Wire flaw detection device and wire diagnostic method

Abstract

To provide a wire flaw detection device capable of determining the extent and damaged portion of the damage state of a wire without individually setting a threshold based on a difference in the type of wires and installation environments.SOLUTION: A wire flaw detection device includes: a magnetization mechanism which forms a magnetic path in a predetermined section of a wire; and multiple magnetic sensors which can detect magnetic signals generated from strands of the wire. The wire flaw detection device detects the damage on the wire. The wire flaw detection device further includes a signal analysis unit which includes a signal collector that collects magnetic signals output from the magnetic sensors and a signal processor that processes the magnetic signals output from the signal collector. The signal analysis unit is configured to detect the damage on the wire by using multiple feature amounts from each of the multiple signal waveforms detected by the multiple magnetic sensors.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to flaw detection for measuring the state of damage to a metal wire, and more particularly to a wire flaw detection device and a wire diagnosis method suitable for use in detecting breaks in wire ropes. [Background technology]

[0002] The metal wires used in elevators are wire ropes made of twisted thin steel wires, and are used as hoisting ropes for elevator cars. In escalators, metal wires are installed inside the resin material of the handrails to maintain their strength.

[0003] From a safety standpoint, it is necessary to periodically check the condition of such metal wires for damage. Conventionally, a magnetic leakage flux inspection method has been known as a representative technique for detecting a damaged state of a wire.

[0004] Furthermore, in the technology for detecting the damage state, it is possible to determine the level of damage state of the wire by analyzing and processing the detected signal.

[0005] Patent document 1 discloses a technology for determining the level of damage to a wire rope based on a level calculated from the correlation between the sensor signal output due to damage to the wire rope moving in the longitudinal direction and the number of times this signal output exceeds a threshold. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2013-35693 A Summary of the Invention [Problem to be solved by the invention]

[0007] The wire rope used in elevators may vary in thickness and number of strands depending on the model of elevator. Furthermore, the damage state of the rope may vary depending on individual differences in rope manufacturing lots and differences in installation conditions such as the weight of the elevator car.

[0008] When damage to a wire rope is detected, it is useful to identify the cause of the damage in order to take corrective measures for maintenance. One possible cause of damage, for example when damage occurs to the surface of the wire rope (breakage), is wear damage caused by excessive contact with the sheave. For example, when damage occurs to the inside of a wire rope (valley break), it is believed to occur due to poor lubrication caused by a lack or depletion of lubricating grease.

[0009] In the assessment method of Patent Document 1, a signal output threshold is set for each longitudinal section of the wire rope, and the number of signals exceeding this threshold is counted to classify the damage state into three levels. However, the threshold value must be set individually based on the wire rope thickness and the number of strands, and therefore, when the wire rope thickness or number of strands is changed, it is often necessary to reset the threshold value. Furthermore, the location of damage (surface mountain cut, inner valley cut) cannot be determined by the threshold value alone.

[0010] An object of the present invention is to provide a wire flaw detector and a wire diagnosis method that can determine the degree of damage to a wire and the location of the damage without having to set individual thresholds depending on the type of wire or installation environment.

[0011] The above and other objects of the present invention and novel features of the present invention will become apparent from the description of this specification and the accompanying drawings. [Means for solving the problem]

[0012] The wire flaw detection device of the present invention comprises a magnetization mechanism that forms a magnetic path in a predetermined section of the wire, and a plurality of magnetic sensors that can detect magnetic signals generated from the strands of the wire, and further comprises a signal analysis unit that includes a signal collector that collects the magnetic signals output from the magnetic sensors and a signal processor that processes the magnetic signals output from the signal collector, and the signal analysis unit detects damage to the wire using a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors.

[0013] The wire diagnosis method of the present invention is a wire diagnosis method using a wire flaw detection device that detects damage to the wire, and is equipped with a magnetization mechanism that forms a magnetic path in a predetermined section of the wire and a plurality of magnetic sensors that can detect magnetic signals generated from the wire strands, and includes the steps of: for a normal wire, using a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors, creating a unit space of the plurality of feature quantities in advance; for a test wire, acquiring a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors; detecting the distance between the unit space in the normal wire and the test wire based on the data of the feature quantities of the signal waveforms; and detecting a damage state of the test wire based on the distance. Effect of the Invention

[0014] According to the above-mentioned wire flaw detector and wire diagnosis method of the present invention, it is possible to determine the damage state of the wire (level of damage and damaged location) without setting individual thresholds depending on the type of wire or installation environment. Furthermore, by determining the damaged parts of the wire, the cause of the damage can be inferred, which will make it possible to optimize maintenance and improve the safety performance of equipment that uses wires.

