JP3135009B2 - Metal detector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention detects whether or not a metal that should not be mixed into a material to be inspected, such as a raw material such as food and medicine, an intermediate product thereof, or a packaged product. In particular, the present invention relates to an apparatus for inspecting an article conveyed by an endless strap such as a conveyor.

[0002]

2. Description of the Related Art Conventionally, some metal detectors use a detection head having one oscillation coil and two reception coils. When a high-frequency current is caused to flow through the oscillation coil, a high-frequency magnetic field is generated by the high-frequency current, and a current is generated in the two receiving coils by the high-frequency magnetic field.
E2 is induced. The two receiving coils are differentially connected so that a detection signal of E1-E2 is obtained.

When an article to be inspected passes through the inside of the detection head, if the article to be inspected does not contain metal, the high-frequency magnetic field generated by the oscillation coil is not disturbed and the metal detection signal (E
1-E2) is 0. On the other hand, when the article to be inspected contains a metal, this disturbs the high-frequency magnetic field, causing a difference between E1 and E2, indicating that the metal has been detected.

The principle of generating a metal detection signal is slightly different between a case where the metal is iron or the like, which is a high magnetic material, and a case where the metal is a nonferrous metal or the like having a low magnetic permeability, but is not directly related to the present invention. The description is omitted.

By the way, such metal detection is carried out, for example, by transporting the article to be inspected by an endless cord band such as a belt conveyor 2 as shown in FIG. 12 (a), and providing a detection head 4 on the conveyor 2. There is. In such a case, dust easily accumulates in the endless seam 8 of the belt 6 of the conveyor 2, and when the seam 8 passes through the detection head 4, a metal detection signal is generated despite the absence of metal. There was something to do. Further, an adhesive is used to join the belt 6 at the seam 8, but the component of the adhesive may behave like a metal and generate a metal detection signal in some cases.

Generally, in a metal detecting device, a detecting head 4
Is compared with a predetermined threshold value, and when the metal detection signal is larger than this threshold value, it is determined that the metal is detected. Therefore, it is possible to prevent the seam 8 as described above from being erroneously determined to be metal. Then, the threshold value must be set higher than the level e1 of the metal detection signal shown in FIG.
There is a possibility that a metal that generates a detection signal having the same level as the level e1 of the detection signal generated when the seam 8 is detected may not be detected.

In order to solve this problem, as a first method, the conveyor belt 6 is periodically cleaned and the seam 8 is cleaned.
In order to prevent the accumulation of dust, the conveyor belt 6 is sometimes replaced. As a second method, as shown in FIG.
The belt 6 is also joined so that a makes an appropriate angle θ with respect to the longitudinal direction of the belt 6.

In the method shown in FIG. 13A in which the seam is inclined, the seam 8a which behaves like a metal is formed as shown in FIG.
Unlike the case of (1), since the light gradually passes through the detection head 4, the level e2 of the detection signal becomes as shown in FIG. 13B, and the level e2 of the straight seam 8 as shown in FIG. It becomes considerably smaller than the level e1 of the detection signal in the case. Therefore, since the threshold value can be set smaller than that in the case of FIG. 12, even a metal that generates a detection signal of a level smaller than the level e1 shown in FIG. 12B can be detected.

[0009]

However, frequent cleaning of the belt 6 is troublesome, and even if the seam is inclined, even a metal that generates only a small level detection signal can be detected to some extent. However, it is still impossible to detect a metal that generates a detection signal smaller than the level e2.

[0010]

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and therefore, in the present invention, there is provided an endless strap which can run while supporting an article on an upper surface thereof. Metal detection means for generating a detection signal whose level changes depending on whether or not the article conveyed by the endless lace contains metal, and an endless lace in a state where the endless lace travels without supporting the article. Partition the total length of the tie into multiple sections,
Each time each section passes through the detection means, the storage means storing the detection signal from the detection means at that time, and the endless lace band in a state where the endless lace band is running while supporting the article. Means for determining which of the sections has passed the detecting means, and a deviation calculating means for calculating a deviation between a value stored in the storing means corresponding to the section determined by the determining means and a detection signal of the detecting means. , Is provided.

