JP4616695B2 - Multi-sensor signal abnormality detection apparatus and method - Google Patents

Description

  The present invention relates to a multi-sensor signal abnormality detection apparatus and method, and more particularly, to a multi-sensor signal abnormality detection apparatus and method capable of detecting an abnormality of each detection signal acquired by a multi-sensor composed of a plurality of sensors. .

  Conventionally, eddy current testing (ECT), which is one of non-destructive inspections, is widely known for flaw detection inspection of heat transfer tubes and the like. A multi-sensor probe has a large number of sensors arranged in one probe for speeding up the inspection. If one of the many sensors is abnormal, the inspection needs to be repeated. Anomaly detection with high reliability is important.

  As a method of detecting an abnormality in the sensor signal, a standard signal is generated and it is checked whether or not a specific item representing the feature of the standard signal is within a prescribed allowable range. For example, in the ECT signal, an abnormality of the sensor signal is detected by checking whether or not the amplitude and the phase angle of the normalized signal obtained by standardizing the signal generated by the standard defect are within a predetermined allowable range. .

  However, in the conventional signal abnormality detection method, for example, relative to a standard signal caused by electrical noise generated at the time of flaw detection, metal contact with a standard defect, or characteristic abnormality of a coil constituting the ECT probe, etc. It is difficult to detect an anomaly that appears only as a small change signal.

  In relation to the above technique, the following technique has been proposed.

  In “Non-destructive inspection data extraction device and collection system using this device” disclosed in Japanese Patent Application Laid-Open No. 2004-163288, the phase angle and amplitude of a signal obtained by standardizing a waveform generated by a standard defect are within a specified range. There has been proposed a method for detecting an anomaly by checking whether or not there is.

JP 2004-163288 A

  The objective of this invention is providing the multisensor signal abnormality detection apparatus and method which can detect abnormality of each detection signal acquired by the multisensor comprised by a some sensor.

  In the following, means for solving the problem will be described using reference numerals with parentheses used in [Best Mode for Carrying Out the Invention]. These symbols are added in order to clarify the correspondence between the description of [Claims] and the description of the best mode for carrying out the invention. ] Should not be used for interpretation of the technical scope of the invention described in the above.

  The multi-sensor signal abnormality detection device (1) of the present invention includes an input unit (3) for reading a detection signal obtained by a multi-sensor probe (5) having a plurality of sensors, and a signal for processing the detection signal. A processing unit (2), and an output unit (4) for outputting a processing result by the signal processing unit. The signal processing unit reads a detection signal obtained by the multi-sensor probe via the input unit, and A standardized signal is generated by a standardization process that equalizes the phase angle and amplitude of the standard signal due to simulated defects in all of the sensors in the sensor, and uniform components are extracted from the standardized signal among multiple sensors. A uniform component removal signal is generated by component removal processing, and anomalies in the detection signal obtained by each of multiple sensors are detected by checking the singular value of the defining factor of the uniform component removal signal. That.

  In addition, the signal processing unit (2) in the multi-sensor signal abnormality detection device (1) of the present invention passes through a part where there is no significant signal generation factor between the normalization process and the uniform component removal process. Offset processing is performed on the detection signals obtained from each of the plurality of sensors after normalization processing so that the values of the detection signals obtained from the plurality of sensors (5a) are constant values.

  Further, the multi-sensor signal abnormality detection device (1) of the present invention is caused by a change in an external environmental factor such as a temperature change or a change in the distance between a plurality of sensors (5a) and detection of a simulated defect. The signal processing unit (2) includes a drift removal filter for removing drift noise, and the signal processing unit (2) uses a drift removal filter between the normalization process and the uniform component removal process to detect a plurality of sensors after the normalization process. From each of the detection signals obtained, drift removal processing is performed to remove drift noise caused by changes in external environmental factors such as temperature changes, and changes in the distance between a plurality of sensors and the subject during detection signal acquisition.

  In the multi-sensor signal abnormality detection device (1) of the present invention, as a singular value check, the amplitude of the detection signal obtained by each of the plurality of sensors (5a) after removal of the uniform component is equal to or greater than a specified threshold value. Find out.

  In addition, the signal processing unit (2) in the multi-sensor signal abnormality detection device (1) of the present invention, as a singular value check, detects all of the detection signals of each of the plurality of sensors (5a) after the uniform component removal. The magnitude of the noise level is obtained, and the uniform component of the detection signal from each of the plurality of sensors with respect to the magnitude of the noise level is compared with the magnitude after removal.

