JP3040949B2 - Noise eliminator for pulse type cable detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a noise elimination device in a pulse type cable detection device for detecting a specific cable at an intermediate portion of a laying route from a plurality of laid cables.

[0002]

2. Description of the Related Art Hundreds to thousands of cables are laid in power plants, substations, and central control panels of plants.
These cables are connected to related electrical equipment at various places in the office. These cables have lengths ranging from tens to hundreds of meters per cable, and are laid through the wall or floor of the building at the entrance to or exit from the building or room. Is done. Therefore, it is very difficult to visually detect a specific cable from a large number of cables in the middle of the cable laying route.

[0003] When repairing electrical equipment in the central control panel or in various places in the plant, a specific cable is detected at an arbitrary intermediate portion other than the control room, and the cable is cut or reconnected in accordance with the repair content of the electrical device. The need to connect often arises.

[0004] Broadly speaking, two types of this type of cable detection device are known: an electromagnetic induction type and an electrostatic coupling type.

In an electromagnetic induction type cable detecting device, a sine wave voltage is applied from one end of a conductor of one cable to the ground from a transmitter, the other end of the conductor is grounded, and a sine wave current flows. To form a loop circuit. At this time, a magnetic field generated around the cable is detected by a search coil, and the detection output is measured by a receiver. A cable for which a measurement value having a significant difference from the detection output of another cable is obtained as the detection output at that time is identified as a cable for detection purpose.

Similarly, in an electromagnetic induction type cable detecting device, two cables are used, a sine wave voltage is applied from a transmitter between one ends of conductors of each cable, and the other ends of both conductors are connected to each other. To form a loop circuit in which the sine wave current circulates by shorting through the conductors. In this case, a magnetic field generated around each cable when a sine wave voltage is applied is detected by the search coil and the receiver. A cable for which a measured value having a significant difference from other cables is obtained as the detection output is identified as a cable for detection purpose. It is known that the detection output is displayed as a level using a bar graph, and a buzzer sounds when the level is equal to or higher than a predetermined value. A detector of this type is disclosed in, for example, Japanese Patent Application Laid-Open No. 57-18 / 1982.
No. 0304.

In an electrostatic coupling type cable detecting device, a sine wave voltage is applied from a transmitter between an outermost layer conductor at one end of one cable (a core wire when this does not exist) and the ground, and a conductor is applied. The other end is insulated from the ground (floating)
State. The electric field generated around the cable is measured by the receiver from above the outermost insulating layer of the cable via the measuring electrode, and as a result of the measurement, a measurement value having a significant difference from other cables is obtained. The identified cable is identified as a cable for detection purposes. A detection device of this type is known, for example, from Japanese Patent Application Laid-Open No. 7-270469.

In the above-mentioned conventional apparatus, since a sine wave voltage is used as a measurement signal, there is a limitation on a usable frequency bandwidth, and there is a problem of frequency resolution, so that a specific operation can be performed at the same time. The number of cables is limited. Furthermore, in the case of transmitting a sinusoidal voltage in the cable detection device of the electrostatic coupling type, in order to obtain the same electric field strength regardless of the cable length, energy or power that increases in accordance with the distribution capacity corresponding to the cable length is required. And

Therefore, the present applicant has proposed a pulse-type cable detecting device capable of performing a task of identifying a specific cable in a short time with high energy and with high accuracy even in the case of a large number of cables (Japanese Patent Application No. Hei 8-8). 95585).

[0010] This pulse type cable detection device is shown in FIG.
And as shown in FIG. 16, a transmitter 30 for injecting a pulse signal from a near-end injection unit into a specific cable, a probe 40 for detecting a pulse signal transmitted through the cable at an intermediate portion of the installation route by electrostatic coupling, A receiver main body for identifying a specific cable based on the detection output of the probe.

