JP3264870B2 - Iron piece detection method and device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an iron piece detecting method and apparatus (needle detector), and particularly to various textile products, sewing clothing, and the like.
The present invention relates to an iron piece detecting method and apparatus effective for detecting iron pieces such as needles and nails mixed in carpets, textiles, felts, nonwoven fabrics, foodstuffs, medical goods, packed goods, and the like.

[0002]

2. Description of the Related Art For example, it is not unusual for a sewing machine needle to break during a sewing operation, and sometimes the user forgets to remove a waiting needle used during a sewing operation. However, if there is a remaining needle such as a broken needle or a waiting needle in the sewing clothing, the remaining needle may harm the user of the sewing clothing, so that the sewing clothing is inspected by an iron piece detecting device. In many cases, iron pieces are also detected for various fiber products, carpets, textiles, felts, nonwoven fabrics, food products, medical products, packed products, and the like.

[0003] Incidentally, iron pieces have been conventionally detected using a metal detector or an iron piece detector mainly for detecting a ferromagnetic member. For example, a direct current magnetic field generation using a permanent magnet is performed. And a sensor unit (a meter reading unit) comprising a detection coil having a winding in which a DC magnetic flux in the DC magnetic field generated by the DC magnetic field generation unit interlinks, and a detection signal generated in the detection coil. Is generally used to output an alarm when the threshold value exceeds a threshold. That is, when the sensor unit and the object to be detected are moved closer to each other and the object to be detected includes iron pieces, the number of magnetic fluxes linked to the detection coil is increased by the sensor unit in the object to be detected. And a detection signal is generated in the detection coil. When the detection signal exceeds the threshold, an alarm is output assuming that the detection target includes iron pieces.

In such an iron piece detecting device, it is assumed that iron pieces having a predetermined size or more are detected when a detection signal generated in a detection coil exceeds a threshold. However, in order to be able to detect the presence of an iron ball having a diameter of, for example, about 1.2 mm, a sensor having a DC magnetic field having a strong magnetic field strength is required as a sensor, and the iron ball is required to have a DC magnetic field. In order to amplify the small detection signal generated in the detection coil when moving in the
An amplifier with high amplification is needed. Therefore, there is a problem that the detection coil easily picks up a small external noise magnetic field and causes a malfunction. Further, since the detection signal has a frequency component corresponding to a relative moving speed between the sensor unit and the detection target, the detection signal is passed through a filter that allows the frequency component to pass, thereby preventing malfunction due to an external noise magnetic field. Can be considered. However, if the noise voltage generated in the detection coil by the external noise magnetic field is in the same frequency band as the frequency component, malfunction due to the external noise magnetic field cannot be prevented.

[0005] By the way, a garment manufacturer or apparel maker that manufactures various textile products and sewing clothing products sometimes uses various clothing accessories, but various textile products and sewing clothing products with the clothing accessories attached thereto. When using a shingle detection device to detect shingles on items, even in the absence of shingles in various textiles and garments, metal clothing accessories (buttons, fasteners, front Is often used, which may cause a malfunction.

[0006] For this reason, an iron piece detecting device maker, a garment maker, an apparel maker, and the like consult with each other, and for example, a broken needle of a sewing machine is converted into an iron ball having a diameter of 1.2 mm. 0.8mm maximum diameter
As a criterion for setting detection sensitivity, for example, a detection signal generated corresponding to an iron ball having a diameter of 1.2 mm indicates that iron pieces having a predetermined size or more when the threshold is exceeded are detected. In a state where the iron piece is detected, and a detection signal generated corresponding to an iron ball having a diameter of 0.8 mm, an iron piece having a predetermined size or less is not detected when the iron ball has a threshold or less.

However, when iron pieces are detected using such an iron piece detecting device, even if the sensor section and the object to be detected are relatively moved at a constant speed, the DC magnetic field generating member and the object to be detected are still in contact. When the distance to the moving position changes, the detection sensitivity of the iron piece changes due to the change in the magnitude and frequency component of the detection signal generated in the detection coil of the sensor unit,
The size of shingles cannot be determined.

