JPS5835422A - Device for recognizing moving direction of moving sound source - Google Patents

Abstract

PURPOSE:To ensure the recognition of the moving direction, by providing a pair of microphones in parallel with the moving direction of the moving sound source, moving the maximal position of the mutual correlation function curve of the detected outputs to the direction corresponding to the moving direction of the moving sound source. CONSTITUTION:The noise detected signals LA and LB from the microphones 3A and 3B are imparted to filters 11A and 11B so as to pass only the frequency band which largely includes the noise part of an aircraft 1 to be measured and to remove the high frequency component which gives adverse effects when A/D converter circuits 12A and 12B are variably operated. The A/D converter circuits 12A and 12B send out digital waveforms signals S1A and S1B so that the positive part of the noise waveform of the aircraft 1 is a logic ''1'' and the negative part thereof is a logic ''0''. The output S1A is stored in a mutual correlation function computing circuit 13 for the time period required for the computation. The output S2A is imparted to the computing circuit 13 through a delay circuit 14, and the mutual correlation function is computed.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus for identifying the direction of movement of a moving sound source, and is intended to identify the direction of movement of a moving sound source by detecting noise generated by, for example, an aircraft, a train of automobiles, etc. using a microphone.

For example, due to the decline in environmental measures around airports, pigeon lofts that monitor aircraft noise over a long period of time, and noise generated by aircraft arriving and departing from airports are being investigated.

Recording sound source information such as noise level, duration, flight direction, etc. is a good way to understand the operational status of an airport! P1!
It is believed that However, in general, the sources of sound around airports are not only aircraft arriving and departing, but also the noise generated by various facilities attached to the airport, as well as the noise from vehicles attending the airport, such as cars and various work vehicles, arriving at random.This is the object of measurement. It is detected by a microphone superimposed on the noise of a certain aircraft.

In this way, conventional moving sound source identification devices have been developed to identify moving sound sources to be measured that are superimposed before arriving from sound sources other than the fixed target. Nothing discernible was proposed.

In consideration of the above points, the present invention attempts to obtain a moving direction edge JIllIIL amount for identifying the moving direction of a moving sound source of an aircraft. Detection - By calculating the maximum position of the cross-correlation function curve at the interval, data representing the state in which the maximum position moves over time is obtained, and based on this data, the moving direction of the moving sound source is determined. This is what I did.

The present invention will be described below in detail with reference to Figure mtc as an embodiment applied to identifying the moving direction of an aircraft.

First of all, let's talk about the position in Figure 1 to Figure 1.
kS moves. On the ground 1 is the moving direction K of the aircraft l? E)
-z A pair of microphones JAl and JB are installed laterally at a predetermined distance d as a dynamic sound detector to detect the noise of Aircraft 9. In practice, the microphones 3 and 7B are placed at predetermined monitoring points on the arc length of the airport runway, parallel to the extension.

For example, if the point M at the center of the microphone comb and JB is the installation point of the moving sound detection aS, the position? .

If the angle of incidence of noise arriving at this installation point M from an aircraft L located at It becomes k. Here, 6 is the speed of sound.

Therefore, by calculating the noise detection signals obtained from the microphones J and JBK, and #)111111111 as time t elapses, we obtain the delay time t of the above equation Q).
, a cross-correlation function curve φ1(τ) as shown in Jll(2), which takes a maximum value at the time T m t of VCsm and simply decreases before and after that, is produced. By the way, the first miter of the aircraft, which is similar to an aircraft, was at the time of JAIC.
Considering the noise sound wave that arrived at -O as a reference, this sound wave will arrive at the distant first microphone phon JIn at a future time τ-t, delayed by the delay time ta, so it will show the highest correlation at this point. It is from.

On the other hand, when the aircraft comes directly above the moving sound detection point 6 (0-de 0°), the first! When the reference noise sound wave arrives at the fourth microphone JB at time τ-0, the first microphone JBK also reaches the same samurai, so! The cross-correlation function curve φ, as shown in FIG.

When (τ) is drawn, it becomes k.

Furthermore, when the aircraft l passes the moving sound detection unit 6 and comes to the first microphone 31@, the first microphone on the side far from the aircraft l
The reference noise sound wave at microphone 3 is at time 0-0.
The first microphone of near Confucianism, when it arrived in? BK
When formulating according to the above equation (1), this noise sound wave had already passed at the past time τ-1, so the current time τ-0 as shown in Fig. 3(o)K. The first and first microphones 3 at a point in the past τ−−1°
The detection signals -1 and Motsu of System and JB show the highest correlation.

Therefore, among the noise coming from the microphone com and JBK, only the frequency components representing the acoustic characteristics present in the aircraft sound are picked up and the noise detection signals of the pair of microphones 3 and 7B are generated.

