sharp emissions on the fan diagram, if the dips from minimum to maximum values reach 15 dB or higher, serves as confirmation of the presence of a vehicle on the highway. It is the fact that the inverse FFT conversion is performed in a characteristic frequency band in which for a given type of vehicle there are regions of partial coherence of the received signal, which determines the presence of such sharp drops in the fan diagram.

If there is one or more peaks of the indicated type selected in amplitude on the fan diagram, the azimuthal direction to the vehicle relative to the center of each receiving antenna for the current n-th time interval is calculated as the weighted average of the angles of the maximum peaks in the fan diagram according to the formula:

,

,

- азимутальный угол, соответствующий максимуму m-го пика на веерной диаграмме n-го временного интервала.where 1≤m≤M, M ~ 3 ... 5, A _m is the relative weight of each of the maximum (two to five) peaks in the angular distribution,

- azimuthal angle corresponding to the maximum of the mth peak in the fan diagram of the n-th time interval.

In FIG. Figure 5 shows diagrams in polar coordinates, each of which indicates the azimuthal direction to the vehicle for a given time interval, calculated by the formula for the weighted average angles.

, либо наоборот

- принимается решение о направлении движения объекта справа налево или слева направо соответственно.Using two fan diagrams constructed for two successive time intervals n-th and n + 1-th, weighted average values of the angles of the maximum peaks in both fan diagrams are found and the direction of movement of the vehicle is determined by the sign of their difference. Whichever

or vice versa

- a decision is made on the direction of movement of the object from right to left or from left to right, respectively.

и

, идущих из центров приемных антенн под указанными азимутальными углами, дает текущую координату x_n транспортного средства на трассе для n-го интервала времени. Эту координату в выбранной системе отсчета всегда можно рассчитать. Так, для приведенного на фиг. 6 примера расположения транспортного средства относительно центров первой и второй приемных антенн, лежащих вдоль оси х с произвольно выбранной начальной точкой отсчета, с координатами в данной системе отсчета x₁ и х₂, рассчитать текущую координату транспортного средства можно с помощью следующих выкладок:Using the fan diagrams constructed for the nth time interval, for both receiving antennas, the weighted average values of the angles of the maximum peaks in both fan diagrams are determined, giving the current azimuthal direction to the vehicle relative to the center of each receiving antenna. Line crossing point

and

going from the centers of the receiving antennas at the indicated azimuthal angles gives the current coordinate x _n of the vehicle on the track for the nth time interval. This coordinate in the selected reference system can always be calculated. So, for the one shown in FIG. 6 an example of the location of the vehicle relative to the centers of the first and second receiving antennas lying along the x axis with an arbitrarily selected starting point of reference, with coordinates in this reference system x ₁ and x ₂ , you can calculate the current coordinate of the vehicle using the following calculations:

.

.

x ₁ + x ₂ -2x = -H (ctgθ ₁ + ctgθ ₂ ),

.

.

Using data on the location of the object x _n and x _{n + 1} in the adjacent n-th and n + 1-th time intervals, determine the speed V = (x _{n + 1-} x _n ) / τ, where τ is the duration of one time interval , the totality of which is the duration of the full implementation of the signal T = N⋅τ.

, где D - дистанция между транспортными средствами на трассе. Но, с другой стороны, можно упомянуть, что и никакие другие из существующих способов обнаружения транспортных средств при достаточно плотном потоке движения транспорта по трассе тоже не работают.It should be noted that the implementation of the invention according to claim 1 of the formula has certain specifics. Firstly, since the computational processing in this embodiment of the method takes a considerable amount of time, it is necessary to separate the receiving antennas at a significant distance from each other L = 50-200 m in order to guarantee the possibility of performing the necessary calculations during the travel time of the vehicle from one receiving antenna to other. In addition, the relation Kd >> λ must be satisfied - each receiving antenna must be well developed, this condition is imposed in order to minimize the parasitic effect of the level of the side lobes in the directivity of the receiving antenna compared to the main lobe and to reduce the negative effect of the “ruggedness” effects radiation patterns. And finally, in order to guarantee a separate vehicle with a fairly intense traffic on the highway, it is necessary to exclude the directivity of the receiving antenna of more than one object within the main lobe. For this, the condition must be met

, where D is the distance between vehicles on the highway. But, on the other hand, we can mention that none of the other existing methods of detecting vehicles with a fairly dense traffic flow on the highway also does not work.

To implement the method according to claim 2, after receiving, digitizing a useful signal and identifying a vehicle, the following operations are carried out.

For the n-th time interval, the cross-correlation function of the individual sections of the complete signal realization g _kn = g _kn (t) recorded by the acoustic sensors of one receiving antenna at a given time interval n⋅τ≤t≤ (n + 1) ⋅τ is calculated by convolution corresponding to these sections of the full implementation of the signal arrays from the output of the acoustic sensors, more and more distributed along the aperture of the receiving antenna. The reference is the first (far left) sensor in the receiving antenna, and the arrays corresponding to all sensors in the receiving antenna, taken in pairs with the reference, are sequentially convolved as follows: 1-1, 1-2, ... 1-k, ... 1-K.