In order to solve such a problem, the present applicant has proposed an iron piece detecting device as disclosed in Japanese Patent Application Laid-Open No. 7-260943. In this iron piece detecting device, FIG.
As shown in FIG. 3, the detection object Dx is moved at a constant speed by the transfer band 5, and the detection signals generated by the detection coils 3 and 4 which are arranged at a predetermined distance L are individually amplified. The digital signals are converted into digital signals, and the calculation of the cross-correlation of each digital signal is performed.
This will be described with reference to FIG. In FIG. 13, reference numeral 1 denotes a permanent magnet as a DC magnetic field generating member, on which a magnetic path 2 is formed.

In FIG. 14, X1 (t) represents the function of the signal generated in the detection coil 3 by the iron pieces, and nl (t) represents the function of the noise generated in the detection coil 3 by the elements other than the iron piece. And Y1 (t)
Is a functional representation of the output signal which is their sum. Similarly, X2 (t) is a function expression of a signal generated in the detection coil 4 by iron pieces, and n2 (t)
(T) is a function display of the noise generated in the detection coil 4 by other than the iron pieces, and Y2 (t) is a function display of the output signal which is the sum of them. And
τ is the time required for the detection target Dx to move the distance L between the detection coil 3 and the detection coil 4. Accordingly, the output signals Y1 (t) and Y2 (t) are given by the following equations 1 and 2, respectively.
Is represented by

[0010]

## EQU1 ## Y1 (t) = X1 (t) + n1 (t)

Y2 (t) = X2 (t) + n2 (t) In this case, the cross-correlation function of the output signals Y1 (t) and Y2 (t) is defined by the following equation (3).

[0011]

(Equation 3) Focusing on the first term following the integral symbol of the above equation 3, the function Xl of the signal generated in the detection coil 3 by the iron pieces
(T) and the function X2 (t) of the signal generated in the detection coil 4 by the iron pieces have the largest correlation with the delay time τ. If the term is sufficiently larger than the term containing, the term containing noise can be ignored, and the presence or absence of iron fragments can be detected based on the correlation value.

However, if the correlation value of the noise is not sufficiently small as compared with the correlation value of the signals generated in the detection coils 3 and 4 by the iron pieces, the iron piece detecting apparatus described above can provide good iron piece detection results. In addition, since the movement of the detection object Dx by the transfer zone 5 is limited to only the movement from the detection coil 3 to the detection coil 4, for example, the detection object Dx is reciprocated. Even if it is desired to perform the iron piece detecting operation in the state, there is a problem that the iron piece detecting operation cannot be performed in such a detection mode, and a solution has been demanded. Another problem is that the level of the detection signal must be constant.

[0013]

As a solution to the above problem, the present applicant has proposed the following apparatus (Japanese Patent Application No. Hei 8-24609).

That is, in FIG. 1, a permanent magnet 1 as a DC magnetic field generating member is arranged so as to generate a DC magnetic field H having a predetermined magnetic field strength in a space where the object Dx moves, and has a thickness. In order to form a DC magnetic field H in a space which is magnetized in the direction and faces one magnetic pole (N pole), the other magnetic pole part and a part excluding a part of the side surface are surrounded by a magnetic path 2. ing. In the DC magnetic field H formed in the space on the N pole side of the permanent magnet 1, two detection coils 3 and 4 are separated by a predetermined distance L in the moving direction (X direction) of the detection target Dx. And a transfer zone 5 is provided in a space between the detection coils 3 and 4 and the permanent magnet 1. The transfer belt 5 is stretched between a driving roller 7 driven by a driving motor 6 and a driven roller 8, and is driven at a predetermined speed V by the driving motor 6.

When the object Dx is placed on the transfer zone 5 traveling at a constant speed V in the X direction, the object Dx passes through the DC magnetic field H generated by the permanent magnet 1 at the moving speed V. I do. If the detection target Dx contains iron pieces, the detection target Dx is generated in the DC magnetic field H generated by the permanent magnet 1.
Move, the state of the magnetic flux distribution in the DC magnetic field H changes depending on the iron pieces in the object Dx to be detected, so that the detection coils 3 and 4 correspond to the change in the magnetic flux distribution in the DC magnetic field H. A detected signal is generated. Since the detection coils 3 and 4 are arranged at a distance L apart from each other and the detected object Dx runs at the same moving speed V as that of the transfer zone 5, two detection signals are output between the detection coils 3 and 4. Has a constant time difference T determined by the distance L of the detection coil 3 and the speed V of the detection target Dx passing between the detection coils 3 and 4.