  The signal processing unit (2) in the multi-sensor signal abnormality detection device (2) of the present invention has a configuration in which a plurality of sensors (5a) are arranged in multiple stages in the traveling direction in which the detection signal is acquired. The uniform component removal processing is performed for each stage of the multistage array.

  In addition, the signal processing unit (2) in the multi-sensor signal abnormality detection device (1) of the present invention includes a standard signal generation factor part of detection signals obtained by each of the plurality of sensors (5a) before the normalization process. Anomalies are detected by performing a singular value check on the prescribed characteristics of the signal that occurs when passing through the.

  Moreover, the multi-sensor signal abnormality detection system (100) of this invention is equipped with the multi-sensor signal abnormality detection apparatus (1) of any one of Claim 1-8.

  In addition, the multi-sensor signal abnormality detection method of the present invention is for processing an input unit (3) for reading a detection signal obtained by a multi-sensor probe (5) having a plurality of sensors (5a) and a detection signal. A multi-sensor signal abnormality detection method by a multi-sensor signal abnormality detection device (1) having a signal processing unit (2) of the above and an output unit (4) for outputting a processing result by the signal processing unit, the input unit A detection signal reading step for reading the detection signal obtained by the multi-sensor probe via the detection signal, and a phase angle and an amplitude of a standard signal due to a simulated defect in all of the plurality of sensors with respect to the detection signal read by the detection signal reading step Are generated by a standardized signal generation step that generates a standardized signal by a standardization process that aligns the Uniform component extraction signal generation step for generating a uniform component extraction signal by uniform component extraction processing for extracting a uniform component among a plurality of sensors from the normalized signal, and uniform component extraction An abnormality detection step of detecting an abnormality of the detection signal obtained by each of the plurality of sensors by performing a singular value check of a defining factor of the uniform component removal signal generated by the signal generation step.

  In addition, the multi-sensor signal abnormality detection method of the present invention includes a plurality of sensors (5a) when passing through a portion having no significant signal generation factor between the normalized signal generation step and the uniform component removal signal generation step. An offset step for performing offset processing on the detection signals obtained from each of the plurality of sensors after the normalization processing is provided so that the value of the detection signal obtained from each of the sensors becomes a constant value.

  In the multi-sensor signal abnormality detection method of the present invention, the multi-sensor signal abnormality detection device (1) further simulates a change in external environmental factors such as a temperature change or a plurality of sensors (5a) during detection signal acquisition. A drift removal filter for removing drift noise caused by a change in the distance to the defect is provided, and each of the plurality of sensors after the normalization process is performed between the normalized signal generation step and the uniform component removal signal generation step. Is provided with a drift removing step for removing drift noise caused by a change in external environmental factors such as a temperature change or a change in the distance between the plurality of sensors and the simulated defect during acquisition of the detection signal.

  In the multi-sensor signal abnormality detection method of the present invention, as a singular value check, it is checked whether the amplitude after removal of the uniform component of the detection signal obtained by each of the plurality of sensors (5a) is equal to or greater than a specified threshold value. A threshold determination step.

  In addition, the multi-sensor signal abnormality detection method of the present invention is a noise level calculation for determining the noise level of all the sensors in the detection signal by each of the plurality of sensors (5a) after removing the uniform component as a singular value check. And a relative comparison step of relatively comparing the magnitude of the noise level obtained in the noise level calculation step and the magnitude after removal of the uniform component of the detection signal from each of the plurality of sensors.

  Further, the multi-sensor signal abnormality detection method of the present invention has a configuration (5a1, 5a2) in which a plurality of sensors (5a) are arranged in multiple stages in the traveling direction in which detection signals are acquired. A multi-stage sensor processing step for performing a uniform component removal signal generation step for each of them.

  In addition, the multi-sensor signal abnormality detection method according to the present invention is a method for detecting a signal generated when passing through the standard signal generation factor part of the detection signal obtained by each of the plurality of sensors (5a) before the standardized signal generation step. A singularity discrimination step for performing a singular value check of the specified feature is provided.

  A multi-sensor signal abnormality detection program according to the present invention realizes the multi-sensor signal abnormality detection method according to claims 9 to 15.