[0011]

According to the present invention, the storage means stores the detection signals of the metal detection means in each section of the endless strap when the endless strap passes through the metal detection means without carrying the metal. I have. For example, a detection signal for a section corresponding to a seam that behaves similarly to a metal as described above ,
Passes the detection means when the seam passes below the detection means
Detection signal for the section where
The detection signal for the section of interest is also stored. Then, when each section passes through the detection means, the determination means determines the section, the storage value corresponding to the section is read from the storage means, and the deviation from the detection signal of the detection means is calculated by the deviation calculation means. Since the calculation is performed, when the seam passes through the detecting means in a state where the article is actually being conveyed by the endless strap, the stored value corresponding to the seam is read from the storage means, and the seam is actually detected. Since the deviation from the detection signal from the detecting means is calculated, the deviation is 0, and the seam is not erroneously determined to be metal. Similarly, when the seam passes below the detecting means or when metal dust passes through the detecting means, the deviation becomes zero, and it is possible to prevent erroneous determination.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The first embodiment, as shown in FIG. 2, has an endless cord band, for example, a belt conveyor 10, on which a metal detection head 12 is provided.
The article to be inspected is carried into the belt conveyor 10 by the carry-in conveyor 14, and is conveyed on the conveyor 10. However, it is detected by the detection head 12 whether the conveyed article to be inspected contains metal or not, and The finished inspection object is carried out by the carry-out conveyor 16.

As shown in FIG. 1, the detection head 12
It has two oscillation coils 18 and two reception coils 20 and 20. The high frequency signal oscillated by the oscillator 22 is amplified by the amplifier 24 and supplied to the oscillation coil 18. As a result, the oscillation coil 18 generates a high-frequency magnetic field, and the high-frequency magnetic field is supplied to the receiving coils 20, 20. The receiving coils 20, 20 are differentially connected, and generate a detection signal as described above when an inspection target including metal passes through the inside of the detection head 12.

This detection signal is amplified by an amplifier 26 and then synchronously detected by a synchronous detector 28. A high-frequency signal is supplied from the amplifier 24 via the variable phase shifter 30 for the synchronous detection. In general, the level of the detection signal obtained from the receiving coils 20 and 20 varies not only according to the size of the metal piece included in the inspection object, but also changes in phase according to the magnetic permeability and the electric resistance value of the metal piece. I do. That is, even if the metal pieces have the same size, the level and phase of the level of the detection signal are different between a magnetic material such as iron and a non-magnetic material such as aluminum having low magnetic permeability. Also,
A highly conductive product containing salt or the like may generate a detection signal in spite of not containing metal. To cope with such a situation, a variable phase shifter 30 is provided, and by adjusting the phase of the high-frequency signal supplied to the synchronous detector 28, an optimum phase shifter for a magnetic or non-magnetic material is provided. The sensitivity is obtained, and unnecessary signals (product signals) generated from conductive products are removed.

The output signal of the synchronous detector 28 is supplied to a sample-and-hold circuit 34 via a filter 32, where it is sequentially sampled and held, converted to digital by an analog-digital converter 36,
The signal is supplied to a signal processing unit 38 composed of a PU.

The signal processing section 38 processes the digitized detection signal from the detection head 12 using the RAM 42 as a work area in accordance with a program stored in the ROM 40. In order to use for this processing, the signal processing section 38 is supplied with a signal from a belt single rotation detecting means 46 for detecting that the belt 44 of the belt conveyor 10 has made one rotation.

The belt one rotation detecting means 46 includes:
For example, as shown in FIG. 5A, a notch 48 is provided at one position on the longitudinal edge of the belt 44, and a reflection type photoelectric switch 50 shown in FIG. A conveyor provided at an arbitrary position of the conveyor 10 can be used. Also, as shown in FIG. 6A, a minute hole 52 is formed in the belt 44, and this is
The detection using the transmission type photoelectric switch 54 including the light receiving element 54b and the light receiving element 54b can also be used as the belt one rotation detection means 46.