[0015] Problems, configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief description of the drawings]

[0016] [Figure 1] 1 is a schematic diagram (cross-sectional view) showing a magnetic sensor and a wire rope of a wire rope flaw detector according to a first embodiment of the present invention. [Diagram 2] 1 is a functional block diagram showing a configuration of a wire flaw detector according to a first embodiment of the present invention. [Diagram 3] FIG. 2 is a schematic diagram showing types of feature quantities of a signal waveform. [Figure 4] 4 is a flowchart of a process for determining whether a wire rope is normal or abnormal according to the first embodiment of the present invention. [Diagram 5] 4 is a flowchart of a process for determining a location where a wire break has occurred in a damaged wire rope according to the first embodiment of the present invention. [Figure 6] 3 is a diagram showing an example of the relationship between the value of a signal feature amount and normal and abnormal levels according to the first embodiment of the present invention. FIG. [Figure 7] 4 is a diagram showing an example of a relationship between a value of a signal feature amount and a damaged portion of a wire according to the first embodiment of the present invention. FIG. [Figure 8] FIG. 4 is a diagram showing an example of a display of a determination result according to the first embodiment of the present invention. [Figure 9] FIG. 11 is a schematic diagram showing the arrangement of wires and sensors according to a second embodiment of the present invention. [Figure 10] FIG. 5 is a functional block diagram showing a configuration of a wire flaw detector according to a second embodiment of the present invention. [Figure 11] 10 is a flowchart of a process for determining whether a wire is normal or abnormal according to a second embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] Hereinafter, the embodiments and examples of the present invention will be described with reference to text and drawings. However, the structures, materials, and other specific configurations shown in the present invention are not limited to the embodiments and examples described here, and can be appropriately combined or improved without changing the gist of the invention. Elements that are not directly related to the present invention are not shown.

[0018] The wire flaw detector of the present invention is a wire flaw detector that detects damage in the wire and is equipped with a magnetization mechanism that forms a magnetic path in a specified section of the wire and a plurality of magnetic sensors that can detect magnetic signals generated from the strands of the wire. In addition, the wire flaw detector of the present invention further includes a signal analysis unit including a signal collector that collects the magnetic signal output from the magnetic sensor, and a signal processor that processes the magnetic signal output from the signal collector. In the wire flaw detector of the present invention, the signal analysis unit detects damage to the wire by using a plurality of feature quantities of each of a plurality of signal waveforms detected by a plurality of magnetic sensors.

[0019] According to the wire flaw detection device of the present invention, the signal analysis unit detects damage to the wire using multiple feature quantities of each of the multiple signal waveforms detected by the multiple magnetic sensors, making it possible to determine the damage state of the wire (level of damage and damaged location) without individually setting threshold values ​​depending on the type of wire or installation environment. Furthermore, by determining the damaged parts of the wire, the cause of the damage can be inferred, which will lead to the optimization of maintenance and improvement of the safety performance of equipment that uses wires, such as elevators.

[0020] The wire diagnosis method of the present invention is a wire diagnosis method using a wire flaw detector that is equipped with a magnetization mechanism that forms a magnetic path in a specified section of the wire and a plurality of magnetic sensors that can detect magnetic signals generated from the wire strands, and detects damage to the wire. The wire diagnosis method of the present invention includes the steps of: for a normal wire, using a plurality of feature quantities of each of a plurality of signal waveforms detected by a plurality of magnetic sensors, creating in advance a unit space of the plurality of feature quantities; for a wire of a test subject, acquiring a plurality of feature quantities of each of a plurality of signal waveforms detected by a plurality of magnetic sensors; detecting the distance between the unit space in the normal wire and the wire of the test subject based on the data of the feature quantities of the signal waveforms; and detecting a damage state of the wire of the test subject based on the distance.

[0021] According to the wire diagnosis method of the present invention, the damage state of the wire is detected using multiple feature quantities of each of multiple signal waveforms detected by multiple magnetic sensors, making it possible to determine the damage state of the wire (level of damage and damaged location) without having to set individual threshold values ​​due to differences in wire type or installation environment. In addition, the distance between the unit space in a normal wire and the wire of the subject is detected, and the damage state of the wire of the subject is detected based on this distance. This makes it possible to more easily and reliably determine the level and location of damage to the wire of the subject without having to set individual threshold values ​​depending on the type of wire or installation environment.

[0022] In the above-mentioned wire flaw detection device, the signal analysis unit can be configured to calculate the distance between a unit space and multiple features of multiple signal waveforms detected by multiple magnetic sensors based on a unit space created from multiple features of signal waveforms previously measured on a normal wire, and to detect the degree of damage to the wire under test based on that distance. With this configuration, the distance between the unit space and multiple features of multiple signal waveforms detected by multiple magnetic sensors is calculated based on the unit space, and the degree of damage to the test wire is detected based on that distance, making it possible to more easily and reliably determine the level and location of damage to the test wire.

[0023] In the above-mentioned wire flaw detection device, when it is determined that the wire under test is damaged, the signal analysis unit can be configured to calculate the distance between the unit space and multiple feature amounts of multiple signal waveforms detected by multiple magnetic sensors of the wire under test based on a unit space created from multiple feature amounts of signal waveforms previously measured on a normal wire, and detect the damaged portion of the wire under test by comparing the distance with a threshold value based on the distance. In this configuration, the damaged part of the subject's wire is detected by comparing the distance from the unit space with a distance-based threshold, making it possible to more easily and reliably determine the damaged part of the subject's wire.

[0024] In the above-mentioned wire flaw detection device, the signal analysis unit can be further configured to detect the damaged portion of the test wire by using multiple feature quantities including at least a feature quantity indicating kurtosis as the multiple feature quantities when it is determined that there is damage in the test wire. Furthermore, the signal analysis unit can be configured to detect a damaged portion of the wire of the subject by using any combination of kurtosis and amplitude, kurtosis and half-width, or kurtosis and area as multiple feature quantities. With these configurations, a plurality of feature quantities including a feature quantity indicating kurtosis are used, so that it is particularly easy to determine the location where damage has occurred.