The transmitter 30 shown in FIG. 14 includes a power supply circuit 34, a code generator 36, and a switching circuit 38 housed in a case 32. The power supply circuit 34 receives a commercial power supply or a battery power supply, outputs a DC voltage of several volts to several tens of volts, and supplies the DC voltage to the code generator 36 and the switching circuit 38. The code generator 36 generates a unique code pattern preset for each cable to be inspected, and inputs the generated code pattern to a code input terminal of the switching circuit 38. The switching circuit 38 has a number of code input terminals corresponding to the number of cables to be inspected and an output terminal corresponding thereto, and based on each input code pattern and power supply voltage, changes a code pattern unique to each cable. The corresponding pulse voltages P1, P2, ..., P
N is sequentially and cyclically output to each output terminal in a time-division manner. Each pulse has a time width (pulse width) shorter than the propagation time.
In this case, the voltage may be about 0.1 to 10 μs (microsecond), and the voltage value is about several volts to several tens of volts. In principle, (2 ⁿ -1) types of code patterns can be generated according to the number n of bits of the code. For example, n = 8
Then, 255 kinds of code patterns can be generated, and up to 255 cables can be specified. FIG. 15 (a)
Shows an example of a code pattern in a case where all of 8 bits are "1", and FIG. 4B shows a case where the bit arrangement is "1, 0, 1, 1, 0, 1, 0, 1". Is shown.

Thus, even when a plurality of cables are to be inspected, a plurality of code pulses do not instantaneously overlap each other, and it is not necessary for the receiving side to synchronize with the transmitting side. The cable can be easily identified.

A receiver is constituted by the probe 40 and the receiver main body 50 shown in FIG. Probe 40
Is functionally comprised of a measuring electrode 42 for electrostatically coupling to a cable to be inspected, and a preamplifier 44 for amplifying an electric signal detected by the measuring electrode 42. The amplifier 44 is housed in a case 48 in which a shield 46 is stretched. The receiver main body 50 includes a power supply unit 52,
An attenuator 54, a main amplifier 56, a code pattern setting switch 58, a discriminator 60, and a level indicator 62 are provided. In general, it is preferable that the power supply unit 52 be a battery power supply because the receiver is used in a place where the cable is laid. A multi-core connection line 70 is provided between the probe 40 and the receiver main body 50, through which the output signal of the preamplifier 44 is input to the attenuator 54 and the power supply 52 of the main body 50 Operating power is supplied to the preamplifier 44 of the probe 40. Here, the output lines of the power supply unit 52 are not shown in order to avoid complicating the drawing. In the main body 50, the power supply unit 52 includes components such as an attenuator 54, a main amplifier 56,
0 and the level indicator 62 are also supplied with operating power. Here, the operating power supply lines are not shown.

The attenuator 54 includes, for example, 0, 6, 14, 2
The configuration is such that seven types of attenuation amounts of 0, 26, 34, and 40 dB can be switched and set, and are switched according to the input signal level. 40, 34, 26, depending on the input signal level
When the output of the discriminator 60 is changed to 20, 14, 6, 0 dB in order to indicate that the output coincides with a preset code, the level is detected as the amount of attenuation of the attenuator 54, and the level is detected by the level indicator 62. indicate.

In the receiver, the electric field generated by the traveling wave pulse of the cable is detected by electrostatic coupling using the measuring electrode 42 of the probe 40. The detection signal is supplied to the preamplifier 4
Amplified at 4 and guided to the receiver body 50. The detection signal guided to the receiver main body 50 is attenuator 54 and main amplifier 5
The measurement signal is input to the discriminator 60 via the signal generator 6. Here, when the input pulse is equal to or more than a predetermined value, the input pulse is set by the code pattern setting switch 58 and input as a reference signal.
It is determined whether or not it matches the code pattern of the pulse injected into the cable to be specified. The threshold value of the input level of the discriminator 60 is fixed, and the maximum attenuation that the discriminator 60 determines to be the same is displayed as the reception sensitivity by variously changing the attenuation of the attenuator 54.