If we draw the phase 1 III number curve φ cod (τ) of and Ll according to the progression of WI # darkness, as shown in Figure 1, we can see that the aircraft l is at the position PHe P in Figure 1. If you move sequentially, this is Kl! Therefore, the mutual phase 1lIIn number curve is a.

? ,,? , from right to left in the order of [maximum value φ1 of each curve]. I&X aφ1 magazine X *φ1m5L
The small group moves from right to left. Conversely, if the aircraft l moves sequentially through the positions Asel't*I'x in Figure 1, the cross-correlation function curve will change to '1'HaA3 in Figure 1 accordingly. P.

Move from left to right in this order.

Therefore, in the present invention, the cross-correlation function l! To identify in which direction the position of the maximum value of l1II moves, and K
This is to identify the moving direction of the moving sound source of the aircraft cape.

Based on the above principle, the moving direction of the moving sound source is determined by the present invention as shown in FIG. 3.

1gt and Tateko's microphone com and J! IF)lI
The sound detection signal and IIB are filters and //B.
In addition to passing only the frequency band that strongly contains the sound component given to the measurement target, i.e., the aircraft If), the relatively Remove high frequency components. The KA/D conversion circuits 1 com and 1-B are digital waveform signals 81N in which the positive part of the noise waveform (consisting of an alternating current waveform) of the aircraft I is r/J, and the negative part is r/J. and BIB. In this example, ν
The sampling time of φ conversion circuit/Jmu and l-B is jtr4
#mk has been selected.

Here, the output B snow of the first toVD conversion circuit lcom is directly given to the mutual phase a@numerical calculation circuit/JK and accumulated for the time required for calculation, and the output B of the first toVD conversion circuit/com is The output of is applied to the cross-correlation function calculation circuit /Jk via the delay circuit /da, and thus the cross-correlation function is calculated.

There are 1M mutual phase WR11 number calculation circuits/3 (for example, W-S
For the sample point a), calculate the cross-correlation function between the stored data of the output of the conversion circuit 12 and the output currently being sent from the conversion circuit 12, and thus perform the calculation as described above with respect to FIG. , it is possible to obtain a cross-correlation function curve φ, (τ) that has a maximum value at the point in time when the output 81 of the conversion circuit 1 has the strongest correlation with the reference value of the conversion circuit 1. .

However, if a built-in delay circuit is inserted at the output end of the first conversion circuit/JOB, the position where the cross-correlation function curve φ(τ) is drawn will be the position shown in FIG. When there is an aircraft l at P, the maximum value will be at the center position of #tFK5
IIC and is drawn to the right of position 11 in the first @ or position I1 in FIG. Then it will be drawn to the left of that.

The calculation operation of the cross-correlation function calculation circuit 13 is repeated at a predetermined relatively large period (for example, a 2 second period).

A cross-correlation function calculation is performed for each calculation operation and KJli sample points, thus creating a cross-correlation function curve for each scheduled period.

Note that this cross-correlation function calculation circuit is
/Tei 00 Kude No. IIK, or the paper "One method of optimization of polarity correlation calculator" (Suzuki document, Sasaki Fumiyoshi,
Written by Hajime Nishimiya, Japan Acoustics ◆WII Collected Papers (October 1952)
Moon)) Kl! It can be composed based on the principles of a song.

When a cross-correlation function curve of 1 wA is created, this is the data switching circuit l! The cross-correlation function is temporarily stored in the cross-correlation function storage circuit l&. The data stored in this storage circuit 14 is given to the moving average circuit/l via the data switching circuit 13, and after smoothing the curve, it is transferred to the cross-correlation function storage circuit/l: be remembered. Here, the moving average means fiIU constant value! ,,! ,... are obtained one after another, take a fixed number of 1lK pieces starting from 6 piths [6, which refers to the entire arithmetic mean made] For example, when taking the average of 3 measured values, (X weight + X weight + Xs) / J, (x,
+x Hatake+14)/a, <x Zheng+X4+1g)/J...
・is the moving average.

In this way, as shown in FIG. 6, the data of sample time '11 * tg... (one each in the positive and negative directions) is sequentially stored in the memory circuit 16, and the entire stored data is stored in the direction of enlarging the 1-hour position. When viewed as a mutual phase gtuI that changes smoothly without any fine irregularities! It is designed to read and sheath several curved spears φ cod (τ). The reason why this is not done is because the phase 1III number is calculated for a relatively short period of time, which may cause local unevenness, and this unevenness must be removed.