[0025] In the above-described wire flaw detector, the signal analysis unit may be configured to use the Mahalanobis distance as the distance. In this configuration, the Mahalanobis distance is used as the distance to the unit space, so that the level and location of damage to the wire of the subject can be determined more easily.

[0026] In the above-mentioned wire diagnosis method, the step of determining the damage state of the subject's wire further includes a step of detecting the degree of damage to the subject's wire and a step of detecting the damaged portion of the subject's wire, and the step of detecting the damaged portion of the subject's wire can be configured to use a plurality of features including at least a feature indicating kurtosis. Furthermore, in the above-mentioned wire diagnosis method, the step of detecting a damaged portion of the wire of the subject can be configured to use any combination of kurtosis and amplitude, kurtosis and half-width, or kurtosis and area. With these configurations, a plurality of feature quantities including a feature quantity indicating kurtosis are used, so that it is particularly easy to determine the location where damage has occurred.

[0027] In the above wire diagnosis method, in the step of detecting the distance, a Mahalanobis distance may be used as the distance. In this configuration, the Mahalanobis distance is used as the distance to the unit space, so that the level and location of damage to the wire of the subject can be determined more easily.

[0028] The wire flaw detector and wire diagnosis method of the present invention can be applied to inspecting damage to wires in various devices that use wires. For example, the present invention can be applied to the wire rope of the hoisting cable of an elevator car, the internal wire of an escalator handrail, and the like.

[0029] Hereinafter, specific embodiments of the wire flaw detector and wire diagnostic method of the present invention will be described.

[0030] (First embodiment) Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. In this embodiment, the wire flaw detector and wire diagnosis method of the present invention are applied to a wire rope used in an elevator.

[0031] The wire ropes used in elevators are made by twisting several thin wires together to form strands, which are then twisted around a core rope to form a single bundle. Wire ropes made of twisted magnetic metal wires are used as hoisting ropes for elevator cars.

[0032] Wire ropes in elevators, which are used as moving ropes, undergo deterioration over time through bending fatigue, wear, corrosion, and other factors, and therefore require regular inspections and checks. Visual inspection is effective for detecting deterioration that shows obvious changes in appearance, such as rust or corrosion, but it can be difficult to detect abnormalities that occur inside the wire rope, as there is variation in inspection accuracy depending on the inspector.

[0033] Therefore, an effective technique for monitoring the safety of wire ropes is to use magnetic leakage flux testing, which detects the deterioration of wire ropes based on the leakage magnetic flux generated from the wire rope. Magnetic leakage flux testing is a method in which the wire rope is excited and the leakage magnetic flux generated is detected by a magnetic sensor.

[0034] A measuring device based on the magnetic leakage flux inspection method (hereinafter referred to as a magnetic leakage flux inspection device) is provided with a magnetization mechanism for magnetizing the wire rope, and a magnet is attached to this magnetization mechanism. A magnetic flux leakage inspection device for wire ropes works by forming a magnetic path when the wire rope is attached to a magnetization mechanism, as a result of the magnetic field emitted from the magnet causing magnetic flux to flow back between the wire rope and the magnetization mechanism.

[0035] If the wire rope in which the magnetic path is formed is damaged, such as by a broken wire, the flow of magnetic flux is impeded and magnetic flux leaks onto the surface of the wire rope. By detecting this leakage magnetic flux with a magnetic sensor, it is possible to detect the state of a broken wire in the wire rope.

[0036] The leakage magnetic flux flaw detector of the present embodiment described below can detect the level of damage state based on the frequency of wire breakage and the type of wire breakage. In this embodiment, the leakage magnetic flux flaw detector will be referred to as a "wire rope flaw detector."

[0037] FIG. 1 is a schematic diagram (cross-sectional view) showing a magnetic sensor and a wire rope of a wire rope flaw detector according to a first embodiment of the present invention. As shown in FIG. 1, a large number of magnetic sensors 3 constituting a sensor unit 20 of a wire rope flaw detector 1 (see FIG. 2) are arranged around a wire rope 2 having a circular cross section.

[0038] The wire rope 2 illustrated in Fig. 1 has eight strands 23, and the entirety including these eight strands 23 are twisted together into one. Note that the wire rope 2 having such a composition is merely one example of an object to be inspected by the wire rope flaw detector 1 of this embodiment, and wire ropes having a different number of strands can also be diagnosed by the wire rope flaw detector 1 of this embodiment in the same manner.

[0039] The wire rope 2 and the strand 23 are made up of a large number of wires 24, and if any breaks occur in the wires 24, the wire rope flaw detector 1 is required to reliably detect the breaks. Furthermore, breaks in the wire 24 vary depending on the location: when they occur on the outer periphery of the strand 23, they are called mountain cuts 25 in the wire 24, and when they occur halfway between adjacent strands, they are called valley cuts 26 in the wire 24. In other words, the valley cuts 26 are breaks in the wire 24 halfway between the outermost and innermost parts of the strand 23 in the circumferential direction of the strand 23, or at a location where adjacent strands 23 come into contact with each other.

[0040] A plurality of magnetic sensors 3 shown in FIG. 1 are installed on a circumference close to the wire rope 2. Representative types of the magnetic sensor 3 include a detection coil and a Hall element, but for example, a tunnel magneto resistive (TMR) sensor, an anisotropic magneto resistive (AMR) sensor, or a giant magneto resistive effect (GMR) sensor can also be used.