  With the multi-sensor abnormality detection device and method of the present invention, each detection signal acquired by a multi-sensor composed of a plurality of sensors is normalized based on a standard signal, and then zero-point correction and uniform component removal are performed. By performing an abnormality check on the processed signal, it is possible to detect a minute abnormality generated in combination with the standard signal. Further, by adding a drift removal process, it is possible to detect an abnormality having an amplitude smaller than the drift component. Further, by performing a relative comparison with the noise level as a singular value check, a fine abnormality can be detected without excessive detection. Further, the abnormality detection effective for the multi-sensor probe having a plurality of stages can be performed by the uniform component extraction process and the singular value check process for each stage according to the sensor arrangement. Also, abnormality of sensor characteristics can be detected by singular value check between sensors of standard signals before normalization.

  Due to the above effects, the present invention can provide a multi-sensor abnormality detection apparatus and method capable of detecting an abnormality of a signal acquired by a multi-sensor composed of a plurality of sensors with high accuracy.

  With reference to the attached drawings, a best mode for carrying out a multi-sensor signal abnormality detection apparatus and method according to an embodiment of the present invention will be described below. In the multi-sensor signal abnormality detection device according to the embodiment of the present invention, an ECT signal for a simulated defect or the like previously added to a test piece portion is acquired in a flaw detection inspection of a heat transfer tube or the like, and this acquired signal is used as a specified factor signal. And a signal to be used as a standard for standardization among the prescribed factor signals is a standard signal. And the following signal processing is implemented about each of the said standard signal acquired by the some sensor with which a sensor probe is equipped, or a prescription | regulation factor signal.

(1) Based on the standard phase angle and standard amplitude of the standard signal determined in advance, the phase angle adjustment and amplitude adjustment are performed on the signals in all sensors, and the phase angle and amplitude of the standard signals in all sensors are determined. Perform standardization processing to align the same. Then, an offset process is performed with respect to all the standardized sensor signals so that the signal value of a part having no significant signal generation factor is a constant value.

(2) A uniform component between the sensors (for example, the median of the signal value of each sensor) of the signals acquired by all the sensors subjected to the normalization processing is obtained and obtained from the respective detection signals subjected to the normalization processing. Remove the uniform component.

(3) By checking whether the magnitudes of the defined factor signals obtained by all sensors that have performed the process of removing the uniform component are singular values, the case where there is no singular value is regarded as normal. The case where there is even one singular value is regarded as abnormal.

  By the above-described processing by the multi-sensor signal abnormality detection device according to the embodiment of the present invention, the abnormality of the multi-sensor signal that is conventionally missed as a standard signal or only a relatively small change with respect to the specified factor signal is overlooked. Can be reliably detected. Thereby, the abnormality detection performance acquired by the multisensor can be dramatically improved.

(Embodiment 1)
FIG. 1 shows a schematic configuration of a multi-sensor signal abnormality detection apparatus according to Embodiment 1 of the present invention. The multi-sensor signal abnormality detection device 1 according to the present embodiment includes an input unit 3 for reading a detection signal obtained by a multi-sensor probe having a plurality of sensors, a signal processing unit 2, and a process by the signal processing unit 2. And an output unit 4 for outputting the result.

  As a reference, FIG. 2 shows a form in which a flaw detection inspection of a heat transfer tube in ECT is performed using the sensor probe 5. The sensor probe 5 includes a multi-sensor 5a in the head portion. The multi-sensor 5a surface is arranged so as to be parallel to the wall surface of the subject 40, and the signal is detected by the sensor probe 5 moving in the direction of the axis 5b. FIG. 3 shows a schematic configuration when the multi-sensor 5a is arranged in two stages as an example. As shown in FIG. 3, the multi-sensor 5a has an upper sensor 5a1 and a lower sensor 5a2 in the axial direction.

  FIG. 4 shows a schematic diagram for constructing a Lissajous waveform with a signal acquired by the multi-sensor 5a of the sensor probe 5. FIG. The signal acquired by each sensor of the sensor probe 5 includes an X signal that is a complex real axis component and a Y signal that is a complex imaginary axis component. And a Lissajous waveform is constituted by plotting a specific part (waveform of a specified time zone) of each time series waveform on the XY plane.

