JP2008051564A - Position measuring device - Google Patents

Abstract

An object of the present invention is to provide a technique capable of measuring the position of a moving body with high accuracy with a simple configuration.
In order to solve the above-described object, a position measuring apparatus according to the present invention includes a microphone unit 2 provided on a moving body, a plurality of speaker units 1, and M of different audible ranges for each speaker unit 1. A sound source data supply device that supplies sequence sound source data and a synchronization signal generator, and a synchronization signal is transmitted from the synchronization signal generator to the microphone unit and the sound source data supply device in response to a synchronization signal request signal transmitted from the microphone unit. The signal is transmitted to the speaker unit via the power line 6, and the position of the moving body is measured based on the transmission time difference of the sound from the speaker unit 1 to the microphone unit 2.
[Selection] Figure 15

Description

The present invention relates to a technique for measuring the position of a moving object using audible sound.

In recent years, a lot of research has been conducted on robots (home robots) whose home is an active area. Home robots are required to have advanced obstacle avoidance ability and flexible trajectory planning ability in order to be able to work in human living spaces. In order to realize these, it is necessary for the robot to grasp its own position with high accuracy (considered to be on the order of several centimeters).

A typical system using ultrasonic waves is called an active bat, and an accuracy of about 3 cm 2 is put into practical use using a 40 kHz ultrasonic signal (Non-patent Document 1). However, the accuracy of ultrasonic waves deteriorates because the time of flight of reflected waves is measured in an environment where there are many obstacles and strong environmental noises, such as in a home where the straight wave is strong. As a countermeasure, it is necessary to install a large number of receivers on the ceiling, which does not meet the object of the present invention. On the other hand, in the position detection using audible sound, even if there is an obstacle, the diffracted wave can be used, so the degree of accuracy reduction is reduced. In addition, since it diffracts, it is difficult to become a hidden terminal, and there is a high possibility that a small number of sensors (speakers) are required. If the sound speed is 343 m / s, if it is sampled at 10 kHz, detection is ideally possible with a resolution of 3.43 cm, and high-precision detection can be realized with less expensive hardware than using RF signals. On the other hand, when the distance is measured by diffraction, a position different from the true coordinate is detected as a result, and there is a problem that the detection result needs to be verified. Girod has proposed to spread the spectrum of audible sound against accuracy degradation due to obstacles and environmental noise (Non-Patent Document 2). However, the measurement range is limited to one dimension, and there are comments on the environment including obstacles, but the measurement results are not shown.
In addition, in order to measure the time of flight of the sound wave, the receiver needs to know the time when the sound wave was transmitted accurately (referred to as a synchronization problem). Although the present invention transmits synchronization signals by radio waves in the same manner as Girod's system of Non-Patent Document 2, in order to cover radio waves all over the home, it is necessary to install a large number of transmission antennas.

Therefore, the inventor has proposed to perform two-dimensional position measurement using M-sequence sound source data in the audible range (Non-Patent Document 3).
Nissanka B. Priyantha, Anit Chakraborty, and Hari Balakrishnan, The Cricket Location-Support System, 6th ACM International Conference on Mobile Computing and Networking (ACM MOBICOM), Boston, MA, pp. 32≡43, August 2000. Lewis Girod, Development and Characterization of an Acoustic Rangefinder, Tech.rep.TR-00-728, Information Sciences Institute (ISI), the University of Southern California (USC), March 2000. Naoyuki Takeichi, Shinji Tsuzuki, Michihiro Imaoka, Yoshiro Yamada, “Examination of high-accuracy position detection method using audible sound DS-CDM”, 2005 Annual Conference of the Shikoku Branch of the Electrical Engineering Society, pp.88, 2005

In the inventions of Non-Patent Document 1 and Non-Patent Document 2, the position of the moving body cannot be two-dimensionally measured at a place where there is an obstacle. An object of this invention is to provide the technique which can measure the position of a moving body with high precision by simple structure. When the distance is measured by diffraction, a position different from the true coordinate is detected as a result, so a method for verifying the detection result will be clarified. Clarify how to easily extend the transmission range of synchronization signals by radio waves.