[0041] The arrangement of the magnetic sensors 3 constituting the sensor unit 20 may be varied depending on the type of the magnetic sensors 3. In addition, the arrangement of the magnetic sensors 3 is preferably such that multiple sensors are installed at a predetermined interval (pitch) around the outer periphery of the wire rope 2 in order to efficiently detect breaks in the wire 24 using a minimum number of magnetic sensors 3.

[0042] FIG. 2 is a functional block diagram showing the configuration of a wire rope flaw detector 1 according to the first embodiment of the present invention. As shown in FIG. 2, the wire rope flaw detector 1 includes a magnetization mechanism 30, a sensor unit 20, a magnetic sensor circuit (magnetic sensor circuit unit) 5, a signal analysis unit 7, a data display unit 8, and a data input unit 9.

[0043] The data input unit 9 and the data display unit 8 have a general-purpose computer (PC) connected to a power source 10 and a control circuit 11, and the user operates it to control the wire rope flaw detector 1. The control circuit 11 controls the magnetic sensor circuit 5 and the signal analysis unit 7 .

[0044] The magnetization mechanism 30 forms a magnetic path in a predetermined section in the longitudinal direction of the wire. The magnetization mechanism 30 can employ a configuration in which the wire is magnetized by a DC magnetic field using a permanent magnet, or a configuration in which the wire is magnetized in a non-contact manner by an AC magnetic field using an excitation coil.

[0045] The magnetic sensor circuit 5 has a magnetic signal amplifier (magnetic signal amplifier section) 12 and a filter circuit (filter circuit section) 13. The magnetic signal amplifier 12 amplifies the output signal from the magnetic sensor 3. The filter circuit 13 performs general analog filtering on the output signal amplified by the magnetic signal amplifier 12, and outputs an analog signal. The analog filtering removes noise components including commercial frequency noise, and passes signals only in the desired frequency range. In this way, the magnetic sensor circuit 5 performs analog processing on the magnetic detection signal output from the magnetic sensor 3, and outputs an analog magnetic signal to the signal analysis unit .

[0046] The signal analysis unit 7 is composed of an A / D converter (A / D conversion unit) 16, a signal collector (signal collection unit) 17, and a signal processor (signal processing) 18. The A / D converter 16 converts the analog magnetic signal output from the magnetic sensor circuit 5 into a digital signal and outputs it to the signal collector 17. The signal processor 18 includes a one-chip microcomputer, a single board computer, or the like. The signal processor 18 stores the digital magnetic signal output from the A / D converter 16 in the signal collector 17 by the CPU (Central Processing Unit) reading and executing a program stored in a memory or storage. The processing by the signal analysis unit 7 can be realized by program processing in the signal processor 18, and the data display unit 8 and data input unit 9 can also be devices associated with the signal processor 18, such as a keyboard and a liquid crystal display.

[0047] Next, an overview of the analysis process for determining whether or not a wire in a wire rope has been broken in the signal analysis unit 7 of the wire rope flaw detector 1 will be given, and then the process procedure will be described in detail with reference to the drawings.

[0048] The leakage magnetic flux of the wire rope 2 is detected by the magnetic sensor 3 and, as described above, is converted into a digital signal by the A / D converter 16 of the signal analysis unit 7. If the wire rope 2 is damaged, the leakage magnetic flux signal resulting from a broken wire is converted into a digital signal. In a normal wire rope 2 with no broken wires, only the strand signal resulting from the unevenness of the strands 23 is detected. The signal resulting from a broken wire and the strand signal detected in a normal wire rope each have different signal waveform characteristics.

[0049] FIG. 3 is a schematic diagram showing types of feature quantities of a signal waveform. Specifically, the features of a signal waveform (hereinafter also referred to as “signal features”) include, as shown in FIG. 3, amplitude 100, half-width 101, area 102, effective value 103, kurtosis 104, skewness 105, and crest factor 106, which is the ratio of amplitude 100 to effective value 103.

[0050] The signal resulting from a broken wire in an abnormal wire rope and the strand signal detected in a normal wire rope have different signal features. Therefore, by utilizing the difference in signal features, it is possible to determine whether the wire rope is normal or abnormal.

[0051] However, when absolute values ​​of each signal feature are used, differences will arise depending on the type of wire rope and individual differences, so cluster processing using multivariate analysis is necessary. A typical cluster analysis method is the analysis method using the Mahalanobis distance. Specifically, when the signal values ​​in a certain group are expressed by multiple variables x, the average value of each variable in the group is represented by μ, and the covariance matrix of the group is represented by Σ, the Mahalanobis distance Dm(x) with a specific variable is defined by the following formula (1).

[0052]

number

[0053] In addition, the Mahalanobis-Taguchi Method (MT), one of the existing analysis methods, sets a normal product as the unit space, analyzes the Mahalanobis distance from the test specimen, and determines the degree of abnormality of the test specimen.

[0054] The cluster analysis method is not limited to the Mahalanobis distance, and other methods such as Euclidean distance, Manhattan distance, Chebyshev distance, and Minkowski distance may also be used.