(Operation Principle in Embodiment 1)
FIG. 5 shows the operating principle of the multi-sensor signal abnormality detection apparatus 1 according to the present embodiment. The input unit 3 reads the standard signal obtained by each sensor of the multi sensor 5a (step S01). Since the raw data (original signal) of the standard signal obtained by each sensor of the multi-sensor 5a has variations in sensitivity and phase depending on each sensor, the raw data of the read standard signal is a predetermined standard signal. The standard angle processing and the amplitude adjustment are performed on the signals of all the sensors so that the standard phase angle and the standard amplitude are equal to each other, and the normalization process is performed to make the phase angles and the amplitudes of all the sensors the same (step S02). ). Next, the uniform component removing unit 2b provided in the signal processing unit 2 of the present embodiment performs zero point correction on the standard signals acquired by the respective sensors and subjected to the normalization processing (step S03). . Here, the zero point correction, a pretreatment for Hikisa uniform Ingredients described below, the value of the signal obtained from the multi-sensor 5a respectively at positions where there is no significant signal generating factor is constant The offset processing is performed on the standard signal after the normalization processing so that the value is obtained. After the zero point correction is performed, a uniform component is obtained using the standard signal with the zero point corrected. Here, for example, a median (median value) signal of the signal value of the multisensor 5a is used as the uniform component. Then, the obtained uniform component is extracted from each standard signal that has been zero-point corrected (step S04). After performing the process of removing the uniform component, the magnitude of the standard signal from each sensor from which the uniform component has been removed is compared with a specified threshold value. For all the sensors, the case where the value of the standard signal is smaller than the prescribed threshold value is normal, and the case where there is one standard signal larger than the threshold value is abnormal (step S05). When there is an abnormality determination, the output unit 4 outputs and notifies the output result.

  FIG. 6 shows an example of a model of a series of signal processing in the present embodiment. As shown in FIG. 6, normalization processing based on the phase angle and amplitude adjustment, zero point correction, and uniform component removal processing are performed on the signals acquired by the respective sensors of the multi-sensor 5a. As a result, the signal level after the signal processing of the standard signal acquired by each sensor becomes a small value within a prescribed threshold value unless the sensor signal is abnormal.

  Each component of the signal processing unit 2 and the normalization unit 2a, the uniform component removal unit 2b, the singular value detection unit 2c, and the like included in the signal processing unit 2 of the present embodiment is configured by a computer. The constituent units such as the normalization unit 2a, the uniform component removal unit 2b, the singular value detection unit 2c, and the like are read and executed by the respective arithmetic units by reading execution programs stored in the respective storage units. The predetermined signal processing in is performed. Similarly, the signal processing unit 2 performs input / output control of detection signals and processing signals to the respective components such as the normalization unit 2a, the uniform component removal unit 2b, and the singular value detection unit 2c.

  According to the present embodiment, it is possible to reliably detect an abnormality of a multi-sensor signal that has been missed without being caught as a relatively small change with respect to a standard signal. Thereby, the abnormality detection performance of standard signals acquired by a plurality of sensors can be dramatically improved.

(Embodiment 2)
The basic schematic configuration and operation principle of the multi-sensor signal abnormality detection device 1 according to the second embodiment of the present invention are the same as those in the first embodiment. However, the uniform component removing unit 2b of the signal processing unit 2 according to the present embodiment changes external environmental factors such as temperature changes, changes in the distances between the plurality of sensors 5a and the subject 40 during detection signal acquisition, and the like. Are provided with a drift removal filter for removing drift noise caused by the above and a drift removal section (not shown) for controlling the drift removal filter.

(Operation principle in the second embodiment)
FIG. 7 shows an operation principle of the multi-sensor signal abnormality detection device 1 according to the present embodiment. The steps up to step S02 of the multi-sensor signal abnormality detection device 1 according to the present embodiment are the same as those in the first embodiment, and the description thereof is omitted here.

In step S02, the signal processing unit 2 according to the present embodiment performs drift removal processing on the signals acquired by the respective sensors and subjected to normalization processing (step S10). Here, the drift removal, a pre-treatment for Hikisa uniform Ingredient performed in step S04, processing is performed in the uniform component Hikisa portion 2b. As described above, the signal processing unit 2 provided in the multi-sensor signal abnormality detection device 1 according to the present embodiment has a drift removal filter (not shown). The drift removal filter has a function of removing drift noise caused by changes in external environmental factors such as temperature changes and changes in the distance between the plurality of sensors 5a and the subject 40 during detection signal acquisition. The drift removal unit of the signal processing unit 2 uses a drift removal filter to detect changes in external environmental factors such as temperature changes and detection signal acquisition from detection signals sent from each of the plurality of sensors after normalization processing. A drift removal process for removing drift noise caused by a change in the distance between the plurality of sensors and the subject 40 is performed. As the drift removal filter, median filter removal or the like for removing drift noise is used. The signal processing after step S10 in the present embodiment is the same as steps S04 and S05 in the first embodiment.