The signal processing section 38 is also supplied with a signal from an article detecting means 56 for detecting whether or not the article to be inspected has been carried on the conveyor 10.
As the detecting means 56, for example, a photoelectric switch provided between the carry-in conveyor 14 and the conveyor 10 as shown in FIG. 2 can be used. The reason for using the article detecting means 56 will be described later.

The processing performed by the signal processing section 38 will be described with reference to FIGS. FIG. 3A shows that the belt conveyor 10 does not place an article to be inspected,
When one rotation up to the + 1st rotation is performed, the detection head 12
Shows the change in the detection signal that occurs, where the reference numeral A indicates the detection signal when the seam of the belt 44 of the conveyor 10 passes through the detection head 12, and the reference numeral B indicates the detection signal. Is a detection signal when the seam passes below the detection head 12. FIG. 6B shows a pulse signal generated when the one-rotation detecting means 46 detects one rotation. During the time tn between the generation of this pulse signal and the generation of the next pulse signal,
For example, when the detection signal is sampled at a sampling time T as shown in FIG. 3C, as shown in FIG. 4D, the sampling values Z of N (= tn / T) detection signals are obtained.
1, Z2... ZN are obtained.

FIG. 4 shows the detection signal generated by the detection head 12 when the belt conveyor 10 makes one rotation from the Pnth rotation to the Pn + 1th rotation in a state where the belt conveyor 10 travels while placing an article to be inspected including metal. A change signal is indicated by reference numeral C, which is a detection signal when the inspection object including metal passes through the detection head 12, and the maximum level of the detection signal is indicated by reference numerals A, B. Is smaller than the maximum level of the detection signal corresponding to.

When this detection signal is also sampled by the sampling signal shown in FIG. 3C in synchronization with the detection signal shown in FIG. As shown in b), each sampling value G1, G2,... GN is obtained. From each of these sampling values, the correspondence of each sampling value shown in FIG. 4C (this is a re-posting of each sampling value in a state where the article to be inspected shown in FIG. 3D is not placed) is shown. Are subtracted, that is, G1-Z1, G2-Z2,.
When the GN-ZN was calculated, the detection signals corresponding to the joints were removed and the maximum level was smaller than the detection signals corresponding to the joints, as shown in FIG. , Only the sampled value of the signal corresponding to the metal is obtained.

Therefore, in this embodiment, in the state where the article to be inspected is not placed, the signal processing section is provided between the time when the belt 1 rotation detecting means 46 generates the pulse signal and the time when the next pulse signal is generated. The sampling and holding circuit 34 and the analog-to-digital converter 36 sample and hold using a sampling signal of a period T generated by a timer built in the 38, and store the digitized detection signal in the RAM 42 (hereinafter, this is referred to as This is called a preset.). However, the storage address is changed from the time when the belt 1 rotation detection means 46 generates the pulse signal.
It is determined by an address counter n that counts a sampling signal. Then, in a state where the article to be inspected is placed, the belt one rotation detecting means 46 generates a pulse signal, and then causes the address counter n to count a sampling signal, and reads a stored value from the RAM 42 according to the count value. At this time, subtraction is performed from the digitized detection signal.

Note that the rotation of the belt 44 is not strictly constant depending on how the belt 44 of the conveyor 10 driven by the motor is stretched, and the manner of vibration of the belt 44 differs for each rotation. Therefore, the level of the detection signal from the detection head 12 is small but fluctuates due to the vibration of the belt. In order to remove this vibration component, the belt 44 is rotated a plurality of times, for example, M times, instead of data obtained by one rotation of the belt 44, and the average value of each corresponding sampling value is used as a representative sampling value. . For example, when Z11, Z12,..., Z1M are obtained as the first sampling values, the average value of these is used as the representative first sampling value,
It is stored in AM42.

Such a preset operation can be started by the operator confirming that the inspected article is not conveyed on the conveyor 10 and manually giving an instruction to the signal processing section 38. In this embodiment, the article detecting means 56 further detects whether the article to be inspected has approached the detection head 12 and automatically detects the article if it has not been detected for a certain period of time. It may be performed. In this way, for example, if the state where the article to be inspected is not placed continues for a predetermined time, the preset value is automatically calibrated, which is effective when the detection signal changes with time.