[0055] In this embodiment, among the feature amounts of many types of signal waveforms as shown in FIG. 3, a plurality of types of feature amounts are used to perform cluster processing by multivariate analysis. As the multiple types of feature amounts, any combination of two or more types of feature amounts can be used. Preferably, the multiple types of feature quantities include at least the kurtosis 104 shown in Fig. 3. More preferably, the multiple types of feature quantities include any combination of the kurtosis 104 and the amplitude 100, the kurtosis 104 and the half-width 101, and the kurtosis 104 and the area 102 shown in Fig. 3. In this manner, by using a plurality of types of feature quantities including the kurtosis 104, it becomes possible to easily identify the position where a break has occurred in the wire 24 of the wire rope 2.

[0056] In addition, in this embodiment, the signal features previously obtained from a normal wire rope are used as a unit space, and the distance (such as Mahalanobis distance) from the signal features of the test wire rope is calculated, and if there is damage, the level of damage is determined based on the degree of distance from the unit space.

[0057] FIG. 4 is a flowchart of a process for determining whether a wire rope is normal or abnormal according to the first embodiment of the present invention. According to FIG. 4, first, in step S11, under the control of the control circuit 11, a magnetic signal is acquired by the magnetic sensor 3 for a new or normal wire rope, and the signal feature quantity a and the signal feature quantity b are analyzed. Next, in step S12, the signal feature quantity a and the signal feature quantity b of the new and normal wire ropes analyzed in step S11 are created as a unit space in the signal processor 18 and stored. Next, in step S13, a signal of the test wire rope is acquired. Then, from this signal, signal feature quantity a and signal feature quantity b of the test wire rope are acquired. Next, in step S14, it is determined whether the signal feature quantity a and the signal feature quantity b of the wire rope under test are within a reference range. Specifically, the distance (e.g., Mahalanobis distance) between the unit space of the signal feature quantity of the normal wire rope previously stored in step S12 and the point represented by the signal feature quantity a and the signal feature quantity b of the wire rope under test is compared with the reference range. In step S14, if the distance is within the reference range, the process proceeds to step S15, in which the wire rope under test is determined to be a normal product A, and the data display unit 8 displays "normal." If the distance is outside the reference range in step S14, the process proceeds to step S16, where the level of abnormality, B or C, is determined based on the distance value, the data display unit 8 displays that there is an abnormality, and the process proceeds to step S17. At this time, the data display unit 8 displays the level of abnormality (B or C) together with the display that there is an abnormality. In step S17, a process is performed to determine the location of a broken wire in the wire rope determined to have an abnormality level B or C. In this manner, a determination is made as to whether the wire rope is normal or abnormal, and the determination result is displayed on the data display unit 8.

[0058] As the signal feature amount a and the signal feature amount b shown in FIG. 4, any two types of feature amounts can be used.

[0059] 5 is a flowchart of a process for determining a location where a wire break has occurred in a damaged wire rope according to the first embodiment of the present invention. This process shows a processing method for the determination in step S17 shown in FIG. 5, first, in step S21, the signal feature quantities c and d of the test wire rope are acquired. Then, the acquired signal feature quantities c and d are analyzed. Next, in step S22, it is determined whether the signal feature amount c and the signal feature amount d are equal to or greater than a reference range. Specifically, the distance between the unit space based on the signal feature amount c and the signal feature amount d of a normal wire rope stored in advance and the point based on the signal feature amount c and the signal feature amount d is compared with the reference range of distance. In step S22, if the distance is within the reference range, the process proceeds to step S23, in which the data display unit 8 displays a determination that the wire is broken at a valley. In step S22, if the distance is equal to or greater than the reference range, the process proceeds to step S24, in which the data display unit 8 displays a wire break determined to be a break in the wire. In this manner, the location of the wire breakage in the test wire rope is determined.

[0060] For signal feature c and signal feature d shown in Fig. 5, it is desirable to use a combination of features in which at least one feature is different from the combination of features used for signal feature a and signal feature b shown in Fig. 4. It does not matter whether a combination of both features is different, or a combination in which one feature is different and the other feature is the same. This makes it possible to make the unit space and distance for a normal wire rope suitable for detecting the damage level and determining the location where the damage has occurred.

[0061] Preferably, at least one of the signal feature quantities c and d is kurtosis. More preferably, the signal feature quantities c and d are any combination of kurtosis and amplitude, kurtosis and half-width, or kurtosis and area. In this way, by using kurtosis as at least one of the signal feature quantities c and d, the occurrence site of damage can be easily determined.

[0062] The wire rope flaw detector 1 of this embodiment comprises a magnetization mechanism 30 that magnetizes the wire rope 2, and a plurality of magnetic sensors 3 that can detect magnetic signals generated from the wires 24 of the wire rope 2, and further comprises a signal analysis unit 7 that includes a signal collector 17 that collects the magnetic signals output from the magnetic sensors 3, and a signal processor 18 that acquires signal features of the magnetic signals output from the signal collector 17. Then, the signal analysis unit 7 stores multiple signal features previously acquired from a normal wire rope as a unit space, and detects the damage level of the test wire rope based on the distance (such as Mahalanobis distance) between the multiple signal features acquired from the test wire rope and the unit space. Furthermore, the signal analysis unit 7 calculates the distances (Mahalanobis distance, etc.) between a plurality of signal feature amounts and a unit space in the damaged wire rope, and determines the location where the damage has occurred.