The transmitter 30, the probe 40, and the receiver main body 50 configured as described above are connected to a cable as follows.

As shown in FIG. 17, the metal shield layer S, which is the outermost conductor of the cables 2A, 2B,..., 2N to be specified, is exposed at the injection end.
Then, a code pulse is injected with a unique code preset for each cable. In the case of a cable without a shielding layer, the core of the cable may be used instead of the metal shielding layer S. The low voltage side of both output terminals of the transmitter 30 and the case 32 are both grounded. If there is no shielded cable that runs partially or over the entire length, ground the cable to be specified instead.

In the receiver main body 50, a ground terminal is connected to a nearest ground terminal such as a ground terminal of a building. The measuring electrode 42 of the probe 40 is brought into contact with the cable surface made of an insulating layer, or the measuring electrode 42 is probed from a distance of several hundred mm in parallel with the cable surface. For all cables in the cable group including the cable to be specified, the reception sensitivity when the measurement electrode 42 of the probe 40 is brought close to the cable surface while maintaining the same distance and angle is higher than those of other cables. If it is significant and significant, it is determined that this is the cable to be specified. According to the experiment, significant difference determination was possible even when the measurement electrode 42 of the probe 40 was separated from the cable by about 200 mm.

The code pulse injected from the transmitter 30 propagates between the shield layer of the cable to be specified and the core wire or between the shield layers of the parallel cables. In the present invention, the propagation of the pulse is used, but the energy of the pulse does not need to consider the capacitance when the propagation medium is viewed as a lumped constant circuit, and the characteristic impedance when viewed as a distributed constant circuit (almost Tens of ohms) and a slight attenuation from the injection point to the specific operation point. Therefore, even in the case of a long cable, high-sensitivity measurement can be realized without lowering the sensitivity, so that it is not necessary to send out a high-energy signal which leads to a malfunction of the device.

[0020]

In the above-described pulse type cable detecting apparatus, a pulse signal encoded at a specific repetition frequency is injected into a cable, and the encoded pulse modulated signal is transmitted to a measuring electrode of a receiver. And demodulates it into a coded pulse signal on the transmitting side to identify the desired cable. In the case of a pulse signal, the signal to be propagated is less attenuated in accordance with the cable length than a method using a sine wave AC signal, so that it can be applied to a long distance cable without any problem. Another advantage of this pulsed cable detection device is that it can be applied regardless of whether the shield layer at the far end of the cable is grounded or ungrounded.

However, also in the case of the pulse type, when the shield 46 of the probe 40 is grounded, pulse noise invades from the ground point, and when the noise has a specific frequency component on the transmission side, the coded pulse signal May include pulse noise. As a result, the original signal pulse may be buried in the noise, or the noise pulse may be mixed in a place where there is no signal, so that the original code-modulated signal cannot be accurately demodulated. obtain.

Accordingly, the present invention provides a pulse-type cable detection device that satisfactorily removes noise mixed into a received signal on the receiving side, accurately demodulates the original code-modulated signal, and thereby performs accurate cable detection. It is an object of the present invention to provide a noise elimination device that can be used.

[0023]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention provides a method for detecting a specific cable at an intermediate portion of a laying route from a plurality of laying cables. The coded pulse signal is injected at the near-end injection section, the pulse signal propagated through the cable is detected by an electrostatically coupled probe at the middle of the installation route, and a specific cable is detected at the receiver body based on the detection output of the probe. In a noise elimination device in a pulse type cable detection device for identifying a narrow band filter that extracts only the same frequency component as the transmission pulse frequency of a transmitter from an output signal of a probe, a narrow band filter outputs a predetermined signal. A level comparator that converts the signal to a binary signal of “H” or “L” according to a threshold value, and a pulse missing in the output signal of the level comparator A digital noise filter for restoring and removing unnecessary pulses; and a signal / noise discriminating means for restoring an output signal of the digital noise filter to a pulse signal equivalent to a coded pulse signal on the transmission side. I do.

