Wire rope deterioration determination device and wire rope deterioration determination method

Abstract

To provide a wire rope determination device that can determine deterioration of a wire rope more accurately.SOLUTION: The wire rope deterioration determination device includes: a detector 30 for detecting deterioration of a wire rope by a leakage magnetic flux technique; a data input unit 31 for obtaining measurement data from the detector 30; noise component removing means for removing noise components from the data input unit 31; and waveform processing means for extracting data which is determined to be highly relative to what is causing the deterioration of the wire rope.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a wire rope deterioration determination device and a wire rope deterioration determination method, and more particularly to a wire rope deterioration determination device and a wire rope deterioration determination method using a sensor based on a leakage magnetic flux method. [Background technology]

[0002] A conventional wire rope deterioration determination device is disclosed in, for example, JP 2013-035693 A (Patent Document 1). According to this publication, the device has a wire breakage detection unit that detects wire breakage in a wire rope made by twisting together multiple steel wires, and a display unit that displays the diagnosis result of wire breakage in the wire rope according to the data detected by the wire breakage detection unit, and based on a level calculated from the correlation between the wire breakage output detected by the wire breakage detection unit for each certain section and the number of times this output exceeds a threshold, the wire breakage state of the wire rope is divided into stages and displayed on the display unit 9 in correspondence with the elapsed time from the start of measurement. In addition, this device is configured to display the location where the wire breakage has occurred in the measured waveform in correspondence with the elapsed time from the start of measurement on the display unit.

[0003] As a result, an elevator wire rope wire breakage diagnosis system is provided that can easily determine whether the damage to the wire rope is minor or major. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2013-035693 A (Abstract) Summary of the Invention [Problem to be solved by the invention]

[0005] In conventional wire rope deterioration determination devices, for example, when diagnosing a wire rope, a wire breakage detection unit located in the machine room is made to contact the wire rope by passing magnetic flux through it, and the elevator car is allowed to run in this state, thereby detecting wire breakage based on the magnitude of the magnetic flux density of the wire breakage on the surface of the wire rope.

[0006] Conventional wire rope deterioration determination devices are capable of determining whether damage to a wire rope is minor or major, but have not been designed to detect signs of deterioration.

[0007] This invention has been made to solve the problems described above, and its object is to provide a wire rope diagnosis device and method that can identify the cause of deterioration when measuring and diagnosing the deterioration condition of a wire rope. [Means for solving the problem]

[0008] The wire rope deterioration determination device of the present invention includes a sensor that detects wire rope deterioration using a leakage flux method, a data acquisition means that obtains measurement data from the sensor, a noise component removal means that removes noise components from the data, a data extraction means that extracts data that is determined to have a high correlation with factors of wire rope deterioration from the data from which noise has been removed by the noise component removal means, and a deterioration characteristic detection means that detects deterioration characteristics based on the data extracted by the data extraction means.

[0009] Preferably, the measurement data includes a real waveform component and a noise component, and the noise component removing means applies the time series data extracted by the data extracting means to an autoregressive model to separate the real waveform component from the noise component. Note that the noise component may be white noise.

[0010] In one embodiment of the present invention, the data extraction means determines that a location having a high correlation with a reference time series, which is a waveform component due to a reference break, has a high correlation with the deterioration cause of the wire rope.

[0011] The data extracting means may extract an evaluation time series by multiplying the predicted time series by the square of a correlation coefficient with the autoregressive model.

[0012] In another aspect of the present invention, a method for determining deterioration of a wire rope includes detecting leakage magnetic flux data of the wire rope from a sensor using a leakage magnetic flux method, removing noise components from the detected data, extracting data from the noise-removed data that is determined to have a high correlation with factors of deterioration of the wire rope, and detecting deterioration characteristics based on the extracted data. Effect of the Invention

[0013] According to the present invention, a wire rope deterioration determination device and a wire rope deterioration determination method can be provided that can extract data that is highly correlated with the causes of wire rope deterioration and therefore determine wire rope deterioration with greater accuracy. [Brief description of the drawings]