  The multi-sensor signal abnormality detection device 1 according to the present embodiment is particularly susceptible to noise caused by changes in the relative distance between the subject 40 and the multi-sensor 5a and noises caused by changes in external environmental factors such as temperature. This is effective when the inspection is performed using the sensor probe 5 having the above-described sensors.

  According to the present embodiment, it is possible to reliably detect an abnormality in a multi-sensor signal that has been conventionally missed without being caught as a relatively small change with respect to drift noise. Thereby, the abnormality detection performance of standard signals acquired by a plurality of sensors can be dramatically improved.

(Embodiment 3)
The basic schematic configuration and operation principle of the multi-sensor signal abnormality detection device 1 according to the third embodiment of the present invention are the same as those in the first embodiment. However, the signal processing unit 2 of the present embodiment is obtained by a noise level calculation unit (not shown) that calculates a noise level that is a variation component between sensors, and the obtained noise level magnitude and the sensor. A relative value calculation unit (not shown) that obtains a relative value that is a ratio of the magnitude of the signal after the uniform component extraction; Further, between step S04 and step S05 in the first embodiment, a process of calculating a noise level that is a variation component between sensors, and the calculated noise level and the uniform obtained by each sensor. A process of obtaining a relative value that is a ratio to the magnitude of the standard signal after component removal is performed.

(Operation Principle in Embodiment 3)
FIG. 8 shows the operating principle of the multi-sensor signal abnormality detection device 1 according to the present embodiment. The steps up to step 04 of the multi-sensor signal abnormality detection apparatus 1 according to the present embodiment are the same as those in the first embodiment, and thus the description thereof is omitted here.

  In the signals acquired by the respective sensors after the uniform component removal processing, only the variation component between the sensors remains when there is no abnormality, but the noise level differs depending on the sensor probe 5 or the same sensor probe 5 However, the magnitude (noise level) of the variation component between the sensors varies because the attitude of the sensor changes for each flaw detection. For this reason, when the threshold value of the standard signal in each sensor after the process of step S04 is determined as an absolute value, it is necessary to determine an optimum threshold value each time.

  In this embodiment, in order to avoid this, as shown in the detailed processing flow of the part in FIG. 9, noise between sensors is obtained for each signal acquired by each sensor after the uniform component removal processing. The level magnitude (dispersion, median value, etc.) is calculated (step S20), and the relative magnitude of the standard signal obtained by each sensor after the uniform component removal processing with respect to the calculated noise level magnitude is calculated. A value is calculated (step S30).

  Then, after completion of step S30, each relative value is compared with a specified threshold value to determine whether each signal is abnormal (step S05A). In the present embodiment, in the threshold determination based on the absolute value of the signal, it is possible to detect an abnormality of the signal that was overlooked at the time of measurement with a low noise level by comparing the relative value with the threshold.

  According to the present embodiment, it is possible to reliably detect an abnormality of a multi-sensor signal that has been missed without being caught as a relatively small change with respect to a standard signal, regardless of the noise level. Thereby, the abnormality detection performance of standard signals acquired by a plurality of sensors can be dramatically improved.

(Embodiment 4)
The basic schematic configuration and operation principle of the multi-sensor signal abnormality detection device 1 according to the fourth embodiment of the present invention are the same as those in the third embodiment. However, the signal processing unit 2 of the present embodiment includes a sensor stage selection unit (not shown) that selects a sensor at an arbitrary stage of the multi-sensor 5a arranged in multiple stages with respect to the traveling direction of the sensor probe 5. Yes. Then, sensor stages are sequentially selected by the sensor stage selection unit, and step S04, step S20, and step S30 in the third embodiment are repeatedly performed for each selected sensor stage. When step S04, step S20, and step S30 are completed in all sensor stages, step S05A is performed as in the third embodiment, and the operation flow in the present embodiment is terminated.

(Operation Principle in Embodiment 4)
FIG. 10 shows the operation principle of the multi-sensor signal abnormality detection apparatus 1 according to the present embodiment. Steps S01 to S03 and step S05A of the multi-sensor signal abnormality detection device 1 according to the present embodiment are the same as those in the third embodiment, and the description thereof is omitted here.