It is preferable that the digital noise filter executes a process of restoring missing pulses and removing unnecessary pulses in multiple stages.

When the digital noise filter is a multi-stage shift register and all three shift registers before and after the position to be corrected are "H" in order to correct missing one pulse, the shift register at the position to be corrected is forcibly set to "H". H "
If all four stages of shift registers before and after the correction target location are “H” to correct the missing of two pulses, the shift register at the correction target location is forced to “H”, and the correction target is corrected for non-continuous pulse correction. When the shift registers at the seven stages before and after the location are not all “H”, a correction means for forcibly setting the shift register at the location to be corrected to “L” can be included.

It is preferable that a buffer amplifier is provided before the narrow band filter.

The transmitter emits a coded modulation pulse by repeatedly turning on and off in a predetermined time unit.
The signal / noise discriminating means starts reading the input signal at a predetermined timing. When the number of pulses calculated from the cycle of the repetitive pulse in the unit output signal section of the transmitter is substantially a majority, the signal / noise discriminating means determines that the section is the ON section. If it does not reach the majority, it can be determined that the section is an off section.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The noise eliminator of the present invention is used on the receiver side, for example, by being inserted between a probe 40 and a receiver main body 50 in FIG.

FIG. 1 is a block diagram showing a noise removing apparatus 10 according to the present invention. As shown, a buffer amplifier 12, a narrow band filter 14, a level comparator 1
6, a digital noise filter 18 and a signal / noise discriminating unit 20.

The denoising is substantially performed by the blocks 14, 1
It is performed in three stages of 6,18. A detection signal mixed with noise or not mixed with noise detected by the probe 40 is used as a buffer amplifier 12 as a first stage of noise removal.
Through the narrow band filter 14, where the transmitter 3
Only the same frequency components as the transmission pulse frequency from 0 are extracted. By using this narrow band filter 14, of the countless frequency components that can be included as noise,
Frequency components other than the transmission pulse frequency can be removed. As a second step, the original pulse is returned by the level comparator 16. The pulse signal at this time is a signal mixed with unnecessary noise pulses, and may not be comparable to the transmitted pulse of the transmitter 30 yet. In the last digital noise filter 18 in the third stage, digital noise removal for restoring missing pulses and removal of unnecessary pulses is performed in multiple stages, and restored to a pulse equivalent to a coded modulated pulse on the transmission side. Modified to demodulated pulse.

Hereinafter, the function of each section of the block will be described in more detail.

1. Output signal of the transmitter 30 As a premise of the following description, the transmission pulse of the transmitter 30 has a pulse width of 5 to 15 μs, a repetition frequency of 10 kHz, a repetition period of 100 μs, and a peak value of 5 to 40 V as shown in FIG.
As shown in FIG. 3, this is turned on / off in units of 10 ms and output as a coded modulation scheme pulse.
10 ms (corresponding to 100 pulses) in code units is a time width determined in consideration of the operation response speed of the narrow band filter 14. The pulse width is determined according to the cable length. For example, when the cable length is 100 m, the cable propagation speed of the transmitted signal is 150 to 200 m / μs in the coaxial mode in which the signal is applied between the core wire and the shielding layer, but between the shielding layer and the ground. In the earth mode where a signal is added, the installation of a tunnel is 150-1.
80 m / μs and 50-80 m / μs for conduit installation, the propagation time of the reflected pulse from the far end is 1 to 1 for sinus installation.
2μs, pipeline laying is 2-4μs, pulse width 1μs
In the following case, there is no influence of the reflected wave. That is, it is irrelevant whether the far end is open or grounded. However, if the pulse width is too short, the energy of the transmitted pulse will decrease,
Since the pulse does not propagate over a long distance, the pulse width is set to 5 to 15 μs as illustrated in FIG. 3, for example, 5 μs, 10 μs, or 15 μs.