для k=1…K и для (n-1)τ≤t≤(n+1)τ

for k = 1 ... K and for (n-1) τ≤t≤ (n + 1) τ

Then, for all time intervals, graphs of the dependence of τ _n (k) of the convolution correlation maximum delay are plotted for all pairs of sensors searched in the aperture of the receive antenna for each time interval (see Fig. 7 and Fig. 8).

The direction of movement of the vehicle (from left to right or from right to left) and its location relative to the receiving antenna (the current position of the noise source to the left or right of its traverse point) is determined by the sign of the slope of the dependence of the delay time of the correlation maximum of convolution (∂τ _n / ∂k> 0 or ∂τ _n / ∂k <0) and the dynamics of its slope during the transition from the current n-th to the next n + 1-th time interval (see Table 1).

одной из приемных антенн определяют азимутальное направление на транспортное средство как равное 90 градусам относительно центра данной приемной антенны и местонахождение транспортного средства на траверзе данной приемной антенны (т.е. текущая координата транспортного средства на оси х соответствует координате центра данной приемной антенны в выбранной системе отсчета).When the angle of inclination vanishes, the dependence of the delay time of the correlation maximum of convolution

one of the receiving antennas determine the azimuthal direction to the vehicle as equal to 90 degrees relative to the center of the given receiving antenna and the location of the vehicle on the beam of the given receiving antenna (i.e., the current coordinate of the vehicle on the x axis corresponds to the coordinate of the center of this receiving antenna in the selected reference system )

The autocorrelation function of the total data array from both receiving antennas is taken, followed by summing over all elements of each of the receiving antennas and the vehicle speed is determined as the ratio of the distance between the centers of the receiving antennas to the time interval between the characteristic maxima in the autocorrelation function.

To do this, create an array G _{k, n} (t) = g _{1, kn} (t) + g _{2, kn} (t) the sum of the individual sections of the full implementation of the signal g _kn = g _kn (t) for n⋅τ≤t≤ (n +1) ⋅τ from both receiving antennas. Take the autocorrelation function of this array

for each acoustic sensor k = 1 ... K of both receiving antennas and for the time interval (n-1) τ≤t≤ (n + 1) τ.

.Further, in order to increase the signal / background ratio, the functions K _{k, n} (t) are summed over all sensors to both receiving antennas:

.

на n-м временном интервале будет содержать три характерных максимума (см. пример на фиг. 9), любые два соседних из которых разделены во времени на Т=L/V, где L - расстояние между центрами приемных антенн, а V - скорость транспортного средства на трассе. Отсюда, соответственно, вычисляется скорость транспортного средства.Autocorrelation response

in the nth time interval it will contain three characteristic maxima (see the example in Fig. 9), any two of which are adjacent in time divided by T = L / V, where L is the distance between the centers of the receiving antennas, and V is the transport speed funds on the track. From here, respectively, the speed of the vehicle is calculated.

This embodiment of the method based on the correlation approach has an advantage over the implementation option according to claim 1 of the formula in that much milder restrictions are imposed on the possibility of its implementation due to the fact that the computational operations are much faster in time, therefore, for spatial resolution of transport means in the stream is enough diversity of receiving antennas at a distance equal to 2-3 linear dimensions of the vehicle itself. In our version of the method according to claim 2, the distance L between the centers of the receiving antennas is 10-50 m. To the characteristics of the receiving antenna itself, there are also no such stringent requirements as in the case of the method according to claim 1 of the formula.

As a result, both in the implementation of the invention according to claim 1 and in the implementation of the invention according to claim 2, the vehicle is passively detected by its own acoustic noise, including regardless of weather conditions and time of day, the vehicle is identified, its direction of movement is determined and speed, its localization comes to light.

The implementation of the invention according to claim 1 of the formula allows to identify the localization of the vehicle in space, i.e. to determine the azimuthal direction to the vehicle and its coordinates, almost at any time from the moment the reception of the useful signal by the receiving antennas begins, however, since the processing of the received signal takes a considerable amount of time, it is possible to determine the speed of the vehicle in this embodiment only a posterior mode (with a delay in time), in addition, it is necessary to separate the receiving antennas at a significant distance from each other to guarantee To make it possible to carry out the necessary calculations during the passage of a vehicle from one receiving antenna to another.

The implementation of the invention according to claim 2 of the formula allows you to fix the location of the vehicle only at the moments when it traverses one and the other receiving antennas, i.e. to identify its localization with limitations, but it allows you to quickly calculate its speed, and for the implementation of this variant of the method, it is enough to separate the receiving antennas at a distance comparable to the linear dimensions of the vehicle itself. Thus, both embodiments of the invention fulfill the task, but, depending on the observation conditions and the priority of individual detection tasks, you can use either one or the other.