In order to solve the above-described object, a position measurement apparatus of the present invention is a sound source that supplies a microphone unit provided in a moving body, a plurality of speaker units, and M-sequence sound source data in different audible ranges to each speaker unit. A power supply line having a data supply device and a synchronization signal generator, and transmitting a synchronization signal from the synchronization signal generator to the speaker unit via the microphone unit and the sound source data supply device in response to a synchronization signal request signal transmitted from the microphone unit The position of the moving body is measured by the time difference of flight of sound from the speaker to the microphone.
For the error test, we examined a method for measuring the degree of dispersion of the autocorrelation peak error in the audible high frequency region and a method for measuring the synchronous reception energy.

Further, in addition to the above configuration, a power line antenna that transmits a synchronization signal from the synchronization signal generation device to the microphone unit is provided in the synchronization signal generation device, and this power line antenna uses 1 / of the transmission wavelength used for transmission. You may comprise by providing the shield of the length of 2 or its integral multiple in the power line of a synchronous signal generator. Further, by providing an LC resonance circuit in the synchronization signal generator, the electrical length may be adjusted so that the actual length of the shielded power line is ½ of the transmission wavelength or an integral multiple thereof.

As described above, the position measuring apparatus of the present invention has an effect that the position of the moving body can be measured with high accuracy with a simple configuration and can be used even in a place with a relatively small obstacle. By verifying the measurement result, when there is a relatively large obstacle, it is possible to detect that the measurement result has a large error. Moreover, it has the effect that the reach | attainment range of a synchronizing signal can be easily extended by using an existing power line as an antenna.

The best mode for carrying out the present invention will be described with reference to the drawings. FIG. 1 is an explanatory diagram showing the principle of one-dimensional position measurement, and FIG. 2 is an explanatory diagram showing the principle of two-dimensional position measurement.

In the one-dimensional position measurement, as shown in FIG. 1, the measurement system is composed of two PCs (PC1, PC2) using Windows (registered trademark) connected to a speaker 1 and a microphone 2, respectively. The equipment used here will be specifically described. The PC sound card 3 was Creative SoundBlaster Live! Value, the speaker was Advent S4997, and the microphone was Audio-Technica AT-VD3, all of which were inexpensive. A / D conversion and D / A conversion for sound wave input / output were performed at a sampling frequency of 48 kHz and a sampling bit of 16 bits. We prepared multiple 1023chip M series with chip-rate 8kHz as sound source data, and output them from the speaker through BPF of 100 Hz to 8 kHz respectively. The BPF low-frequency cutoff frequency and chip-rate are values that take into account the frequency characteristics (80 Hz to 13 kHz) of the microphone used. The number of M sequences to be prepared is (number of speakers to be used + 1). Note that each sequence can be identified even if it is output from the speakers at the same time. Can also be applied. In two-dimensional and higher measurements, this system is a kind of synchronous DS-CDM (Direct-Sequence Code Division Multiplex), because different M-sequences are output simultaneously from multiple speakers and those sound waves are received simultaneously by a single microphone. I can say that.

As shown in Fig. 1, for one-dimensional measurement, two types of M series are simultaneously output in stereo from the sound card 3 of PC1 (PC 4), and the trigger sound remains the electrical signal of the stereo terminal input of PC2 (PC 5). Directly input to one side. By comparing the autocorrelation peak position of the sound recorded by the microphone installed at a distance of 1 m and the autocorrelation peak position of the trigger sound, the flight time t _f of the sound wave is obtained. Where v _s = 331: 5 + 0: 61C (C: room temperature [° C]), the measured propagation distance lm can be calculated by equation (1).
lm = t _f × v _s + α (1)
Here, it is assumed that the propagation delay time of the trigger sound is sufficiently smaller than the distance measurement sound flying in the air. Α is an offset value, which is a distance corresponding to an electrical delay time between the speaker and the microphone. In Fig. 2, α was calibrated at the point of l = 100 cm. In this example, since the sound source point and the sound collection point are determined visually, the installation accuracy of l is 1 cm or less. The microphone position where the measurement was performed is in the range of l = 10 to 240 cm.