2. 4. Detection Signal of Probe 40 The detection signal taken out by the probe 40 is a 10 kHz coded signal mixed with pulse noise as shown in FIG. The reception signal detected at this time is about 2 V at a transmission output of 20 V.
A noise component reaching as much as 3 V is mixed. That is, the signal may be buried in the noise as shown in FIG. 4A depending on the ground conditions, and as shown in FIG. I do.

3. The output signals of the buffer amplifier 12, the narrow band filter 14, and the level comparator 16 The signal amplified by the buffer amplifier 12 and filtered through the narrow band filter 14 is affected by the reception sensitivity, the ground noise, and the like. 5) and a case with a lot of noise as shown in FIG. 5B. Here, the ground noise has little effect on the effective signal when the noise is small as shown in FIG. 6A, but when there is a lot of noise as shown in FIG. Or as an unnecessary signal. The signals shown in FIGS. 5A and 5B are passed through the level comparator 16 so that the signals shown in FIGS.
As shown in (b), the magnitude of a certain predetermined level, that is, a “threshold”, for example, a predetermined peak value of 5
The digital signal is converted into a binary digital signal by turning on the portion above the 0% value and turning off the portion below the 0% value. FIG. 7 shows the signal waveform thus obtained. FIG. 7A shows a case where there is little noise, and a portion where a signal should exist is shaped into a normal pulse signal. FIG. 7B shows a case where there is a lot of noise, and unnecessary pulses are mixed in a portion where a normal pulse which should have a signal is missing and a portion where there is no signal. There are places. The output signal of the level comparator 16 is sent to the digital noise filter 18 at the next stage. As described above, as illustrated in FIG. 7, the digitized signal possibly includes an unnecessary signal due to noise or a signal that has been lost due to noise.

4. Operation of Digital Noise Filter 18 A fixed pulse group included in one unit output taken into the digital noise filter 18 from the level comparator 16 is originally composed of all “H” level or “L” level pulses. Should be there. The digital noise filter 18 receives the digital noise filter 1 from the level comparator 16 based on this concept.
The unit pulse to be corrected to the “H” level and the unit pulse to be corrected to the “L” level and the unit pulse to be corrected are determined based on the arrangement relationship of the “H” level and the “L” level pulses of the entire pulse group taken into 8. The level is corrected based on the result of the determination.

The digital noise filter 18 includes a CPU, takes in the output pulse of the level comparator 16, and performs the above filter processing by software.
The filter processing means can switch the correction accuracy to four steps by switching the number of repetitions of the filter processing to, for example, four steps. Generally, the greater the number of repetitions, the higher the degree of averaging and the higher the correction accuracy.

The output signal of the level comparator 16 is composed of 100 pulses (10 ms) per unit corresponding to the output signal of the transmitter 30. Therefore, regardless of the valid signal existence section in the ON section or the non-existence section of the valid signal in the OFF section, the ON section (indicated as “H” in the figure) or the OFF section for at least 10 ms (for 100 shots). (Indicated as “L” in the figure). The digital noise filter 18 performs an additional correction for a portion lost due to noise in an extremely fine signal of a valid signal or an invalid signal. This correction is performed not in signal units (10 ms) but in pulse units (100 μs). The method of this correction will be described below.

As the pulse correction method, 1) one pulse missing correction, 2) two pulse missing correction, and 3) non-continuous pulse additional correction are executed. These corrections are performed in a multi-stage shift register included in the digital noise filter 18.

(1) One-Pulse Missing Correction As shown in FIG. 8, for example, a shift register SR4 is focused on as a shift register of a correction target portion specified in advance, and three shift registers SR1 to SR1 each including the shift register SR1 are included.
SR7 is to be checked, only the shift register SR4 is at the “L” level, and the shift registers SR1 to SR1
Only when all three and SR5 to SR7 are "H", the shift register SR4 at the correction target portion is determined to be missing due to noise, and is forcibly corrected to "H". No correction is made for conditions other than the above. Hereinafter, the same correction processing is performed on the data of the correction target portion specified in advance.