The detection signal DT1 of the detection coil 3 is
0, the detection signal DT2 of the detection coil 4 is input to the amplifier 11, and the detection signals DT1 and DT2 amplified by the amplifiers 10 and 11 are subjected to A / D conversion through low-pass filters (LPF) 12 and 13, respectively. To the vessels 14 and 15. The digital signal output from the A / D converter 14 is stored in the memory 16 and the A / D converter 15
The digital signal output from is stored in the memory 17. 20 is, for example, a microprocessor, RAM, ROM
And a control unit 30 such as a digital signal processor (DSP).
Reference numeral denotes an output unit including an indicator light, a sound generator, and the like. The arithmetic unit 30 controls the memory 1 under the control of the control unit 20.
After calculating the difference value of the digital signal based on the detection signals DT1 and DT2 read out from 6 and 17, the calculation of the autocorrelation is performed on the difference value.

The detection signals DT1 and DT2 generated from the detection coils 3 and 4 include signals (detection targets) generated by iron pieces included in the detection target Dx moving at a constant speed by the transfer zone 5. Signal), noise generated by various causes such as vibration of the sensor, an external electromagnetic field and an external magnetic field is also included. The noise is divided into A (i) simultaneously generated in the detection coils 3 and 4 and other noise B (i), and a detection target signal generated in the detection coil 3 by iron pieces in the detection target Dx. To xl
(I), and if the detection target signal generated in the detection coil 4 by the iron pieces is x2 (i), the detection coils 3 and 4
Are generated by the following equations 4 and 5, respectively.

[0018]

## EQU4 ## Y1 (i) = x1 (i) + A (i) + B1 (i)

## EQU00005 ## Y2 (i) = x2 (i) + A (i) + B2 (i) The difference signal Z (i) between the signals Y1 (i) and Y2 (i) generated by the detection coils 3 and 4 is as follows. Is represented by the following equation (6).

[0019]

## EQU6 ## Z (i) = x1 (i) -x2 (i) + B1
(I) -B2 (i) The autocorrelation of the difference value Z (i) expressed by the above equation (6) is obtained by the following equation (7).

[0020]

(Equation 7) When the noise B (i) in the above equation 7 is considered as white noise,
Equation 7 is represented by the following equation 8.

[0021]

R (i) = {x1 (i) -x2 (i)} x
1 (ij) −x2 (ij)｝ The above equation 8 indicates that the detection coils 3 and 4 are in the ideal state in which the noises B1 (i) and B2 (i) are considered as white noise. This shows that the correlation value of the noise having a frequency similar to that of the detection target signal which is simultaneously input is negligible and is represented only by the autocorrelation value of the difference value of the detection target signal.

However, the noises B1 (i) and B2 (i) generated during the actual iron piece detecting operation are not white noises.
However, the autocorrelation value of the difference value of the detection target signal is a value sufficiently larger than the correlation value of the noises B1 (i) and B2 (i). Therefore, the detection signals DT1 and DT
Thus, the signal to be detected can be detected satisfactorily on the basis of the calculation result obtained by calculating the autocorrelation relation with respect to the difference value of the second.

The above-described iron piece detecting device has an effect of canceling noises simultaneously entering two coils, such as pitching vibration with respect to the sensor unit. However, pitching and rolling are mixed in vibration noise and the like, and when the strength of the vibration increases, the correlation value of the noise that enters the two coils in opposite phases due to the rolling becomes larger than the correlation value of the signal of the iron piece. The size cannot be ignored. For this reason, there is a problem that noise and signals cannot be distinguished.

The present invention has been made in view of the above circumstances, and it is an object of the present invention to perform a good detecting operation even when an external unnecessary magnetic field is present, and to suppress noise of pitch and roll. An object of the present invention is to provide a method and an apparatus for detecting an iron piece, which can reliably distinguish the signal from the signal, and can surely detect even when the signal and the noise of the iron pieces are input simultaneously.

[0025]

SUMMARY OF THE INVENTION According to the present invention, there is provided a method for detecting a direction of movement of a detection object so as to interlink with a DC magnetic flux in a DC magnetic field .
And by detecting the output signal of a pair of detection coils that are spaced apart in the longitudinal, the by the <br/> iron compound which is deposited or included in the detected body which moves relatively in the DC magnetic field relates to a method of detecting the iron compound by the output signal changes, the object of the present invention, the difference of the pair of detection coils
Detecting a signal, the autocorrelation value by <br/> plurality of delay time of the differential signal with respect to the signal of the iron compound, prior to noise
Basic pattern and autocorrelation values by a plurality of delay time of the serial difference signal - is stored as down, the detection co upon detection
The autocorrelation value is calculated for a plurality of delay times for the difference signal of the signal, the product-sum operation is performed on the basic pattern, the matrix operation is performed, and the matrix operation value obtained by the matrix operation is determined. the iron compound by comparing the value
Achieved by detecting .