Adjacent data of this smooth cross-correlation function curve ~(τ) is read out sequentially at each sample time, and the difference is calculated in the local maximum position detection circuit/1, and the sign of the difference is determined as positive or negative. be done. However, if the difference is positive, it means that the cross-correlation a - so φ, (T) is increasing, and on the other hand, if the difference 5) is negative, the cross-correlation function aS This means that (τ) is drawing a downward curve. Therefore, when the sign of the difference changes from positive to negative9, the cross-correlation function **φ, (T) reaches the maximum value φwr k x at this sample point.
When it is determined that K has been reached, the contents at the sample time and the contents of the difference are provided to the threshold circuit/l#lc.

This threshold is the path! When the content of the difference is larger than the predetermined threshold value, the content at the time of the maximum value WC@return is given to the SS direction as the maximum position data, and the content of the difference is When it is smaller than the threshold, k should not be given to
, (When a very small change occurs, it is ignored. - Direction of movement is certain I! Circuit X stores the maximum position data of the previous cross-correlation function curve φn(τ). The circuit 11k is stored, and this is compared with the maximum position data of the current correlation function!1141φ3 years old (τ), and if the content of the current maximum position data is smaller than the content of the previous maximum position data, the cross-correlation relationship - It is determined that the line has moved to the left in the figure 1 after the passage of time Km, and a leftward movement line change output 8t1 is given to the movement count circuit n.In contrast, the maximum position data of this time is If the content is larger than that of the previous maximum position data, the cross-correlation function *#
It is determined that 4m(τ) has moved to the right in the figure -)'C as time passes, and a rightward movement confirmation output f1mzt'$ is provided to the movement count circuit JJIC. At this time, the contents of the maximum position storage circuit ν are updated to the new maximum position.

The movement count circuit - is made up of all the intermittent circuits, and the movement confirmation output of movement direction line II + circuit X is 8 times.

If the direction output of the 8-electronic member matches the direction that has been counted up to the life, the counter will be operated to count +, but if the direction output does not match the direction that has been counted so far, the contents of the counter will be cleared. The new number of stitches is then added.

The count content output S of the movement number counting circuit is given to the movement direction number confirmation circuit JJk together with the direction confirmation IIalI force No. 8 of the movement direction number confirmation circuit. Definitely! The circuit N provides a moving direction identification output ss to the output circuit together with a direction confirmation output when the count of the counting circuit exceeds a predetermined number (for example, "J"). The output card is composed of, for example, a printer, a power output data register, etc., and stores the identification output S.

The above circuit elements exchange control signals with the control circuit, and send and receive control signals to the control Ia road meter! ! Next, we will deepen the familiarity function.

In the above configuration, the aircraft l is now in the center position from the right position PS of the 1st II! Suppose you are flying in the direction of . First, when the aircraft I is at position 11, the sound wave of the aircraft I reaches the first microphone JBi'C after a delay time t6 has elapsed after leaving the seventh microphone JAK. At this time, the cross-correlation function calculation circuit/J is at the Jll)k
As shown, a cross-correlation function curve φIt(τ) having a maximum value at a future time T-1 from time τ-O is generated, and after being moved averaged by a moving average circuit I, the cross-correlation function storage circuit 74
to be memorized.

In this way, one round of cross-correlation function curves φ and ri are obtained, and the sampling point of the maximum value is determined by the maximum position detection circuit/
1 is detected and the threshold circuit /? It is compared with the stored contents of the maximum position storage circuit -/ by the moving direction confirmation circuit X via the moving direction confirmation circuit X. However, at this time, the aircraft l is traveling from the right side to the left side, so the maximum position memory circuit ko/ contains the time point τ-t of the JvA (4), a point on the right side (its content becomes an even larger value). Since the maximum position of the K priest is memorized,
The movement direction confirmation circuit X confirms that the movement is to the left, and causes the movement number counting circuit to perform a counting operation.

As a result, when the movement count circuit counts the number rJJ, the flight direction count confirmation circuit judges this together with the left confirmation signal 8 of the movement direction confirmation circuit ” and gives an identification output BI to the output circuit Jk.

Here, when the difference calculated in the local maximum position detection circuit /l is small, it is judged as a disturbance and ignored by the threshold circuit /lK, and the movement direction confirmed by the movement direction confirmation circuit J5 is reversed. If so, move circuit counting circuit n
By clearing the count contents of , this is done to prevent the navigation of aircraft 1 from being erroneously recorded in output simplification:W.

In the above description, the present invention is applied to a mobile airborne sound source, but it can also be applied to various types of mobile sound sources such as a single train.

In addition, in the above embodiment, a pair of microphones were arranged parallel to the direction of movement of the aircraft, but in other words, even if the microphones are arranged in a horizontal direction including diagonally within a range that is not perpendicular to the direction of movement of the aircraft, the same effect can be achieved. The direction of movement can be identified. Incidentally, even if the microphone com and J
This is because a vector component parallel to the direction of arrangement of B can create an sx phase tone curve.