[0014] [Figure 1] 1 is a block diagram of a wire rope deterioration determination device according to one embodiment of the present invention. [Diagram 2] 4 is a flowchart showing a processing procedure of the wire rope deterioration determination device. [Diagram 3] 13 is a diagram showing an example of disconnection detection using a simulated disconnected wire; FIG. 14 is a diagram showing an example of measurement results; [Figure 4] FIG. 11 is a diagram showing an example of a measurement result. [Diagram 5] FIG. 13 is a diagram showing a calculation process required to generate a time series that is determined to have a high relationship with a deterioration cause. [Figure 6] FIG. 1 shows a comparison of an original time series and a predicted time series. [Figure 7] FIG. 13 is a diagram showing a comparison between a predicted time series and a time series multiplied by a correlation coefficient R2. [Figure 8] FIG. 13 is a diagram showing a method of using data for preventive maintenance, showing the probability density distribution of peak amplitudes of an evaluation time series. [Figure 9]FIG. 13 is a diagram showing an estimated value of the probability distribution of peak values ​​extracted when deterioration has progressed considerably. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] An embodiment of the present invention will be described below with reference to the drawings. Fig. 1 is a block diagram showing the overall configuration of a wire rope deterioration determination device 10 according to an embodiment of the present invention. Referring to Fig. 1, the wire rope deterioration determination device 10 includes a control unit 20 including a CPU that controls the entire wire rope deterioration determination device 10, a detector 30 connected to the control unit 20, and a processing unit 40 that processes data detected by the detector 30 by the control unit.

[0016] The detector 30 is provided with a permanent magnet 31 and a coil 32, and a wire rope 50 whose deterioration is to be determined is passed in the vicinity of the coil 32.

[0017] The basic principle of detecting deterioration of the wire rope 50 with the coil 32 is as follows. When the wire rope 50 passes through the detector 30, it is magnetized by the permanent magnet 31. If there is a local reduction in the cross-sectional area due to a break or the like, the leakage magnetic flux changes. When the wire rope 50 passes the coil 32 in the center of the detector 30, the time change in the leakage magnetic flux φ causes an induced voltage E D is generated and detected by the detector 30.

[0018] The processing unit 40 includes a data input unit 41 that inputs data from the detector 30, a noise component removal unit 42 that removes noise components from the input data, a data extraction unit 43 that extracts the data removed by the noise component removal unit 42 from the input data, and a characteristic detection unit 44 that detects the deterioration state of the wire rope based on the data extracted by the data extraction unit 43. The processing of each of these units is performed by an application (not shown) stored in the control unit 20.

[0019] Fig. 2 is a flowchart showing the operation performed by the control unit 20 of the wire rope deterioration determination device 10. Referring to Fig. 2, the control unit 20 inputs detection data via the data input unit 41 (S11), removes noise in the noise component removal unit 42 (S12), extracts data in the data extraction unit 43 (S13), and performs characteristic detection based on the extracted data to detect deterioration of the wire rope (S14).

[0020] 3 is a diagram showing an example of detection of a broken wire using a simulated broken wire. Referring to FIG. 3, five simulated broken wires and the induced voltages (output voltages) (V) in each case are shown.

[0021] 4 is a diagram showing an example of the measurement results extracted by the data input unit 41. The horizontal axis indicates the measurement position, and the vertical axis indicates the standardized induced voltage. In other words, the vertical axis does not indicate the voltage itself, but rather indicates a standardized value obtained by dividing the voltage by the output voltage (about 3 V) at the time when one strand of the wire is broken.

[0022] Figure 4(A) shows the measurement results of a sound wire without any breaks, while Figure 4(B) shows the measurement results of a wire showing signs of deterioration. With particular reference to Figure 4(B), there is noise in the region where the standardized induced voltage is below 0.2, but we believe that there is also a signal output showing signs of deterioration in the part buried in this noise.

[0023] The inventors paid attention to the measurement result data in Figures 3 and 4, and thought that this signal might be related to the cause of deterioration of the wire rope, and devised a method to extract it. Based on that, they completed the basic configuration of the present invention, which is explained in Figures 1 and 2. The specifics will be explained below.

[0024] First, a method for extracting a signal that has a high correlation with a degradation factor, which has been devised by the inventor, will be described.

[0025] The measurement data input to the data input section shows that it contains a mixture of noise components close to white noise, such as drift components that shift the waveform to the positive side overall, and electrical noise, and it is believed that output related to wire rope deterioration is included within this. The measurement data can be expressed as the following formula:

[0026]

number

[0027] where Xm(t) is the total measurement by the detector, X0 is the drift component (positive value), Xtst(t) is the output by the detector (called the "real waveform component"); n(t) is a noise component close to white noise.

[0028] Of these, the drift component can be calculated approximately by taking the time average of the measured waveform. Since the detector responds to the positive side, in reality it is not possible to simply take the time average, but for simplicity's sake, we have considered that it can be approximated by the time average value.

[0029]

number

[0030] where T is the sampling time of the waveform.

[0031] The time series with the drift component removed is expressed by the following equation.