(2) Two-pulse missing correction As shown in FIG. 9, two consecutive shift registers, for example, shift registers SR5 and SR6 are focused on as shift registers of a previously specified correction target portion, and the preceding and following shift registers are included. The four-stage shift registers SR1 to SR10 are to be checked, only the shift registers SR5 and SR6 are at the “L” level, and the shift registers SR1 to SR4
Only when all of SR7 to SR10 are "H", the shift registers SR5 and SR6 at the correction target portion are determined to be missing due to noise and are forcibly corrected to "H". No correction is made for conditions other than the above. Hereinafter, the same correction processing is performed on the data of the correction target portion specified in advance.

(3) Non-continuous pulse additional correction As shown in FIG. 10, when paying attention to the shift register SR8 as the shift register of the correction target portion, the shift registers SR1 to SR7 and SR9 to SR1 at the seven stages before and after the shift register SR8, respectively.
If all 5 are "L", the shift register SR8 at the correction target location is forcibly set to "L". Otherwise, no correction is made. Hereinafter, the same correction processing is performed on the data of the correction target portion specified in advance.

By repeating the above processes (1) to (3) about four times, a considerably high level averaging process can be realized. The data obtained as a result is output as an output signal of the digital noise filter 18. In that case, as a result, the output signal of the noise filter 18 has a phase delay of 100 μs × the number of shift register stages with respect to the input signal. However, since there is no effect on the detection accuracy of the number of pulses and the like, even if a phase delay occurs, there is no problem in cable detection.

5. Operation of Signal / Noise Discrimination Unit 20 The operation of the discrimination unit 20 is also executed by the microcomputer, and counts the number of pulses existing in a signal unit (10 ms) to determine the H / L level in a pulse signal unit. As shown in FIG. 11, signals A, B,
When C is input and the signal is read every 10 ms, if a portion where the signal is lost due to instantaneous noise is read, a misreading of the signal may occur. To cope with this, the determination unit 20 determines the H / L level of the signal based on the number of pulses during 10 ms. However, since the input signal and the discrimination circuit 20 cannot be synchronized, a reading error may occur to some extent. The following processing is performed as a countermeasure for this.

The read timing of the reception pulse is as follows.
Start after about 10 continuous pulses, and as shown in FIGS. 12 and 13, 10 ms corresponding to 100 pulses
If there are more than 100 pulses, for example, 55 or more pulses in between, it is processed as level "H", and if less than 55, it is processed as level "L". By doing so, the opposite decision is made between the first 10 ms section and the last 10 ms section, so that it is possible to restore the transmitted signal as an accurate transmitted signal very close to the transmitted signal from the transmitter 30 side. Although the signal at this time has been subjected to noise removal processing, even if noise is mixed as shown in FIG. 13, this determination processing makes it possible to obtain a highly accurate received signal which is very similar to the transmitted signal. Can be restored.

The output signal of the signal / noise discrimination section 20 obtained as described above, that is, the output signal of the noise removing device 10 is input to the receiver main body 50 for cable detection.

[0046]

According to the present invention, in a pulse type cable detecting apparatus, noise mixed into a received signal is satisfactorily removed on the receiving side, and an original code-modulated signal is accurately demodulated, whereby an accurate cable is obtained. Detection can be performed.

[Brief description of the drawings]

FIG. 1 is a block diagram of a noise removing device according to the present invention.

FIG. 2 is a diagram showing an example of a pulse signal of a transmitter.

FIG. 3 is a diagram showing a modification of the signal output cycle of the transmitter and the pulse width of each pulse signal.

FIG. 4 is a diagram showing an example of a detection signal by a probe.

FIG. 5 is a diagram exemplifying a difference in an output signal of a narrow band filter due to the degree of noise.