In order to perform the above operation, the signal processing section 38 is programmed to execute each routine as shown in FIGS. FIG. 7 is a timer interrupt routine executed every time a timer incorporated in the signal processing unit 38 generates a sampling signal. FIG. 8 is an article detection interrupt executed each time the article detection unit 56 detects an article. FIG. 9 shows a one-rotation interruption routine executed each time the belt one-rotation detecting means generates a pulse signal.

In order to perform the above-described operations, the signal processing unit 38 performs a calibration operation flag, an address counter n, a belt rotation counter B, an article detection timer T, and N detection signal accumulation registers SUMZn ( 1 ≦ n ≦ N), N
A number of preset registers Zn (1 ≦ n ≦ N) and a number N of corrected detection signal storage registers Dn (1 ≦ n ≦ N) are prepared in the RAM 42, for example.

For example, when the article detecting means 56 detects an article to be inspected, an article detection interrupt routine shown in FIG. 8 is executed, and the calibration work flag is turned off (step S).
2) The article detection timer T is reset to 0 (step S4), the belt rotation counter B is reset to 0, and each detection signal accumulation register SUMZn (1 ≦ n ≦
N) are reset to 0 (step S6), and this routine ends.

In this state, if the belt makes one revolution, and the belt one revolution detecting means 46 generates a pulse signal, a one revolution interrupt routine shown in FIG. 9 is executed, and the belt revolution counter B is incremented by one. (Step S8), the address counter n is set to the initial value 1 (Step S10), and this routine ends.

After the timer in the signal processing section 38 generates a sampling signal after the execution of these two routines, a timer interrupt routine shown in FIG. 7 is executed, and the digital signal Gn is first sent from the analog-digital converter 36. Read (step S12). This digital signal is obtained by digitizing the detection signal of the detection head 12 for the section of the belt 44 at the position of the detection head 12. Subsequently, it is determined whether the calibration-in-progress flag is on (step S14). Since this flag has been previously turned off by the article detection interrupt routine, the answer to step S14 is NO, and step S1
6 is executed.

At the stage when step S16 is executed, assuming that each preset value is stored in advance in each preset register Zn (1.ltoreq.n.ltoreq.N), the digital signal Gn is read from the read digital signal Gn. Correction is performed by subtracting the stored value of the preset register Zn specified by the count value, and the corrected detection signal storage register D specified by the count value of the address counter n is corrected.
n, and the address counter n is incremented by one.

Subsequent to step S16, the article detection timer T is incremented by 1 (step S18), and it is determined whether the article detection timer T has counted up to a predetermined value, that is, whether the time has elapsed (step S20). If the answer is no, an instruction is given to calibrate manually (the operator confirms that the article to be inspected is not on the belt conveyor 10 with his own eyes and calibrates the preset value). It is determined whether or not a key (this key is provided on a keyboard not shown) is operated (step S22). If the answer is no, this routine is terminated.

Thereafter, each time a sampling signal is generated, a timer interrupt routine is executed, the digital signal Gn read in this routine is corrected, and the article detection timer T is incremented by one. When T times out,
The answer to step S20 is YES, and the calibration-in-progress flag is turned on (step S24). Note that also when the manual calibration key is on, step S24 is executed, and the calibration in progress flag is turned on.

When the belt-one-rotation detecting means 46 detects one rotation of the belt 44 after the calibration-in-progress flag is turned on, the count value of the belt rotation counter B is reduced by one in the one-rotation interruption routine. The value is incremented, and the address counter n is set to 1.

Then, a timer interrupt routine is executed each time a sampling signal is generated, the digital signal Gn is read in step S12, and it is determined in step S14 whether the calibration in progress flag is ON. Since it is turned on, the answer is yes, and it is determined whether the count value of the belt rotation counter B is a predetermined value M, that is, whether the belt 44 has rotated M times (step S26). If the answer is no, the digital signal Gn at that time is accumulated in the value of the detection signal accumulation register SUMZn designated by the value of the address counter n, and the value of the address counter n is incremented by 1 (step S28). Step S18 and the subsequent steps are performed, and this routine ends.