[0063] FIG. 6 is a diagram showing an example of the relationship between the values ​​of the signal feature amount a and the signal feature amount b and the normal and abnormal levels according to the first embodiment of the present invention. Specifically, for example, the signal feature of a wire rope classified as normal or judged as A is defined as unit space 40, and the level of normality or abnormality of the wire rope is classified according to the reference value of the Mahalanobis distance for each value of signal feature a and signal feature b. When the values ​​of the signal feature amount a and the signal feature amount b are within the unit space 40, they are determined to be normal in step S14 of FIG. If the values ​​of the signal feature quantity a and the signal feature quantity b are outside the unit space 40, it is determined in step S14 of Fig. 4 that there is an abnormality, and the Mahalanobis distances of the values ​​of the signal feature quantity a and the signal feature quantity b from the unit space 40 are compared with a reference value of the Mahalanobis distance. This results in classification into classification 41 of judgment B or classification 42 of judgment C.

[0064] Next, FIG. 7 is a diagram showing an example of values ​​of the signal feature quantity c and the signal feature quantity d and a damaged portion of the wire according to the first embodiment of the present invention. Specifically, for example, for a wire rope classified into classification 41 of judgment B and classification 42 of judgment C in FIG. 6, a mountain cut 25 occurring on the outer periphery and a valley cut 26 occurring on the inner periphery are determined from the Mahalanobis distance for the values ​​of each signal feature c and signal feature d. Here, depending on the location of damage, differences arise in the distance between the damage site and the magnetic sensor 3 and in the distribution of the leakage magnetic field, and therefore, for example, changes in signal features such as kurtosis are conceivable. Furthermore, the Mahalanobis distance between the signal feature quantity c and the signal feature quantity d is different, so that the classification 44 of a mountain cut and the classification 43 of a valley cut are determined. The result of such determination is displayed on the data display unit 8. The data display unit 8 is, for example, a liquid crystal display connected via an output interface, and displays the results of the processing executed by the signal analysis unit 7, etc., as an analysis result display screen.

[0065] FIG. 8 is a diagram showing an example of a display of the determination result according to the present embodiment. 8 shows a specific example of an analysis result display screen by the data display unit 8. The analysis result display screen displays, for example, the presence or absence of an abnormality in the signal waveform, the type of break (valley break or peak break), the number of breaks, the model number, the analysis date and time, and the like. In addition, although not shown, an indication of when to replace the wire rope estimated from the judgment results may be displayed on the analysis result display screen.

[0066] According to the present embodiment described above, in the wire rope flaw detector 1, the level of the damage state (degree of damage) and the damaged location of the wire rope 2 can be determined by using multivariate analysis based on distance such as the Mahalanobis distance. Furthermore, by using multivariate analysis based on distances such as the Mahalanobis distance, the level of damage to the wire rope 2 and the damaged location can be determined without having to set individual thresholds depending on the product type of the wire rope 2 used in the elevator. Furthermore, according to this embodiment, by determining the damaged portion of the wire rope 2, the cause of the damage can be estimated.

[0067] Furthermore, depending on the location of damage, differences arise in the distance between the damage site and the magnetic sensor 3 and in the distribution of the leakage magnetic field, which may result in changes in signal features such as kurtosis. Therefore, by using a plurality of types of feature quantities including kurtosis as the signal feature quantity c and the signal feature quantity d in FIG. 5 and FIG. 7, the location where damage has occurred can be easily determined.

[0068] Second embodiment A second embodiment of the present invention will now be described with reference to the drawings. In this embodiment, the wire flaw detector and wire diagnostic method of the present invention are applied to a wire used in a handrail of an escalator.

[0069] Handrails are installed on escalators to rotate in conjunction with the steps and to prevent passengers from falling. The handrail is covered with a resin material such as urethane, and multiple steel wires are attached inside the resin material to maintain its strength. If the wire becomes twisted or even breaks due to aging, the broken end of the wire may protrude outside the handrail, which is dangerous. Therefore, it is necessary to periodically check the wire inside the handrail for twists or breaks. It is not possible to monitor the condition of the wires from the outside of the handrail, but by using magnetic leakage flux testing, a non-destructive testing method, it is possible to measure twists and breaks in the internal wires.

[0070] FIG. 9 is a diagram showing the arrangement of wires and sensors according to this embodiment. 9, a sensor unit 60 is placed at a position facing the wire 53 to be inspected. The sensor unit 60 is composed of a pair of magnetization mechanisms (a first excitation coil 50 and a second excitation coil 51) and a magnetic sensor (detection coil) 52 placed in the middle of the two excitation coils 50, 51, and is placed in a row in the extension direction of the wire 53. The first excitation coil 50 and the second excitation coil 51 generate AC magnetic fields in opposite directions to each other, and contactlessly magnetize the wire 53. At this time, the magnetic sensor 52 detects a magnetic signal based on the magnetic fields generated by the first excitation coil 50 and the second excitation coil 51.

[0071] When a normal wire 53 is magnetized by the opposing first excitation coil 50 and second excitation coil 51, the magnetic fields input to the magnetic sensor 52 are in opposite directions and therefore cancel each other out, and are not detected as a magnetic signal. On the other hand, if the wire 53 is twisted or broken, leakage magnetic flux occurs, causing a bias in the magnetic field input to the magnetic sensor 52, which is detected as a magnetic signal. Furthermore, the magnetic signal of the leakage magnetic flux resulting from twisting or disconnection of the wire 53 has signal features similar to the signal features 100 to 105 shown in Fig. 3 of the first embodiment. Therefore, it is possible to determine the degree of twisting or disconnection of the wire 53 by multivariate analysis based on distance (such as Mahalanobis distance).