Fig. 2 shows the method for two-dimensional measurement. If the M series for ranging is p (t); q (t), and the output amplitude coefficients from the speakers A and B (respectively 1a and 1b) are a and b, the ranging series output from the speakers A and B The sound is p _A (t) = a × p (t); q _B (t) = b × q (t). The trigger M sequence is r (t), and each received sequence is p _A '(t); q _B '(t); r '(t). The frequency characteristics of BPF are the same as in Fig. 1.

The M sequences p _A (t); q _B (t); r (t) of one cycle time Tm are simultaneously output from the speakers A and B arranged at intervals of 2 m as shown in FIG. 2 (FIG. 3). Calculate the sound flight time tf _A ; tf _B from each speaker from the difference between the autocorrelation peak position of p _A (t); q _B (t) and the autocorrelation peak position of r (t), and substitute it into equation (1) Then, the measured propagation distance l _mA ; l _mB from each speaker was obtained. Thereafter, the measurement coordinates (x _m ; y _m ) were obtained from the position of each known speaker and lm _A ; lm _B. Prior to measurement, the value of α was calibrated at the point of (x, y) = (100cm, 100cm), and the installation accuracy of (x, y) was 1cm or less, respectively, as shown in paragraph 0011, x = 0 to 200 The range of [cm], y = 10 to 220 [cm] was measured at 10 cm intervals.

Compared to FIG. 1, in addition to adding one speaker in FIG. 2, PC1 and PC2 are connected by Ethernet (registered trademark). Transmission power control is necessary as a countermeasure for the near-far problem, and an Ethernet connection is used to feed back the output amplitude coefficient a; b obtained by PC2 to PC1. Output amplitude coefficient is k (k = a or b), p _A (t); q _B (t) autocorrelation peak value is Radj (j) (j = A or B), and autocorrelation peak value of received signal is If Rrcv (j), k is expressed by the following equation.

A test to check the measurement accuracy when an obstacle was included (hereinafter referred to as case 2) was performed as follows. The following two methods were implemented to test the reliability of the measurement results.

つまり式(3) から算出されるスピーカＡ，Ｂ の自己相関ピーク位置誤差の分散をそれぞれv_A; v_B とする。 Although the audible sound reaches the microphone by diffraction even when an obstacle is present, it becomes difficult to diffract as the frequency increases. Using this property, we investigated whether a reliability test could be performed. Autocorrelation peak position for every m = (n + 1) bands, where the M-sequence chip-rate used is higher than the audible range and the audible outer band is divided into n and the total audible high frequency band is combined. Dh (i; j) is obtained (i = 1, 2, ..., m; j = A or B). If Dall (j) is the autocorrelation peak position obtained using the entire band including the audible range of the sequence output from speaker j, Dall (j) and the autocorrelation peak position in the audible outer range are The error variance vj can be calculated.

That is, the variances of the autocorrelation peak position errors of the speakers A and B calculated from the equation (3) are v _{A and} v _B , respectively.

A method for realizing the above-described index is shown in FIG. As shown in the figure, a cylinder with a radius of 9 cm and a height of 81 cm (corresponding to the thickness of a human foot) at the location of (x; y) = (100 cm; 100 cm) and (x, y) = (150 cm , 150cm) to (x, y) = (201cm, 150cm), a 51cm wide, 81cm high, 0.8cm deep plate was placed. In addition, zone A, B, and C in the figure are ranges where errors due to obstacles are expected to increase.
As in Case 1, the speaker and microphone were connected to two personal computers WindowsPC (PC1, PC2). However, the equipment used has been changed to support playback and recording of audible high-frequency sound. Sound Blaster Audigy4Digital Audio made by Creative was used as the sound card for the playback PC. This card is guaranteed to work with signals from 10Hz to 46kHz at a sampling frequency of 96kHz. The audio capture box for the recording PC is Elandolua-25 made by Roland. This box is guaranteed to work with signals between 20Hz and 40kHz at a sampling frequency of 96kHz. The speaker used was Sony SRS-88. The microphone used was SINGLE-DIAPHRAGM CONDENSER MICROPHONE B-5 (omnidirectional) manufactured by BEHRINGER. This microphone is guaranteed to work with signals from 20Hz to 20kHz. A / D conversion and D / A conversion for sound wave input / output were performed at a sampling frequency of 96 kHz and a sampling bit of 16 bits.