Also, the present invention provides a method of interlocking with a direct current magnetic flux in a direct current magnetic field, and
Spaced by detecting a difference signal placed a pair of detection coils, wherein in said difference signal changes by attachment or inclusion has been iron compounds through the said DC magnetic field to said detected body relatively moving An object of the present invention relates to an iron piece detecting device that detects iron pieces, and a delay time setting unit that sets a delay time for the difference signal, and calculates an autocorrelation value of the difference signal according to the delay time. An autocorrelation value calculator,
An inverse matrix operation unit that calculates an inverse matrix based on the autocorrelation value; a memory that stores the autocorrelation value and the inverse matrix of the shingles and noise; and an autocorrelation value and the basic pattern at the time of detection. This is achieved by providing a matrix operation unit for obtaining a matrix value by an inverse matrix, and a comparison unit for comparing the matrix value with a predetermined value and outputting an alarm signal.

[0027]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a perspective view of an iron piece detecting device according to the present invention.

The hardware configuration of the present invention is the same as that shown in FIG. 1. Conventionally, the arithmetic unit 30 calculates the autocorrelation value of the difference value with respect to the only delay time τ to detect the presence or absence of iron fragments. However, in the present invention, a plurality of delay times τ1, τ
The auto-correlation value of each difference value for
Ask according to.

[0029]

R (τ) = Σx (t) x (t + τ) That is, for example, the delay time τ is set to τ1, τ2, τ3 (τ1 <τ
2 <τ3), assuming that a, b, and c are the results, the detection signals of iron pieces are a>b> c, and in the case of floor vibration or other noise, a> b, b <c. Since it has been found, the present invention applies this principle to detect the presence or absence of shingles based on the magnitude of the autocorrelation value of a plurality of delay times for each difference value.

FIGS. 2 to 4 show characteristic examples which are the principles of the present invention. In each case, the horizontal axis represents the delay time τ and the vertical axis represents the level of the autocorrelation value. The time difference is shown. FIG. 2 shows data of a 1.0 mm iron ball, FIG. 3 shows data of vibration noise of a floor or the like, and FIG. 4 shows data of noise of a conveyance path. From these figures, it can be seen that only the iron ball data has the characteristic that the level decreases in proportion to the delay time τ, and the level changes irregularly with respect to the delay time τ in other noise data. It has become something. Therefore, in each figure, for example, the level when the delay time τ = 20 is τ20, the level when the delay time τ = 40 is τ40, and the level when the delay time τ = 60 is τ6.
Assuming that the level at time 0 and the delay time τ = 70 is τ70, the iron pieces and noise can be determined from the magnitude relationship of the levels as follows.

(1) When τ20> τ40, τ60> τ70
In the case of → Iron pieces (2) In the case of τ20 <τ40 and τ60τ <70 →
Vibration noise (3) When τ20> τ40 and τ60 <τ70 →
Conveyance path noise Here, in the above detection method, iron pieces and noise can be surely distinguished. However, as is clear from FIGS. 2 and 3, the correlation value of vibration noise has a level that is several times smaller than that of a signal of small iron pieces. Since noise and a signal are input at the same time, the magnitude relationship between the levels is that of noise. For this reason, there is a problem that the presence or absence of iron pieces cannot be accurately detected.