[0072] FIG. 9 shows one set of excitation coils 50, 51 and magnetic sensor (detection coil) 52, but in practice, multiple sets of excitation coils 50, 51 and magnetic sensor (detection coil) 52 are used. Then, multiple sets of excitation coils 50, 51 and magnetic sensors 52 are arranged in a direction perpendicular to the paper surface of Fig. 9 or in the left-right direction of Fig. 9 (see, for example, JP 2018-4555 A and JP 2021-134081 A). This allows inspection of a wide range of the wire 53 by multiple sets of multiple magnetic sensors 52.

[0073] In addition, the measurement of the normal wire 53 is actually affected by differences in the installation environment of the wire 53. For example, because external noise such as vibration and marker signals for position detection are mixed in, a specific magnetic signal is detected. In this embodiment, a magnetic signal including a disturbance signal accompanying measurement of a normal wire 53 is set as a unit space.

[0074] FIG. 10 is a functional block diagram showing the configuration of a wire flaw detector 61 according to one embodiment of the present invention, which is configured by connecting the sensor unit 60 of FIG. As shown in FIG. 10, the wire flaw detector 61 is configured to include a pair of magnetization mechanisms (first and second excitation coils) 50, 51, a magnetic sensor 52, a magnetic sensor circuit (magnetic sensor circuit section) 65, a signal analysis section 67, a data display section 88, and a data input section 69.

[0075] The data input unit 69 and the data display unit 88 have a general-purpose computer (PC) connected to the power supply 70 and the control circuit 71 , and the user operates it to control the wire flaw detector 61 . The control circuit 71 controls the magnetic sensor circuit 65 and the signal analysis unit 67 .

[0076] The magnetization mechanisms 50 and 51 form a magnetic path in a non-contact manner in a predetermined section in the longitudinal direction of the wire 53. The magnetization mechanisms 50 and 51 magnetize the wire 53 by an AC magnetic field generated by the magnetization mechanisms 50 and 51 in response to an AC frequency input from the control circuit 71 to the magnetization mechanisms 50 and 51.

[0077] The magnetic sensor circuit 65 has a magnetic signal amplifier (magnetic signal amplifier section) 62 and a filter circuit (filter circuit section) 63. The magnetic signal amplifier 62 amplifies the output signal from the magnetic sensor 52. At this time, the AC frequency input from the control circuit 71 to the magnetization mechanisms 50 and 51 is also input to the magnetic signal amplifier 62 as a reference signal, and is synchronized with the output signal from the magnetic sensor 52. The filter circuit 63 performs general analog filtering on the output signal amplified by the magnetic signal amplifier 62, and outputs an analog signal. The analog filtering removes noise components including commercial frequency noise, and passes signals only in the desired frequency range. In this way, the magnetic sensor circuit 65 performs analog processing on the magnetic detection signal output from the magnetic sensor 52 and outputs an analog magnetic signal to the signal analysis section 67.

[0078] The signal analysis unit 67 is composed of an A / D converter (A / D conversion unit) 76, a signal collector (signal collection unit) 77, and a signal processor (signal processing) 78. The A / D converter 76 converts the analog magnetic signal output from the magnetic sensor circuit 65 into a digital signal and outputs it to the signal collector 77. The signal processor 78 includes a one-chip microcomputer, a single board computer, or the like. The signal processor 78 stores the digital magnetic signal output from the A / D converter 76 in the signal collector 77 by having a CPU (Central Processing Unit) read and execute a program stored in a memory or storage. The processing by the signal analysis unit 67 can be realized by program processing in the signal processor 78, and the data display unit 88 and the data input unit 69 can also be devices associated with the signal processor 88, such as a keyboard and a liquid crystal display.

[0079] FIG. 11 is a flowchart of a process for determining whether a wire is normal or abnormal according to the second embodiment of the present invention. According to FIG. 11, first, in step S31, a magnetic signal is acquired by the magnetic sensor 52 for a new or normal wire, and the signal feature amount e and the signal feature amount f are analyzed. Next, in step S32, the signal feature quantity e and the signal feature quantity f of the new and normal wires analyzed in step S31 are created as a unit space in the signal processor 88 and stored. Next, in step S33, a signal from the test wire is acquired, and from this signal, a signal feature amount e and a signal feature amount f are acquired. Next, in step S34, it is determined whether the signal feature amount e and the signal feature amount f are within a reference range. Specifically, the distance (e.g., Mahalanobis distance) between the unit space of the normal wire pre-stored in step S32 and the points represented by the signal feature amount e and the signal feature amount f is compared with the reference range. In step S34, if the distance is within the reference range, the process proceeds to step S35, in which the test wire is determined to be normal product A and the data display unit 88 displays "normal." If the distance is outside the reference range in step S34, the process proceeds to step S36, where the level of abnormality is determined as B or C depending on the distance value, and the data display unit 88 displays that an abnormality is present. Next, in step S37, the result determined as normal A, abnormality level B or C is displayed on the data display unit 88, and the process ends. In this manner, a determination is made as to whether the wire is normal or abnormal, and the determination result is displayed on the data display unit 88.