Therefore, the digital signal G of the detection signal for each section of the belt 44 until the belt 44 makes one revolution.
n is accumulated in each corresponding detection signal accumulation register SUMZn. When one rotation is completed, the one-rotation interrupt routine shown in FIG. 9 is executed again, the belt rotation counter M is incremented by one, and the address counter n is set to one. Then, each time a sampling signal is generated again, a digital signal corresponding to each detection signal accumulation register SUMZn is accumulated in the same manner as described above.

When the belt 44 rotates M times in this manner, the N detection signal accumulation registers SUMZn store the detection head 1 for the corresponding section of the belt 44, respectively.
M digital signals of the detection signal detected by the step 2 are accumulated.

Accordingly, when the belt 44 rotates M times, the answer in step S26 becomes YES, the belt rotation counter B is reset to 0 (step S30), and the value of each of the N accumulation registers SUMZn is divided by M. Is stored in the corresponding preset register Zn (step S32), and furthermore, each accumulation register SUM
The Zn is reset to 0 (step S34), the flag under calibration is turned off, the address counter n is set to 1 (step S36), and the steps after step S18 are performed. As a result, the value of each preset register Zn stores, as a preset value, an average value of M rotations of each detection signal of the detection head 12 for N sections of the belt 44 in a state where the article to be inspected is not placed.

When a new preset value is stored in each preset register Zn in this manner, after the belt 44 makes one revolution and the address counter n is set to 1 in the one revolution interrupt routine, every time a sampling signal is generated. In step S1 of the timer interrupt routine,
2, S14 and S16 are executed, and the preset value Zn corresponding to the digital signal Gn of the detection signal for the section of the belt 44 at the position of the detection head 12 at that time is subtracted, and the corrected detection signal storage register Dn Is stored. These corrected detection signal storage registers Dn
The stored values are processed by a program (not shown) to determine whether or not metal is detected.

FIG. 10 shows a second embodiment. In this embodiment, the detection signal from the amplifier 26 is
a and 28b, filters 32a and 32b, sample and hold circuits 34a and 34b, and analog-digital converters 36a and 36b. The high-frequency signal from the amplifier 24 is supplied to the synchronous detector 28a as it is, and the synchronous detector 28b is supplied to the synchronous detector 28b.
A 90 ° phase shifter 58 supplies a 90 ° phase-shifted high-frequency signal from the amplifier 24.
That is, with reference to the high frequency signal of the high frequency amplifier 24,
The high-frequency signal having the same phase as this is supplied to the synchronous detector 28b, and the synchronous detector 28a is supplied with a high-frequency signal having a phase different from that of the high-frequency signal by 90 °. The other points are configured in the same manner as in the first embodiment, and a detailed description thereof will be omitted.

By providing the phase shifter 58 in this manner, the high-frequency signal supplied to the synchronous detector 28a and the synchronous detector 28b
The X and Y components of the detection signal are supplied to the signal processing unit 38 by setting the phase difference with the high-frequency signal supplied to the signal processing unit to 90 °. Here, the X component is a component having the same phase as the phase of the high-frequency signal from the amplifier 24, and the Y component is equal to the X component and 9 components.
Components having different phases of 0 °.

[0041] When the article to be inspected passes the detection head 12, also include metal the article to be inspected, its conductivity,
Depending on the magnetic permeability, not only the amplitude but also the phase changes. That is, the detection signal can be considered as a vector. For example, a detection signal of a magnetic material such as iron is represented as a vector V1 as shown in FIG.
It is represented as a vector V2, and usually both have a phase difference of 90 °. Therefore, the X component V1X of the detection signal of a magnetic substance such as iron is large, and the Y component V2Y of the detection signal of a nonmagnetic substance such as copper is large. Therefore, when the magnetic substance is detected, the X component is used. When detecting a non-magnetic material, a variable phase shifter 30 is provided as in the first embodiment by utilizing the Y component, and the phase of the high-frequency signal needs to be adjusted depending on which one is to be detected. There is no.