Prior to measurement, the transfer characteristics of these audio devices were measured. The sound characteristics, which were swept at a constant amplitude from 20Hz to 48kHz using the test signal generation software WaveGene with a single speaker, were recorded at a point 10cm away, and the transfer characteristics were measured using the spectrum analyzer software WaveSpectra. FIG. 5 shows the measurement results. In the figure, the upper part is the transmission characteristic of the sweep sound, and the lower part is the environmental noise level in the laboratory. From the figure, it can be seen that playback recording is possible up to around 35kHz. Based on these results, the M-series chip-rate used was set to 32 kHz to include the audible high frequency range. The sequence length is 1023 chips, which is the same as Case 1.

Calibration of α (x, y) = ( 100cm, 100cm) is performed in the absence of obstructions in, was placed a cylindrical after calibration. The measured plane with 2 × 2m ^2, lm on the plane at 10cm intervals _A and lm _B were measured and (xm, ym) was calculated. However, the y = 0cm line segment was excluded because the microphone sensitivity was poor.
Each of BPF1, 2, and 3 in Fig. 4 uses an FIR filter (Humming window, 51 taps) based on the window function method. The frequency characteristics of BPF1 and BPF2 are 100Hz to 32kHz. On the other hand, frequency division for detecting the autocorrelation peak position of sound waves outside the audible high frequency range is divided into n = 11 as shown in Table 1 in BPF3, and m = 12 (= n + Vj was calculated from equation (3) as 1).
The structure of the sound card output data series used is shown in Figs. 6 and 7. The measurement procedure is as follows.

Procedure 1 (Transmission power control)
As shown in FIG. 6, sound waves are simultaneously output from three channels, output amplitude coefficients a and b are obtained according to the equation (1), sound waves are output again, and the process is repeated until the received energy reaches a certain level.

Procedure 2 (distance measurement)
When step 1 (transmission power control) is completed, the measured propagation distances lm _A and lm _B from each speaker are calculated according to equation (1) to obtain (xm, ym).

(Synchronous reception)
According to the arrival times tf _A and tf _B from each speaker obtained by the distance measurement, the output time is adjusted as shown in FIG. 7 in order to simultaneously receive the sequences s _A (t) and s _B (t) at the location of the microphone. Steps 3a and 3b are performed.

Procedure 3a (Dispersion of autocorrelation peak error in audible high frequency range)
p _A (t), q _B (t), p _0A (t), and q _0B (t) are divided according to Table 1, and the variance of the autocorrelation peak error is obtained for each speaker from Equation (3).

Procedure 3b (sound wave synchronous reception energy)
The synchronous received energy Erj (x, y) of s _A (t) and s _B (t) is obtained from the equation (4). In steps 1, 2, and 3, s _A (t) and s _B (t) are not used, but are output from the speaker for simplicity. Similarly, in step 3b, p _A (t) and q _B (t) are not used but are output.

The measurement result will be described.
Coordinate error ev and distance data error e _A = | l _A −l _mA |, e _B = | l _B −l _mB |
8 to 10 show the distributions of points that cannot be measured due to obstacles.
In FIG. 8, the error caused by the obstacle is up to 3.6cm in zone1, up to 4.6cm in zone2, and up to 164cm in zone3. In FIG. 9, the error caused by the obstacle is a maximum of 4.6 cm in zone 4 and a maximum of 162 cm in zone 5. In FIG. 10, the error caused by the obstacle is a maximum of 4.6 cm in zone 6 and a maximum of 162 cm in zone 7.
In FIGS. 8 to 10, the maximum error occurrence point is (x, y) = (190 cm, 150 cm) immediately behind the plate-like obstacle. In addition, in zones 3, 5, and 7 behind the plate-like obstacle, the error is distributed from 7 to 164 cm, which shows that the influence on the obstacle is great. On the other hand, it can be seen that the errors of zones 1, 2, 4, and 6 corresponding to zones A and B (FIG. 4) under the influence of the cylinder are within the same error range as FIG. Therefore, in the case of obstacles in this test, it is necessary to detect that the reliability of the measurement results of zones 3, 5, and 7 corresponding to zone C (FIG. 4) is low. The results are shown below.