[0032]

number

[0033] Next, we will explain how to separate the waveform components and noise components. A typical method for removing noise is to perform frequency analysis of the time series, remove certain frequency components (low-pass filter), and then use Fourier transform and inverse Fourier transform to return it to the time domain. Since there are some unknowns, such as the selection of the frequencies to be removed, we adopted a method this time to fit the time series to an autoregressive model (hereinafter abbreviated as "AR model") as shown below, and separate the real waveform components from the noise components.

[0034] There are two methods for fitting the AR model: the Burg method and the Yule-Walker method (described in Hino: Spectral Analysis, Asakura Publishing, 1968), but there was little difference between them. Here, we use the Burg method.

[0035]

number

[0036] where a(m) is the autoregressive coefficient (order M), e(t) is the prediction error.

[0037] Here, if optimal prediction is possible with equation (4), e(t) will be white noise that has no correlation with the previous and next time steps. Here, we assume that n(t) ≒ e(t), and the time series based on the autoregressive model given by equation (4) is the output (actual waveform component) from the detector. However, even if a certain amount of noise components have been removed, it is believed that there will still be signals that are not necessarily related to degradation factors, such as noise components that cannot be completely removed and outputs that react to subtle changes in the cross section of the wire.

[0038] Next, a method for extracting a signal from a detector that reacts to a deterioration factor will be described.

[0039] According to data from basic experiments conducted by the inventors, it is known that the output signal reacts to the positive side due to breaks, dents, and floating of the wire rope. Here, the waveform portion due to the reference break (several seconds around the break position) is used as the reference time series, and the correlation coefficient with the time series based on the autoregressive model is calculated while moving this reference time series on the time axis. The time series due to the reference break was obtained from an indoor experiment, and almost no drift or noise components were found mixed in.

[0040] The parts with high correlation with the reference time series are regarded as outputs with high correlation with the deterioration factors, and the square of the correlation coefficient ( R 2 By multiplying the time series generated by the autoregressive model (with the peak output being made to match the time series generated by the autoregressive model), a time series determined to have a high correlation with the degradation factors was generated.

[0041] The above calculation process is shown in Figures 5(A) to 5(H). Figure 5(A) shows the original time series (after removing the drift component), Figure 5(B) shows the frequency analysis result of Figure 5(A) (solid line: power spectrum by FFT method, dashed line: power spectrum by autoregressive model), and Figure 5(C) shows the time series (predicted time series) by the autoregressive model applied to the original time series of Figure 5(A). Here, the AR order is 15.

[0042] In addition, since it is expected that the reproduction of the time series will be affected if the AR order is too small or too large, the inventors changed the order and investigated the degree of influence on the detection of peak values. As a result, it was found that the degree of influence converges if the order is about 5 to 10, so here, a slightly longer order of 15 was adopted.

[0043] FIG. 5(D) shows the difference between the original time series and the predicted time series (values ​​equivalent to e(t)). FIG. 5(E) shows the reference time series based on the reference line break (solid line portion, time width of 2 seconds before and after the line break position). FIG. 5(F) shows the correlation analysis between the reference time series and the predicted time series (vertical axis: correlation coefficient). Rest 2) and Fig. 5(G) shows the predicted time series and the square of the correlation coefficient. Rest 2 4 is a diagram showing the multiplication of (hereinafter referred to as an evaluation time series; in the figure, circles indicate peak detection results).

[0044] As can be seen from FIG. 5(G), characteristic detection is performed to detect deterioration of the wire rope, and outputs highly related to the causes of deterioration of the wire rope are obtained as circles in the figure.

[0045] Figure 6 shows a comparison between the original time series and the predicted time series. It can be seen that the two match well. Figure 7 shows the correlation coefficient for the predicted time series. R 2 The time series is compared with the time series multiplied by . It can be seen that the time series that reacts to the positive side, which is characteristic of the detector, is extracted.

[0046] The inventors have analyzed various other cases of wire rope deterioration in addition to those mentioned above, and extracted time series that responded on the positive side in all cases.

[0047] Furthermore, by comparing a series of data during the ascent and descent of the wire rope, reproducible data was also detected.

[0048] As described above, it was found that by removing noise components from the obtained original time series signal, a signal highly relevant to the causes of wire rope deterioration can be obtained.

[0049] From the above results, the inventors realized that this signal can be used to perform preventive maintenance against breakage of the wire rope, etc.

[0050] Next, the use of data for preventive maintenance against breakage of wire ropes, etc. will be described.

[0051] The method for removing noise components and extracting data that is determined to have a high correlation with deterioration factors was explained above, but it is unclear to what extent the extracted data (evaluation time series) is related to deterioration factors. However, since it contains data with high reproducibility, it is considered to be a data group that shows some kind of sign. From the perspective of preventive maintenance, it is important not only to monitor data that shows a clear reaction, but also to monitor signs that are buried in noise.