  As shown in the present embodiment, when sensors for signal acquisition are arranged in multiple stages, probe postures (eccentricity and inclination) when passing through standard factors are not necessarily the same for each stage, so between sensors at different stages Then, there may be a difference in the standard signal waveform. In addition, because the sensor feed rate is not constant, the time required to pass the same position varies depending on the sensor at each stage. There is a possibility of deviation. In the present embodiment, as shown in FIG. 11, the uniform component removal process (step S04), the noise level calculation process (step S20), and the relative value calculation process (step S30) are selected. This is performed only for the sensors in the first stage. Thereby, it is possible to detect an abnormality without adding a difference in the standard waveform due to a difference in sensor arrangement to the noise level. Then, this series of processing is sequentially performed on all the stages of sensors (loop processing between step S40a and step S40a).

  Thereby, in this Embodiment, a highly accurate signal abnormality detection can be implemented also in the multi-sensor signal arranged in multiple stages.

(Embodiment 5)
The basic schematic configuration of the multi-sensor signal abnormality detection device 1 according to the fifth embodiment of the present invention is the same as that of the first to fourth embodiments. The operation principle can be applied to each of the first to fourth embodiments. That is, the signal processing unit 2 of the present embodiment detects a sensor signal abnormality detection unit (not shown) that detects whether there is an abnormality in the waveform characteristics of the standard signal (original signal) acquired by each sensor of the sensor probe 5. ) And before the step S02 in each embodiment, it is detected whether there is an abnormality in the waveform characteristics of the standard signal (original signal) acquired by each sensor of the sensor probe 5.

(Operation Principle in Embodiment 5)
FIG. 12 shows a case where the operation principle of the multi-sensor signal abnormality detection device 1 according to the present embodiment is applied to the first embodiment. Step S01 of the multi-sensor signal abnormality detection device 1 according to the present embodiment and step S02 and subsequent steps are the same as those in the first embodiment, and description thereof is omitted here.

  Based on the original signal obtained in step S01, the defining characteristics (phase angle and amplitude) of each defining factor signal are measured. Investigate the variation of the value obtained by subtracting the standard value (for example, median) between the coils from the measured characteristic value (see Fig. 4), and there is an abnormality if there is at least one sensor whose relative value to the variation exceeds the defined value (Step S50). If an abnormality is determined based on the determination result of step S50, the multi-sensor signal abnormality detection ends there. If it is normal, the process proceeds to signal processing in step S02.

  In the operation principle according to the present embodiment, only the case where it is applied to the first embodiment has been described. However, the present embodiment is not limited to the first embodiment but is also applied to the second to fourth embodiments. Is possible. In the present embodiment, even a sensor signal such as a sensor characteristic abnormality that cannot be identified by the standardized waveform of the standard signal can be identified by examining the abnormality with the characteristics of the waveform of the original signal.

(Embodiment 6)
FIG. 13 shows a schematic configuration of a multi-sensor signal abnormality detection system according to Embodiment 6 of the present invention. The multi-sensor signal abnormality detection system 100 according to the present embodiment includes an input unit 300 for reading a detection signal, a signal processing device 200, and an output unit 400 for outputting a processing result by the signal processing device 200. .

(Operation Principle of Embodiment 6)
The basic configuration and operation principle of the present embodiment are the same as those of the multi-sensor signal abnormality detection device 1 described in the first to fifth embodiments, and the standardization device 200a provided in the signal processing device 200 is uniform. The component removal device 200b and the singular value detection device 200c correspond to the normalization unit 2a, the uniform component removal unit 2b, and the singular value detection unit 2c of the signal processing unit 2 of Embodiments 1 to 5, respectively. However, a large amount of acquired data acquired by the multisensor is input to the input unit 300 of the present embodiment via a network or a storage medium. The signal processing result is output to the output unit 400 as necessary.

  The signal processing device 200 according to the present embodiment, the normalization device 200a, the uniform component removal device 200b, and the singular value detection device 200c provided in the signal processing device 200 are each configured by a computer. The normalization device 200a, the uniform component removal device 200b, and the singular value detection device 200c execute the execution programs stored in the respective storage units by the arithmetic units of the respective devices. Thereby, predetermined signal processing in each device is executed. Similarly, the signal processing device 200 executes input / output control of signals to the normalization device 200a, the uniform component removal device 200b, and the singular value detection device 200c.