In addition, in the tg-th embodiment, 1 teg! is used as a separate circuit for each component! As described above, the functions of each component may be realized by using data arithmetic processing using a microcomputer. Ru.

That is, in step 8, the mutual phase am number is calculated and the data is transferred, and in the first step 8-, moving averaging is performed, and in the first step 8, the difference calculation (i.e., -th order differentiation) is performed. The local maximum position is detected by k, and in the next step 8, it is confirmed to which time position the maximum value has moved, and in the first step 8, it is determined whether or not it has moved based on the previous cycle. Ru.

As a result, if there is no movement, the aircraft is not flying, so it returns to station 118 and enters the next cycle.

If there is a move against 9, the next step is BT.

The time point of the maximum value is reached, and the next step is 8T.

It is determined whether the counted number has reached raJK or not, and if this number has not been reached, step 8 is followed by <14 to enter the next cycle. On the other hand, if the count exceeds rJJ, the next step st@Ice is performed and the moving direction identification output S!
, returns to Stella ζ8- and enters the next cycle.

Even in this case, the same function as described above with reference to FIG. 3 can be realized.

In the above, the cross-correlation lIl number curve curve 1 (') obtained from the cross-correlation function calculation circuit/JWC is used, and the unevenness of the line φn (') is removed by taking the moving average in the moving average circuit l for the IC. I'll do it 5kt but this is k
However, even if other smoothing circuits such as digital or analog filtering circuits are used, the same effect as in the above embodiment can be obtained.

As mentioned above, the main f! According to the method, a pair of microphones are arranged parallel to the direction of movement of a moving sound source, and the maximum position of the cross-correlation of the detected outputs moves in the direction of movement of the moving sound source. The moving direction of the moving sound source can be reliably identified.

[Brief explanation of drawings]

Figures 1 through 3 are illustrations of the principles of the present invention.
Fig. S is a block diagram showing an example of the movement direction identification device for a moving sound source according to the present invention, Fig. 6 is a curved diagram and Fig. γ for explaining its operation. 5 is a flowchart showing a processing procedure when the function of FIG. 5 is realized using a microcomputer. C... Aircraft, C... Ground, Com, 3B... Microphone, D... Moving sound detector, //A, //B...
Filter, l com, /! B...~Xi conversion circuit, 13.
・・Cross-correlation function calculation circuit, / reference ・・Slow grinding circuit, 15
...Data switching circuit, /4...Mutual phase ttin number storage circuit, 17...Moving average circuit, /1...Local maximum position detection circuit, 19...Threshold value circuit, I...Movement Direction confirmation circuit, U...Maximum position storage circuit, n...Movement number counting circuit, R...Movement direction number confirmation I! Circuit, scolding...
Output circuit. Applicant's agent Kiyoshi Inomata Figure 3 τ=-t2 τ=0 Continued from page 1 ■Applicant New Technology Development Corporation 2-5-2 Nagatacho, Chiyoda-ku, Tokyo

Claims (1)

[Claims] 1. Based on the detection outputs obtained from a pair of microphones arranged at a predetermined interval in the direction of lateral force, a predetermined time and Kl [order mutual 5III number - 1 are obtained. When each cross-correlation function ** is obtained, the movement direction of the maximum position is confirmed, and based on the result, the movement direction identification output for 11k before movement is obtained as 4IgIL. Movement of the moving sound source. Direction identification device. 1. Based on the detection outputs obtained from a pair of microphones arranged horizontally at a predetermined interval of 10"C, a cross-correlation function curve was obtained sequentially at a predetermined time. 1 is smoothed, the unevenness of the face on the curve is removed, and each of the smoothed mutual phases 1l
Once you have obtained several songs, check the direction of movement of the maximum position, and! 4111. A moving direction identification device for a moving sound source, which obtains a moving direction identification device for a moving sound source based on i&wrinkle confirmation. 3. Based on the detection outputs obtained from a pair of microphones arranged horizontally at a predetermined interval, a cross-correlation function curve is sequentially obtained at a predetermined time, and each cross-correlation - several songs fs is obtained. The moving direction of the maximum position is confirmed, and when the confirmed moving direction is the same as the previously confirmed moving direction, the number of movements is counted, and when the number of movements exceeds the planned number, the above-mentioned moving direction identification output V for the needle moving sound source is returned. A moving direction identification device for a moving sound source with 4I. 5. The moving direction identification device for a moving sound source according to claim 3, wherein the counted content of the number of movements is cleared when the moving direction of the moving sound source is different from the previously confirmed moving direction.