[0052] Below, we will show an example of how to use the above data to judge signs. Here, we will explain an example of use focusing on the frequency distribution of the evaluation time series. Here, we assume that the frequency distribution is information related to signs of deterioration. Evaluation can be made using the following probability method.

[0053] Figure 8 shows the probability density distribution of the peak amplitude of the aforementioned evaluation time series. In the figure, the curve obtained by fitting the measured values ​​(○) with a log-normal distribution and the corresponding extreme value distribution (extreme value type II distribution) are also shown. The calculation method for the log-normal distribution and the extreme value distribution are as described in the literature ("Revised Fundamentals of Probability and Statistics for Civil Engineering and Architecture, Maruzen, 2007"). The downward arrow in the figure ↓ is the expected value of the extreme value. If 1,000 data sets that follow the log-normal distribution are obtained, the expected value of the extreme value is about 0.6. In this example, since the output voltage at the time of breakage is set to around 3V, it can be determined that there is no problem in terms of safety even if the maximum value is not measured. Conversely, when the extreme value reaches around 1.5V (assuming a safety factor of 2 for the output voltage of 3V at the time of breakage), it can be considered that the deterioration has progressed considerably (Figure 9).

[0054] 9 is a diagram showing estimated values ​​based on FIG. 8, and shows the probability distribution of the extracted peak values. As degradation progress data, the data in FIG. 8 is doubled.

[0055] In the above embodiment, a case has been described in which the wire rope deterioration determination device 10 has a control unit including a CPU that controls the entire device, but this is not limited to the above. The device may also be configured so that data detected by the detector is stored in a data logger or the like, and then separately processed by a personal computer or the like.

[0056] In the above embodiment, an example of using an AR model has been described as a method for removing noise. However, the present invention is not limited to this. A low-pass filter that performs time-series frequency analysis and removes certain frequency components, or a Fourier transform and inverse Fourier transform that return the data to the time domain may also be used.

[0057] Furthermore, in the above embodiment, an example in which the square of the correlation coefficient is used has been described, but the present invention is not limited to this, and the fourth power or the first power of the correlation coefficient may also be used.

[0058] Although the embodiment of the present invention has been described with reference to the drawings, the present invention is not limited to the illustrated embodiment. Various modifications can be made to the illustrated embodiment within the same scope or equivalent scope of the present invention. [Industrial Applicability]

[0059] According to the present invention, a wire rope deterioration determination device capable of determining deterioration of a wire rope can be provided, and therefore the device can be advantageously used as a wire rope deterioration determination device. [Explanation of symbols]

[0060] 10 wire rope deterioration determination device, 20 control unit, 30 detector, 31 permanent magnet, 32 coil, 40 processing unit, 41 data input unit, 42 noise component removal unit, 43 data extraction unit, 44 characteristic detection unit, 50 wire rope.

Claims

1. A sensor that detects deterioration of a wire rope using a leakage magnetic flux method; A data acquisition means for acquiring measurement data from the sensor; a noise component removing means for removing noise components from the measurement data; a data extraction means for extracting the measurement data determined to have a high correlation with a deterioration factor of the wire rope from the measurement data from which noise has been removed by the noise component removal means; a degradation characteristic detection means for detecting a degradation characteristic based on the measurement data extracted by the data extraction means, the measurement data includes a real waveform component and a noise component; The noise component removing means applies the measurement data to an autoregressive model to separate the real waveform component and the noise component, The data extraction means extracts an evaluation time series by multiplying a predicted time series by the square of a correlation coefficient with an autoregressive model, in this wire rope deterioration determination device.

2. The wire rope deterioration determination device according to claim 1 , wherein the noise component is white noise.

3. The wire rope deterioration determination device according to claim 1 or 2, wherein the data extraction means determines that a location having a high correlation with a reference time series, which is a waveform component due to a reference break, has a high correlation with a deterioration factor of the wire rope.

4. Detecting leakage magnetic flux data of the wire rope from a sensor using a leakage magnetic flux method; Remove noise components from the detected data, From the noise-removed data, data that is determined to have a high correlation with the deterioration factors of the wire rope is extracted, A method for determining deterioration of a wire rope, comprising: detecting deterioration characteristics based on the extracted data; the leakage magnetic flux data includes a real waveform component and a noise component, A method for determining deterioration of a wire rope, in which noise components are removed by fitting the leakage magnetic flux data to an autoregressive model to separate actual waveform components from noise components, and the leakage magnetic flux data is extracted by extracting an evaluation time series obtained by multiplying a predicted time series by the square of the correlation coefficient with the autoregressive model.

Images