If the article to be inspected is electrically conductive,
Assuming that the detection head 12 generates a detection signal represented by, for example, the vector V3 even in a state where no metal is included, if the inspection object contains a non-magnetic material such as copper by mistake. And the detection signal is V3 + V2
Is obtained. Therefore, the X component and the Y component of the conductive test object are stored in advance, and each time a detection signal is obtained, the X component of the test object is subtracted from the X component of the detection signal, and the detection signal is obtained. By subtracting the Y component of the article to be inspected from the Y component of the above, and comparing the X component and the Y component obtained by the subtraction with the X component and the Y component when the nonmagnetic material is detected, respectively, It can be determined whether or not a body metal is contained. Of course, similarly, it can be determined whether or not a magnetic material is included.

In the above embodiment, the belt one rotation detecting means 46 is formed by a transmission type or reflection type photoelectric switch.
Although the one that detects the notch or the hole provided in the belt 44 is used, a marker may be provided on the belt 44 instead of detecting the notch or the hole, and the marker may be detected by an optical sensor. . Further, the belt conveyor 10 is used as the endless cord band, but a roller conveyor or the like can also be used as long as the belt has an insulating property.

[0044]

As described above, according to the present invention, in a state where the endless strap is traveling without supporting the article, the entire length of the endless strap is divided into a plurality of sections, and each section is provided with a metal detecting means. Each time the vehicle passes, the detection signal from the detection means at that time is stored in the storage means, and in a state where the endless strap is traveling while supporting the article, any section of the endless strap passes the metal detection means. It is configured such that the deviation between the storage value of the storage unit corresponding to the section determined by the determination unit and the detection signal of the metal detection unit is calculated by the deviation calculation unit. So
For example, even joint endless ties went behave like a metal, even through the lower seam metal detector, or be attached dust such as metal on the endless ties, which
They will not be erroneously detected . Further, even if the level of the signal actually detecting the metal is lower than the detection signal generated based on the joint or the like , the metal can be detected.

[Brief description of the drawings]

FIG. 1 is a block diagram of a first embodiment of a metal detection device according to the present invention.

FIG. 2 is a diagram showing a schematic configuration of the first embodiment.

FIG. 3 is a diagram illustrating outputs of respective units in a state where the belt conveyor according to the first embodiment is running without placing a non-inspection item.

FIG. 4 is a diagram illustrating outputs of respective units when the belt conveyor according to the first embodiment is running with a non-inspection article placed thereon.

FIG. 5 is a diagram showing an example of one rotation detecting means used in the first embodiment.

FIG. 6 is a diagram showing another example of the one-rotation detecting means used in the first embodiment.

FIG. 7 is a flowchart showing a timer interrupt routine of the first embodiment.

FIG. 8 is a flowchart showing an article detection interrupt routine according to the first embodiment.

FIG. 9 is a flowchart showing a one-turn interruption routine of the first embodiment.

FIG. 10 is a block diagram of the second embodiment.

FIG. 11 is a vector diagram showing a detection signal in the second embodiment.

FIG. 12 is a diagram illustrating an example of a conventional metal detection device and a detection signal of a detection head thereof.

FIG. 13 is a diagram showing another example of a conventional metal detection device and a detection signal of the detection head thereof.

[Explanation of symbols]

 Reference Signs List 10 belt conveyor (endless cord band) 12 detection head (metal detection unit) 38 signal processing unit (judgment unit, deviation calculation unit) 42 RAM (storage unit) 46 belt one rotation detection unit (judgment unit)

Claims (1)

(57) [Claims] 1. An endless cord band that can run while supporting an article on its upper surface, and metal detection means that generates a detection signal whose level changes according to whether or not the article conveyed by the endless strap contains metal. In a state where the endless strap is traveling without supporting an article, the entire length of the endless strap is divided into a plurality of sections, and each time each section passes through the detection means, the detection from the detection means at that time is performed. Storage means for storing signals; means for determining which section of the endless strap passes through the detection means while the endless strap is traveling while supporting an article; And a deviation calculating means for calculating a deviation between a value stored in the storage means and a detection signal from the detecting means corresponding to the section determined by the above.