Here, when the level of the autocorrelation value is three-dimensionally expressed in accordance with the change of the delay time τ and the elapse of the time t for iron pieces and each noise, the results are as shown in FIGS. That is, FIGS. 5 to 9 show delay time τ,
FIG. 5 is a three-dimensional display of the elapse of time t and the level relationship of the autocorrelation values. FIG. 5 shows noise data of a transport system such as a motor and a transport zone, and FIG. 6 shows a state in which the transport system is stopped. 7 shows autocorrelation value data for a signal of an iron ball having a diameter of 1.0 mm, and FIG. 8 shows autocorrelation value data for a signal of a 1.2 mm diameter. FIG. 9 shows autocorrelation value data with respect to a combined signal of a 1.0 mm-diameter iron ball and floor vibration noise. These data
As is apparent from the data, the pattern of the noise of the transfer system (FIG. 5) and the noise of the floor vibration (FIG. 6) are greatly different, and the noise (FIGS. 5 and 6) and the iron ball (FIGS. There is a great difference in the pattern with 8). The pattern of the iron ball is almost the same regardless of its diameter. Therefore, FIG.
A noise pattern of the transfer system as shown in FIG. 6, a noise pattern of floor vibration as shown in FIG. 6, and a signal pattern of an iron ball as shown in FIG. 7 or FIG. For example, even if a combined signal of noise and a signal as shown in FIG. 9 is input, noise and iron pieces can be distinguished by extracting the component amount included in the signal as a coefficient value. Can be detected.

As described above, the following processing is performed in the present invention.

By the way, a relational expression when a signal of iron pieces and noise are inputted at the same time is expressed by the following equation (10).

[0035]

X (t) = s (t) + n (t) where s (t) represents an iron piece signal and n (t) represents noise.

Then, the relational expression of the autocorrelation at that time is given by the following equation (9).
Substituting Equation 10 into Equation 11 results in Equation 11 below.

[0037]

[Equation 11] When this is expanded, the following equation 12 is obtained.

[0038]

(Equation 12) Since the first term on the right side of the above equation 12 represents the autocorrelation value of a signal of iron pieces, this is replaced with As · Rs (τ), and the second term on the right side represents the autocorrelation value of noise. This is replaced by An · Rn (τ), and the third and fourth terms on the right-hand side represent the cross-correlation value between the shingle signal and noise. It is replaced by Expression 13. Where Rs
(Τ) is the autocorrelation value of the iron piece signal, Rn (τ) is the autocorrelation value of the noise, both of which are values obtained in advance and detected as a basic pattern prior to the detection operation, and As and An are the values thereof. Represents a coefficient indicating the degree of inclusion, ε (τ)
Is an error component due to approximation.

[0039]

R (τ) = As · Rs (τ) + An · Rn
(Τ) + ε (τ) Here, in order to obtain the coefficients As and An, the least squares method is used so that the approximation error component ε (τ) is minimized.
The least squares method is a method of obtaining each coefficient so that the squared value of the error component ε is minimized.
Therefore, the sum of the values must be minimized. Let J be the minimum value,
An evaluation function J in a quadratic form such as 4 is shown.

[0040]

[Equation 14] In order to minimize the evaluation function J, the differentiated value only needs to be 0. Therefore, the partial differentiation of the coefficients As and An is calculated as in the following equation.

[0041]

(Equation 15) By substituting Equations 13 and 14 into the above equation, the following Equation 16 is derived.

[0042]

(Equation 16) Therefore, the coefficients As and An for minimizing the evaluation function J in the above equation (14) are obtained by the following equation (17).

[0043]

[Equation 17] In the case of Equation 17, since there is only one type of noise, a 2 × 2 determinant is obtained.
The correlation pattern has two noise terms as shown in the following equation.

[0044]

R (τ) = As · Rs (τ) + An1 · Rn
1 (τ) + An2 · Rn2 (τ) + ε (τ) Therefore, the determinant is also 3 × 3 in this case. By increasing the number of matrices to 2 × 2, 3 × 3, 4 × 4,..., Multiple noises can be dealt with.

In order to execute the above operation, in the present invention, the configuration of the arithmetic unit 30 is a functional block as shown in FIG. That is, the calculation unit 30 calculates the difference signal DS from the signals of the detection coils 3 and 4 stored in the memories 16 and 17 and outputs the calculated difference signal DS, and the delay time in synchronization with the calculation of the difference signal DS. Delay time setting unit 302 for setting τ
And the autocorrelation value S based on the difference signal DS and the delay time τ
An autocorrelation value calculation unit 305 for calculating C;
An inverse matrix calculator 303 for obtaining an inverse matrix of the first term on the right side of Expression 13 based on C, and a memory 304 for storing the autocorrelation value SC as a basic pattern and for storing the calculated inverse matrix. are doing. Further, the arithmetic unit 30 includes the memory 3
13 from the inverse matrix and the autocorrelation value SC stored in
And a matrix value TV calculated
Is a predetermined value (threshold) set in the setting unit 307
And a comparison unit 308 for comparing with the SV. Comparison section 3
If the matrix value TV exceeds the predetermined value SV at 08, it is determined that iron pieces have been detected and an alarm signal AL is output to the output unit 31 to issue an alarm.