(Dispersion of autocorrelation peak position error in audible high frequency range)
11 and 12 show the variances v _A and v _B of the autocorrelation peak position error in the audible high frequency range. The distribution of the error variance v _A due to the obstacle from the speaker A (FIG. 11) was 2.15e + 04 cm ² at the maximum in zone 8 and 2.04e + 03 cm ² at the maximum in zone 9. The maximum error dispersion point is the back (x, y) = (190cm, 150cm) of the zone-like obstacle in zone 8, which is consistent with FIG. The variance distribution of error vB due to the obstacle from speaker B (Fig. 12) was 1.76e + 03cm ² at the maximum in zone 10, 3.20e + 03cm ² at the maximum in zone 11, and 2.08e + 03cm ² at the maximum in zone 12. The maximum error variance point is (x, y) = (190 cm, 160 cm) of zone 11 and is consistent with FIG.

Therefore, it can be seen that it is possible to detect that the reliability of the measurement result is low in zones 8 and 11 corresponding to zone 3 where the error is large in FIG. Corresponding to zone 1 in FIG. 8, zone 10 in FIG. 12 could be detected, but zone 2 around the cylinder could not be detected. It is considered that the accuracy deterioration due to the obstacle could not be detected because the diffraction of the audible sound and the high frequency range outside the audible range are the same. However, since zone 2 originally has an error range similar to that in FIG. 3, it was allowed here that it could not be detected in FIG. On the other hand, there are many points with high error variance in zones 9 and 12 where no obstacle exists. This is thought to be because the SNR (Signal to Noize Ratio) of the audible high-frequency component has decreased due to the directivity of the microphone-speaker, but can be excluded if used in combination with the following method.

(Synchronous reception energy)
FIG. 13 shows the relationship between the synchronous reception energy Er and the error ev. From the figure, it can be seen that there is a correlation between the received energy Er and the error ev. FIG. 14 shows a distribution of synchronous reception energy. It can be seen that the reception energy is significantly reduced by the obstacle in zone 13 which is the back side of the plate-like obstacle. The lowest point is (x, y) = (190 cm, 150 cm), which coincides with Fig. 8. When the threshold for judging whether the error is the same level as Fig. 3 is ev = 5cm, ev in Fig. 13 The maximum Er for> = 5cm was 0.40.
The measurement point of area1 where Er <0.4 and ev> 5 exists in zone13, and it is possible to detect a decrease in accuracy due to an obstacle, while Er2 is low in area2 where Er <0.4 and ev <5. Ev is small. This is considered to be because even if there is no obstacle, the received energy tends to decrease as the measurement distance increases. The influence of the directivity between the speaker and microphone (zone 9, 12) shown in the paragraph 0026 was not observed in FIG. Therefore, it can be said that the reliability of the zone in which both the autocorrelation peak position error in the high frequency range outside the audible range and the synchronous reception energy drop occurs is low.