Claims (2)

1. A passive method for detecting a vehicle by its own acoustic noise, in which a useful signal is recorded that is emitted into the atmosphere by the vehicle using at least one receiving antenna, consisting of acoustic sensors placed in a line with a constant pitch, and determining the azimuthal direction to the vehicle relative to the center of the receiving antenna, characterized in that they begin to register a useful signal if it exceeds the threshold the level of the acoustic background of the surrounding space in calm weather P = 60 dB, and provided that the level of the useful signal exceeds the threshold level of the acoustic background by 6 dB during the duration of the full implementation of the signal 10-20 s synchronously receive a useful signal to two receiving antennas, each of which consists of K≥16 acoustic sensors placed in a line along the road at a known distance H from it with a step d of not more than 1/4 of the sound wavelength λ of the portion used by the receiving antennas in the mid-frequency sound pazone 50-1000 Hz and Kd >> λ, and the distance L between the centers of the receiving antennas is 50-200 m, digitize the received signal from each receive antenna at a sampling frequency

and the signals received by each k-th sensor of each receiving antenna are placed in the memory of the computing device in the form of the k-th digital array, for each sensor, the duration of the complete signal implementation is divided into N equal time intervals of duration τ, the vehicle is identified by the observed in it the amplitude spectrum of the behavior of the maxima in the frequency band ΔF characteristic of the vehicle, create two quadrature channels for each k-th sensor of each receiving antenna by constructing

and

FFT spectra of the signal at each nth time interval and form additives to the phase of the spectral components of the signal for subsequent scanning along the maximum angle of the main lobe of the directivity of the receiving antenna

where

,

,

- frequency tuning step, tuning is carried out in the frequency band ΔF characteristic of a given vehicle, and it performs the inverse FFT conversion to obtain the total signal from the output of each of the two receiving antennas in the nth time interval, for each n-th time interval of the interval according to the total signal from the output of each of the two receiving antennas in polar coordinates, build fan diagrams characterizing the azimuthal distribution of the maxima of the modulus of the amplitude response of the receiving antenna to its own acus static noise of the vehicle, and make sure that the vehicle is on the track by the presence in the fan diagram of peaks with a difference from the maximum to the nearest minimum of at least 15 dB, then the azimuthal direction to the vehicle relative to the center of each receiving antenna in the current n-th time interval is determined as the weighted average of the angles of the maximum peaks in the fan diagram, by the sign of the difference between the average values of the angles of the maximum peaks in the fan diagram for two consecutive the hom of the n-th and n + 1-th time intervals of any of the receiving antennas decide on the direction of the vehicle along the track, the current coordinate x _n of the vehicle on the track is determined as the point of intersection of the lines

and

coming from the centers of the receiving antennas along the directions of the weighted average values of the angles of the maximum peaks in the fan diagrams of both receiving antennas for the n-th time interval, using the vehicle coordinates in the adjacent n-th and n + 1-th time intervals x _n and x _{n + 1} determine its speed V = (x _{n + 1} -x _n ) / τ. 2. A passive method for detecting a vehicle by its own acoustic noise, in which a useful signal is recorded that is emitted into the atmosphere by a vehicle using at least one receiving antenna, consisting of acoustic sensors placed in a line with a constant pitch, and the azimuthal is determined direction to the vehicle relative to the center of the receiving antenna, characterized in that they begin to register a useful signal provided that it exceeds the threshold about the level of the acoustic background of the surrounding space in calm weather P = 60 dB, and provided that the level of the useful signal exceeds the threshold level of the acoustic background by 6 dB during the duration of the full implementation of the signal 10-20 s synchronously receive a useful signal to two receiving antennas, each of which consists of K≥16 acoustic sensors placed in a line along the road at a known distance H from it with a step d of not more than 1/4 of the sound wavelength λ of the portion used by the receiving antennas in the mid-frequency sound range it is 50-1000 Hz, and the distance L between the centers of the receiving antennas is 10-50 m, the received signal from each receiving antenna is digitized with a sampling frequency

and the signals recorded by each k-th sensor of each receiving antenna are placed in the memory of the computing device in the form of the k-th digital array, for each sensor, the duration of the complete signal implementation is divided into N equal time intervals of duration τ, the vehicle is identified by the observable in it the amplitude spectrum of the behavior of the maxima in the frequency band ΔF characteristic of the vehicle, determine the direction of movement of the vehicle and its location the receiving antenna according to the sign of the slope of the dependence of the delay time of the correlation maximum of convolution of arrays corresponding to individual sections of the complete signal implementation g _kn = g _kn (t) from more and more distributed pairs of sensors in one receiving antenna, and by the dynamics of its inclination when moving from the current n to the subsequent n + 1-th time interval, determine the azimuthal direction to the vehicle relative to the center of one of the receiving antennas as equal to 90 ° and fix the location of the vehicle on the beam of this reception antenna at the time the angle of inclination of the dependence of the convolution maximum convolution maximum vanishes, take the autocorrelation function of the total data array from both receiving antennas, followed by summing over all elements of each of the receiving antennas and determine the vehicle speed as the ratio of the distance between the centers of the receiving antennas to the interval time between characteristic maxima in the autocorrelation function.

Images