In such a configuration, the operation is shown in FIG.
1 and FIG. 12 will be described.

FIG. 11 is a flowchart showing the acquisition of a basic pattern as pre-processing. Prior to the use of the detecting device, data suitable for the installation environment is stored as a basic pattern in a memory. 304. That is, the data of the two detection coils 3 and 4 are read from the memories 16 and 17 in which the data is temporarily stored, and the difference signal DS of the two is obtained by the difference signal calculation section 301 (step S1). Are input to an autocorrelation value calculation unit 305 to calculate the autocorrelation value of the shingle signal and the autocorrelation value of each noise, and store them as the basic pattern of the shingle signal and the basic pattern of each noise, respectively. Stored in step 304 (step S2). The autocorrelation value of the iron piece signal and the autocorrelation value of each noise calculated by the autocorrelation value calculation unit 305 are calculated by an inverse matrix calculation unit 303.
And the operation of the inverse matrix is executed and the result is stored in the memory 3
04 (step S3). Thus, the operation as preprocessing ends.

FIG. 12 shows a flow chart at the time of detection. Also at the time of detection, first, the difference signal DS of the two detection coils 3 and 4 is taken by the difference signal calculation section 301 (step S10).
In synchronization with the calculation, the delay time setting unit 302 sets the delay time τ
Is set to 0 (step S11), and the autocorrelation value calculation unit 3
In step 05, the autocorrelation value SC of the difference signal DS is calculated with the delay time τ = 0 and stored in the memory 304 (step S12).
Next, the delay time setting unit 302 updates the delay time τ by +1 (step S13), and repeats the calculation and storage of the autocorrelation value until a predetermined delay time τ ′ is reached (step S13).
14). Then, when the predetermined delay time τ ′ is reached, the matrix calculation unit 306 calculates the autocorrelation values for the calculated and stored delay times τ = 0 to τ ′ and the steps S1 and S2 of the preprocessing. The product-sum operation with each of the basic patterns obtained in step (1) is performed (step S15), and the matrix calculation with the inverse matrix obtained in step S3 of the preprocessing and stored (step S16). The comparison unit 308 determines whether or not the coefficient of the term of the iron piece signal is equal to or larger than the predetermined value SV set by the setting unit 307 with respect to the matrix value TV obtained by the matrix calculation unit 306 (step S17). If it is equal to or smaller than the predetermined value SV, it is determined that noise is present, and the process returns to step S10.
If the coefficient of the term of the signal of the iron pieces is equal to or more than the predetermined value SV, since the iron pieces are included, the alarm signal AL is output and an alarm such as a buzzer is transmitted (step S18). Thereafter, the process returns to step S10.

[0049]

As described above, according to the present invention, the autocorrelation value of the difference signal with respect to a plurality of delay times is obtained, and iron pieces and noise are distinguished according to the magnitude of the autocorrelation value. Reliable detection is possible even for the vibration noise of the roll. In addition, the autocorrelation values for the signals and noises of iron pieces are stored in advance as basic patterns,
At the time of actual detection, the coefficient value of each term is obtained by a product-sum operation with the basic pattern, etc., and iron pieces and noise are distinguished according to the magnitude of each term coefficient value. Even if input, it can be detected reliably.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an iron piece detecting device of the present invention.

FIG. 2 is a diagram illustrating a characteristic example of a delay time with respect to an iron ball and an autocorrelation value.

FIG. 3 is a diagram illustrating a characteristic example of a delay time with respect to floor vibration noise and an autocorrelation value.

FIG. 4 is a diagram illustrating a characteristic example of a delay time with respect to a carrier system noise and an autocorrelation value.

FIG. 5 is a characteristic diagram showing an autocorrelation value of a signal with respect to a carrier system noise in a relationship between a delay time and an elapsed time.

FIG. 6 is a characteristic diagram showing an autocorrelation value of a signal with respect to floor vibration noise in a relationship between a delay time and an elapsed time.

FIG. 7 is a characteristic diagram showing an autocorrelation value of a signal with respect to an iron ball having a diameter of 1.0 mm as a relationship between delay time and elapsed time.

FIG. 8 is a characteristic diagram showing an autocorrelation value of a signal with respect to an iron ball having a diameter of 1.2 mm as a relationship between delay time and elapsed time.