Next, a method for exchanging synchronization signals using power lines will be described.
FIG. 15 is an explanatory diagram showing a position measuring device having a power line antenna. Similar to Girod's system in Non-Patent Document 2, audible sounds and radio waves are used. However, although the synchronization signal (M-seq. (0) in the figure) output from the synchronization signal generator uses radio waves for notification to mobiles, the radio wave radiation method is normal. The difference is that power lines are used as antennas instead of antennas. In the Girod system, the synchronization signal generator and the sound source data supply device are integrated, but in the present invention, they can be separated, and the synchronization signal (M-seq. (0) in the figure) is supplied to the sound source data supply device. This notification is different in that it is performed as a common mode signal passing through the power line.
In the example of FIG. 2 is a baseband signal M-seq. (0) is, PC (2) has been directly output from the multi-channel sound card, ASK (Amplitude Shift Keying in the carrier of FIG. 15, the angular velocity omega _c (Also called On-Off Keying (OOK)). The modulated signal is injected into the power line as a common mode signal via a circuit equivalent to an LCL (Longitudinal Conversion Loss) probe, and is thus radiated from the power line. Here, an antenna tuner composed of L (coil) C (capacitor) is used to adjust the electrical length of the AC cord to the half wavelength of the carrier wave. In this example, the case where the carrier frequency fc (= ω _c / 2π) is 25 MHz is tested, and the length of the AC code is 2 m. The circuit equivalent to the LCL probe is used as a mode isolator for common mode and differential mode.
The shielded AC cord has three cores (hot, cold, ground) and functions as a feeder line for the power line antenna. The common mode signal propagates through the shield line, and the connection point with the power line (VVF (Vinyl insulation and Vinyl sheath Flat) cable in FIG. 15, most commonly used as domestic wiring material in Japan) corresponds to the feeding point. . Compared to FIG. 2, although the method of transmitting the synchronization signal is different in FIG. 4, it is electrically transmitted and is sufficiently faster than the speed of sound, so there is no difference in distance measurement accuracy.

FIG. 16 shows the measurement result of the transfer function when a 2-core VVF cable (core wire diameter: 1.6 mm) having a total length of 14.6 m is wired on the ceiling of the test room. Line 1 in the figure is the “AC code-containing common mode”, and is the case where the AC code adjusted to resonate at 25 MHz by the antenna tuner of FIG. 15 is inserted on the transmission side.
Since the transfer function of the common mode is generally uncontrollable, it has not been actively used as a communication means in the past. However, it can be seen that if a shielded AC cord as shown in FIG. 15 is used, low-loss communication is possible in a desired frequency band.
FIG. 17 shows the excitation state of the power line antenna of the present invention. There are two types of power lines: VVF (2-core, 10m) and VCT-SB (2-core shielded vinyl cabtire cable, 1.25mm ² , 70m). Since the length of the shielded AC cord is 2 m, the feeding point is the point of l = 2 m in the figure. Regardless of the power line connected to the AC cord, it can be seen that the AC cord always functions as an antenna element that resonates at half the wavelength of the carrier wave. On the other hand, the power line on the right hand side of the feeding point can be seen to function as an antenna element, although the electrical length varies depending on the wire.

(Radiation characteristics of power line antenna)
FIG. 18 shows a measurement method for comparing the power line antenna cover area with a conventional antenna, and FIG. 19 shows the result.
FIG. 19 (a) shows the electric field strength at a point separated by a distance x along the VVF cable when a common mode signal is injected into the two-core VVF cable as shown in FIG. 18 (a). However, even if x changes, the distance to the nearest VVF cable is always 190 cm. On the other hand, in the case of the conventional antenna, as shown in FIG. 10 (b), the radio wave is assumed to be radiated from one loop antenna, and the electric field strength at a distance x is measured. Here, E (f, x) is defined as the electric field strength of frequency f at distance x [m], and relative strength Hr (f, x) with E (f, 1) is defined as follows.
H _r (f, x) = E (f, x) / E (f, 1) (4)
This relative intensity is shown in FIG. 19. In both cases, the relative intensity gradually decreases as x increases, but it can be seen that the power line antenna has a smaller degree. This is because the loss when the common mode signal propagates through the VVF cable is smaller than the loss when it propagates through the air as shown in FIG. Accordingly, it can be said that in a room where a large number of power lines are wired, the power line antenna is more likely to have a wider cover area than the conventional antenna.