[0080] As the signal feature amount e and the signal feature amount f shown in FIG. 11, any two types of feature amounts can be used.

[0081] According to the present embodiment described above, in the wire flaw detector 61, the level of the damage state (degree of damage) of the wire 53 can be determined by using multivariate analysis based on distance such as the Mahalanobis distance. By using a multivariate analysis based on distance such as the Mahalanobis distance, the level of damage to the wire 53 can be determined without having to set individual threshold values ​​according to differences in the installation environment of the wire 53.

[0082] The present invention is not limited to the above-described embodiments, and includes various modified examples. For example, the above-described embodiments are described in detail to easily explain the present invention, and the present invention is not necessarily limited to those having all of the configurations described.

[0083] (Modification) In each of the above-described embodiments, two feature amounts are used as the multiple feature amounts to calculate the distance from the unit space, and multivariate analysis is performed based on the distance. In the present invention, three or more feature quantities may be used to calculate the distance to the unit space, and a multivariate analysis based on the distance may be performed to detect the damaged state of the wire. For example, when the Mahalanobis distance is used as the distance, the Mahalanobis distance to the unit space can be calculated even if the number of variables of the multivariate x in the above-mentioned formula (1) is three or more. [Industrial Applicability]

[0084] INDUSTRIAL APPLICABILITY The present invention may be employed as a diagnostic method for a wire rope flaw detector that inspects the hoisting cables of an elevator and a wire flaw detector used in the handrails of an escalator in the safety monitoring of elevators. [Explanation of symbols]

[0085] 1...wire rope flaw detector, 2...wire rope, 3,52...magnetic sensor, 5,65...magnetic sensor circuit, 7,67...signal analysis section, 8,88...data display section, 9,69...data input section, 17,77...signal collector, 18,78...signal processor, 20,60...sensor section, 23...strand, 24...wire, 25...mountain cut, 26...valley cut, 50...first excitation coil (magnetization mechanism), 51...second excitation coil (magnetization mechanism), 53...wire, 61...wire flaw detector

Claims

1. A wire flaw detector for detecting damage to a wire, comprising: a magnetization mechanism for forming a magnetic path in a predetermined section of the wire; and a plurality of magnetic sensors for detecting magnetic signals generated from strands of the wire, a signal analyzer configured to collect magnetic signals output from the magnetic sensor and a signal processor configured to process the magnetic signals output from the signal collector; The signal analysis unit detects damage to the wire by using a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors. A wire flaw detector characterized by:

2. The wire flaw detector of claim 1, characterized in that the signal analysis unit calculates the distance between a unit space created from multiple feature quantities of signal waveforms measured in advance on a normal wire and multiple feature quantities of multiple signal waveforms detected by the multiple magnetic sensors, and detects the degree of damage to the wire under test based on the distance.

3. The wire flaw detector of claim 2, characterized in that when it is determined that the wire of the test subject is damaged, the signal analysis unit calculates the distance between the unit space and multiple feature amounts of multiple signal waveforms detected by the multiple magnetic sensors of the wire of the test subject based on a unit space created from multiple feature amounts of signal waveforms previously measured on a normal wire, and detects the damaged portion of the wire of the test subject by comparing the distance with a threshold value based on distance.

4. The wire flaw detector of claim 3, characterized in that when it is determined that there is damage in the wire of the test subject, the signal analysis unit detects the damaged portion of the wire of the test subject using a plurality of feature quantities including at least a feature quantity indicating kurtosis as the plurality of feature quantities.

5. The wire flaw detector of claim 4, characterized in that when it is determined that there is damage in the wire of the test subject, the signal analysis unit detects the damaged portion of the wire of the test subject using any combination of kurtosis and amplitude, kurtosis and half-width, or kurtosis and area as the multiple feature quantities.

6. 3. The wire flaw detector according to claim 2, wherein the signal analysis unit uses a Mahalanobis distance as the distance.

7. A wire diagnosis method using a wire flaw detector that includes a magnetization mechanism that forms a magnetic path in a predetermined section of the wire and a plurality of magnetic sensors that can detect magnetic signals generated from strands of the wire, and that detects damage to the wire, creating a unit space of a plurality of feature quantities in advance using a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors for a normal wire; acquiring a plurality of feature quantities of each of a plurality of signal waveforms detected by the plurality of magnetic sensors for a wire of a test subject; detecting a distance between the unit space of the normal wire and the wire of the subject based on data of the feature amount of the signal waveform; and detecting a damage state of the wire of the subject based on the distance. A wire diagnostic method comprising:

8. The wire diagnosis method according to claim 7, characterized in that the step of detecting the damaged state of the wire of the subject further includes a step of detecting the degree of damage to the wire of the subject and a step of detecting the damaged part of the wire of the subject, and the step of detecting the damaged part of the wire of the subject uses a plurality of feature quantities including at least a feature quantity indicating kurtosis.

9. The wire diagnosis method according to claim 8, characterized in that the step of detecting a damaged portion of the wire of the subject uses any combination of kurtosis and amplitude, kurtosis and half-width, or kurtosis and area.

10. 8. The wire diagnosis method according to claim 7, wherein in the step of detecting the distance, a Mahalanobis distance is used as the distance.

Images