  Due to the above-described configuration and operation principle, the multi-sensor signal abnormality detection system 100 according to the present embodiment is a relatively small change with respect to a standard signal with respect to a large amount of multi-sensor signals read from the input unit 300 in the past. However, it is possible to reliably detect an abnormality in a multi-sensor signal that has been overlooked without being caught, and dramatically improve the abnormality detection performance of a standard signal acquired by a plurality of sensors. Thereby, a multi-sensor signal abnormality detection system with high flaw detection accuracy can be realized.

It is a figure which shows schematic structure of the multi sensor signal abnormality detection apparatus of this application. It is a figure which shows the usage pattern of the sensor probe in Embodiment 4. FIG. It is a figure which shows the structure of the multi sensor of the sensor probe in FIG. It is the figure which showed the model of the signal processing which forms the Lissajous waveform in this application. FIG. 3 is a diagram showing an operation flow of the first embodiment. FIG. 4 is a diagram illustrating a schematic diagram of signal processing in the first embodiment. FIG. 10 is a diagram illustrating an operation flow of the second embodiment. FIG. 10 is a diagram showing an operation flow of the third embodiment. FIG. 10 is a diagram showing details of noise level calculation and relative value calculation in the operation flow of the third embodiment. FIG. 10 is a diagram illustrating an operation flow of the fourth embodiment. FIG. 10 is a diagram illustrating a schematic diagram of a detection signal obtained by a sensor selection function in a fourth embodiment. FIG. 10 is a diagram showing an operation flow of the fifth embodiment. FIG. 10 is a diagram showing a schematic configuration of a sixth embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Multi sensor signal abnormality detection apparatus 2 ... Signal processing part 2a ... Standard part 2b ... Uniform component removal part 2c ... Singular value detection part 3 ... Input part 4 ... Output part 5 ... Sensor probe 5a ... Multi sensor 5b ... Shaft 40 ... Subject 100 ... Multi-sensor signal abnormality detection system 200 ... Signal processing device 200a ... Standard part 200b ... Uniform component removal part 200c ... Singular value detection part 300 ... Input part 400 ... Output part

Claims (17)