Power line communication (PLC) A speaker with a built-in modem and the sound source data that you want to hear when it is plugged into the nearest power outlet is called a PLC speaker. In recent years, with the widespread use of DVDs and digital broadcast receivers, 5.1ch surround can be easily enjoyed at home, and there is a great need for wireless surround speakers. There is also a need to listen to content stored in a portable digital music player with the nearest speaker. PLC speakers are useful for these needs.
In FIG. 15, all the speakers are connected to the PC (1) with electric wires, but the present invention can also be implemented with a PLC speaker. An image of the above positioning system based on the use of the PLC speaker is shown in FIG.
Normally, a speaker is used for music streaming as described above, but when positioning is performed, an M-sequence signal for positioning is superimposed and output from the speaker in response to the request. The PLC speaker has a built-in sound source data supply device. What is important here is the synchronization of the speaker and the microphone. In order to make the error due to the synchronization error within 5 cm 2, it is necessary to synchronize between devices at 145 μsec (= 5 cm / vs, vs is the speed of sound) or less. It is difficult to guarantee this value with the CSMA type protocol often used in LAN. Therefore, a method for synchronizing each device with a commercial frequency is proposed. According to the CEPCA technical specifications, the zero cross point is displaced by the load of the commercial power supply, but the displacement per house is said to be 100 μsec or less at maximum, which is effective as a means to make the error due to synchronization deviation within 5 cm. It is.
The modulated M sequence (0) shown in FIG. 15 is transmitted as a common mode signal through the power line. During the transmission, it is also radiated into the air as radio waves, transmitted to the microphone as a synchronization signal, demodulated on the microphone side, and recorded as a time stamp sound. When the M-sequence (0) signal transmitted through the power line arrives at the sound source data supply device built in each speaker, the measurement M-sequence (1) (2) is output from the speaker. The microphone built in the moving body also records the incoming M-sequence (1) (2) sound. The delay time from the arrival of the M sequence (0) signal to the speaker output is expressed as a fixed time td.
Difference between delay time t_1 and t_2 from recording M sequence (0) sound to recording M sequence (1) and (2) sound t_1 ≡ td and t_2 ≡ td It is a flight time, and each flight distance can be calculated from a known sound speed. If the position of the speaker is known, the position of the microphone can also be obtained.

It is explanatory drawing which shows the principle of a one-dimensional position measurement. These are explanatory drawings showing the principle of two-dimensional position measurement. It is explanatory drawing which shows the output data series of a sound card. It is explanatory drawing which shows a two-dimensional position measuring apparatus. It is a graph which shows the transfer characteristic and environmental noise of an audio equipment. It is explanatory drawing which shows the output data series of a sound card, and shows the time of synchronous measurement. It is explanatory drawing which shows the output data series of a sound card, and shows the time of transmission power control and distance measurement. It is a graph which shows a 2-dimensional position measurement coordinate error vector. It is a graph which shows a measurement distance error (speaker A). It is a graph which shows a measurement distance error (speaker B). 6 is a graph showing a distribution of autocorrelation peak position error variance in an audible high frequency range (speaker A). It is a graph which shows the distribution of the autocorrelation peak position error dispersion | distribution in an audible high frequency range (speaker B). 5 is a graph showing the relationship between received energy Er and error vector ev of a sequence s (t) during synchronization. It is a graph which shows the received energy distribution of the series s (t) at the time of synchronization. It is explanatory drawing which shows the position measuring apparatus which has a power line antenna. It is a graph which shows the measurement result of the transfer function of a 2 core VVF cable. It is a graph which shows the excitation state of a power line antenna. It is explanatory drawing which shows the measuring method of the cover area of a power line antenna. It is a graph which shows the measurement result of the cover area of a power line antenna. It is a conceptual diagram which shows the positioning system containing a PLC speaker. It is a chart figure which shows the synchronization of the positioning system containing a PLC speaker.

Explanation of symbols

1. Speaker 2. 2. Microphone Sound card 4,5. computer

Claims (2)

A microphone unit provided on the moving body, a plurality of speaker units, a sound source data supply device that supplies M-sound source data in different audible ranges to each speaker unit, and a synchronization signal generator, and transmitted from the microphone unit In response to the synchronization signal request signal transmitted, the synchronization signal is transmitted from the sound source data supply device to the microphone unit and the speaker unit via the power line, and the position measurement is performed to measure the position of the moving body according to the sound transmission time from the speaker to the microphone. apparatus. A power line antenna that transmits a synchronization signal from the sound source data supply apparatus to the microphone unit is provided in the sound source data supply apparatus, and the power line antenna has a length that is 1/2 of the transmission wavelength used for transmission or an integral multiple of the transmission wavelength. The position measuring device according to claim 1, wherein the position measuring device is configured by providing a shield on a power line of the sound source data supply device.

Images