FIG. 6 is a view for explaining the influence of a ground noise on an effective signal.

FIG. 7 is a diagram exemplifying a difference in an output signal of a level comparator depending on the presence or absence of noise.

FIG. 8 is a view for explaining a first example of pulse missing correction by a digital noise filter.

FIG. 9 is a view for explaining a second example of pulse missing correction using a digital noise filter.

FIG. 10 is a view for explaining a third example of pulse missing correction using a digital noise filter.

FIG. 11 is a diagram illustrating an on / off state of an output signal of a transmitter.

FIG. 12 is a view for explaining the operation of a digital noise filter when there is no noise.

FIG. 13 is a view for explaining the operation of a digital noise filter in the presence of noise.

FIG. 14 is a block diagram of a transmitter according to the prior art.

FIG. 15 is an explanatory diagram showing an example of a code pattern generated in the transmitter of FIG. 1;

FIG. 16 is a block diagram of a receiver used in combination with the transmitter of FIG. 1;

FIG. 17 is a diagram for explaining how to use a transmitter and a receiver according to the prior art.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Noise removal apparatus 12 Buffer amplifier 14 Narrow band filter 16 Level comparator 18 Digital noise filter 20 Signal / noise discrimination part 30 Transmitter 40 Probe 42 Measurement electrode 44 Preamplifier

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-57-180304 (JP, A) JP-A-6-91425 (JP, A) JP-A-7-270469 (JP, A) JP-A 9-Japanese 145769 (JP, A) JP-A-9-281175 (JP, A) JP-A-52-82066 (JP, A) JP-A-60-121812 (JP, A) JP-A-4-91514 (JP, A) Hikaru 4-59655 (JP, U) (58) Fields investigated (Int. Cl. ⁷ , DB name) H02G 1/00-1/10 G01R 31/02

Claims (5)

(57) [Claims] In order to detect a specific cable at an intermediate portion of a laying route from a plurality of laying cables, a pulse signal encoded by a near-end injection unit is injected from a transmitter into the specific cable. Noise removal in a pulse-type cable detection device that detects a pulse signal propagated through a cable at an intermediate portion of the installation route by an electrostatic coupling type probe and identifies the specific cable in a receiver main body based on a detection output of the probe. In the apparatus, a narrow-band filter that extracts only the same frequency component as the transmission pulse frequency of the transmitter from the output signal of the probe, and the output signal of the narrow-band filter is set to “H” or A level comparator that converts the signal into a binary signal of “L”; A pulse comprising: a digital noise filter that removes noise; and a signal / noise discriminating unit that restores an output signal of the digital noise filter to a pulse signal equivalent to a coded pulse signal on the transmission side. A noise elimination device in a cable detection device. 2. The noise eliminator according to claim 1, wherein said digital noise filter executes a process of restoring a missing pulse and removing an unnecessary pulse in multiple stages. 3. The digital noise filter comprises a multi-stage shift register, and when all three shift registers before and after the correction target location are "H" for correcting one pulse missing, the shift register at the correction target location is changed to "H". When the shift registers at the four stages before and after the correction target location are all "H" to correct the lack of two pulses, the shift register at the correction target location is forcibly set to "H".
When the shift registers at the seven stages before and after the correction target portion are not all "H" for non-continuous pulse correction, a correction means for forcibly setting the shift register at the correction target portion to "L" is included. The noise removing device according to claim 1, wherein 4. The noise eliminator according to claim 1, further comprising a buffer amplifier preceding said narrow band filter. 5. The transmitter according to claim 1, wherein the transmitter repeats ON and OFF in a predetermined time unit to transmit a coded modulation pulse, and the signal / noise discriminating means starts reading an input signal at a predetermined timing. When the number of pulses calculated from the cycle of the repetitive pulse in the unit output signal section of the transmitter is substantially a majority, it is determined to be an ON section, and when it does not reach the majority, it is determined to be an OFF section. The noise removing device according to claim 1.