FIG. 9 is a characteristic diagram showing an autocorrelation value of a composite signal of an iron ball having a diameter of 1.0 mm and floor vibration noise in a relationship between delay time and elapsed time.

FIG. 10 is a block diagram illustrating a configuration example of a calculation unit according to the present invention.

FIG. 11 is a flowchart showing an operation example of preprocessing according to the present invention.
It is.

FIG. 12 is a flowchart showing a detection operation example of the present invention.

FIG. 13 is a side view illustrating the operation of detecting iron pieces by a DC magnetic field.

FIG. 14 is a diagram for explaining calculation of an autocorrelation value.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Permanent magnet 3, 4 Detector coil 5 Transfer zone 6 Drive motor 12, 13 Low-pass filter (LPF) 14, 15 A / D converter 16, 17, 304 Memory 20 Control unit 30 Operation unit 31 Output unit 301 Difference signal Calculation unit 302 Delay time setting unit 305 Autocorrelation value calculation unit 306 Matrix calculation unit 308 Comparison unit

──────────────────────────────────────────────────続 き Continuing on the front page (72) Michio Nakano 35-35, Sumiyoshidai, Aoba-ku, Yokohama-shi, Kanagawa 34 (72) Inventor Bichai Sachau 1-36-10 Kita-senzuku, Ota-ku, Tokyo Kita-senzuku lodgings RA35 Room (56 References JP-A-7-260943 (JP, A) JP-A-8-233946 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G01N 27/72-27/90 G01V 3/00

Claims (6)

(57) [Claims] (1) to link with a DC magnetic flux in a DC magnetic field;
One detects the output signal of a pair of detection coils that are spaced apart back and forth relative to the moving direction of the detected body, said attached or included in the detected body which moves relatively in the DC magnetic field A method of detecting the iron pieces by changing the output signal depending on the iron pieces, the method comprising :
A difference signal of the same, a plurality of autocorrelation values are obtained for each of the plurality of delay times with respect to the difference signal, and the iron pieces are detected based on a magnitude relationship between the autocorrelation values. Iron piece detection method. 且 to 2. A in DC magnetic field DC magnetic flux interlinked so
One detects the output signal of a pair of detection coils that are spaced apart back and forth relative to the moving direction of the detected body, said attached or included in the detected body which moves relatively in the DC magnetic field A method of detecting the iron pieces by changing the output signal depending on the iron pieces, the method comprising:
Detecting a difference signal of Le, the difference with respect to the signal of the iron compound
An autocorrelation value based on a plurality of delay times of the signal and an autocorrelation value based on a plurality of delay times of the difference signal with respect to noise are stored as basic patterns, and are stored at the time of detection.
An autocorrelation value based on a plurality of delay times is also calculated for the difference signal of the detection coil , a product-sum operation is performed on the basic pattern, a matrix operation is performed, and a matrix operation value obtained by the matrix operation is calculated. iron detecting how to characterized in that so as to detect the iron compound by comparison with a predetermined value. 3. The iron piece detecting method according to claim 2 , wherein said noise is of a plurality of types, and a basic pattern is stored for each of said noises. 4. The iron piece detecting method according to claim 2 , wherein an inverse matrix required for the matrix operation is stored together with the basic pattern. 且 5. A in DC magnetic field DC magnetic flux interlinked so
One detects a difference signal of a pair of detection coils that are spaced apart back and forth relative to the moving direction of the detected body, said attached or included in the detected body which moves relatively in the DC magnetic field a piece of iron detection device to detect the iron compound by the difference signal is changed by iron compounds, and delay time setting unit for setting a delay time with respect to said difference signal, said difference signal according to the delay time An auto-correlation value calculator for calculating an auto-correlation value, an inverse matrix calculator for calculating an inverse matrix based on the auto-correlation value, and a memory for storing the auto-correlation value and the inverse matrix of the shingles and noise. A matrix operation unit for obtaining a matrix value from the autocorrelation value, the basic pattern, and the inverse matrix at the time of detection, and a comparison unit for comparing the matrix value with a predetermined value and outputting an alarm signal. An iron piece detection device, characterized in that: 6. The iron piece according to claim 5 , wherein said delay time setting section sequentially outputs different delay times, and said autocorrelation value calculation section calculates said autocorrelation value for each said delay time. Detector.