An input unit for reading a detection signal obtained by a multi-sensor probe having a plurality of sensors;
A signal processing unit for processing the detection signal;
An output unit for outputting a processing result by the signal processing unit,
The signal processing unit reads the detection signal obtained by the multi-sensor probe via the input unit, and calculates a phase angle and an amplitude of a standard signal due to a simulated defect in all of the plurality of sensors with respect to the detection signal. A standardization signal is generated by a standardization process that aligns to the same, and a preprocessing for a uniform component extraction process that extracts a uniform component among the plurality of sensors from the standardization signal has been standardized. This is performed on a signal, a uniform component removal signal is generated by the uniform component removal processing, and an abnormality of the detection signal is detected by performing a singular value check of a defining factor of the uniform component removal signal. Multi-sensor signal abnormality detection device. In the multi-sensor signal abnormality detection device according to claim 1,
The pretreatment for the uniform component removal process is:
The signal processing unit, between the uniform component Hikisa process and the normalization process, the value of the detection signal obtained from each of the plurality of sensors when passing through the site has no significant signal generating factor The multi-sensor signal abnormality detection device is to perform offset processing on the detection signals obtained from each of the plurality of sensors after the normalization processing so that becomes a constant value. In the multi-sensor signal abnormality detection device according to claim 1,
In addition, equipped with a drift removal filter to remove drift noise,
The pretreatment for the uniform component removal process is:
The signal processing unit, between the uniform component Hikisa process and the normalization process, by the drift removal filter, from the detection signal obtained by each of the plurality of sensors after the normalizing process, it is to perform the drift removal processing for removing the drift noise <br/> multi-sensor signal abnormality detection device. In the multi-sensor signal abnormality detection device according to claim 1,
As the singular value check, a multi-sensor signal abnormality detection device that checks whether an amplitude after the uniform component removal of the detection signal obtained by each of the plurality of sensors is equal to or greater than a predetermined threshold. In the multi-sensor signal abnormality detection device according to claim 1,
The signal processing unit obtains noise level magnitudes of all sensors of the detection signals by the plurality of sensors after the uniform component removal as the singular value check, and the noise level magnitudes with respect to the noise level magnitudes A multi-sensor signal abnormality detection device that compares the uniform component of the detection signal of each of a plurality of sensors with a magnitude after removal. In the multi-sensor signal abnormality detection device according to claim 1,
The signal processing unit performs the uniform component extraction processing for each stage of the multi-stage arrangement when the plurality of sensors have a multi-stage arrangement with respect to a traveling direction in which the detection signal is acquired. Multi-sensor signal abnormality detection device. In the multi-sensor signal abnormality detection device according to claim 1,
The signal processing unit performs a singular value check of a prescribed characteristic of a signal generated when passing through a standard signal generation factor unit of the detection signal obtained by each of the plurality of sensors before the normalization processing. Multi-sensor signal abnormality detection device for detecting abnormality by means of In the multi-sensor signal abnormality detection device according to at least one of claims 1 to 7,
Each of the plurality of sensors is a multi-sensor signal abnormality detection device which is a sensor for ECT (Eddy Current Test).   A multi-sensor signal abnormality detection system comprising the multi-sensor signal abnormality detection device according to claim 1. An input unit for reading a detection signal obtained by a multi-sensor probe having a plurality of sensors, a signal processing unit for processing the detection signal, and an output unit for outputting a processing result by the signal processing unit A multi-sensor signal abnormality detection method by a multi-sensor signal abnormality detection device comprising:
A detection signal reading step of reading the detection signal obtained by the multi-sensor probe via the input unit;
A standardized signal for generating a standardized signal by a standardization process in which the phase angle and the amplitude of the standard signal due to simulated defects and the like in all of the plurality of sensors are made equal to the detection signal read in the detection signal reading step Generation step;
A pre-processed signal is generated by pre-processing for uniform component removal processing for extracting uniform components among the plurality of sensors from the normalized signal generated by the normalized signal generation step. A pretreatment step for removal of the like component;
Uniform component removal by uniform component removal processing that removes uniform components among the plurality of sensors from the preprocessed signal generated by the preprocessing step for the uniform component removal processing. A uniform component removal signal generating step for generating a signal;
A multi-sensor signal abnormality comprising an abnormality detection step of detecting an abnormality of the detection signal by performing a singular value check of a defining factor of the uniform component removal signal generated by the uniform component removal signal generation step Detection method. The multi-sensor signal abnormality detection method according to claim 10,
The pre-processing step for the uniform component removal process includes:
There between the uniform component Hikisa signal generating step and the normalized signal generating step, the value of the detection signal obtained from each of the plurality of sensors when passing through the site has no significant signal generating factor A multi-sensor signal abnormality detection method comprising an offset step of performing an offset process on the detection signals obtained from each of the plurality of sensors after the normalization process so that becomes a constant value. The multi-sensor signal abnormality detection method according to claim 10,
The multi-sensor signal abnormality detection device further includes a drift removal filter for removing drift noise,
The pre-processing step for the uniform component removal process includes:
There between the uniform component Hikisa signal generating step and the normalized signal generating step, from the detection signal obtained by each of the plurality of sensors after the normalizing process, to remove the drift noise A multi-sensor signal abnormality detection method comprising a drift removal step. The multi-sensor signal abnormality detection method according to claim 10,
As the singular value check, a multi-sensor signal abnormality detection method comprising a threshold determination step for checking whether an amplitude after removal of the uniform component of the detection signal obtained by each of the plurality of sensors is equal to or greater than a predetermined threshold . The multi-sensor signal abnormality detection method according to claim 10,
As the singular value check, a noise level calculation step for obtaining the magnitude of the noise level of all sensors of the detection signal by each of the plurality of sensors after the uniform component removal,
And a relative comparison step of relatively comparing the magnitude of the noise level obtained in the noise level calculation step with the magnitude after removal of the uniform component of the detection signal by each of the plurality of sensors. Multi-sensor signal abnormality detection method. In the multi-sensor signal abnormality detection method according to claims 10 to 14,
A multi-stage sensor processing step for performing a uniform component removal signal generation step for each stage of the multi-stage arrangement when the plurality of sensors have a multi-stage arrangement in the traveling direction in which the detection signals are acquired; Multi-sensor signal abnormality detection method. In the multi-sensor signal abnormality detection method according to claims 10 to 15,
A singularity determination step of performing a singular value check of a prescribed characteristic of a signal generated when passing through a standard signal generation factor portion of the detection signal obtained by each of the plurality of sensors before the normalization signal generation step; A multi-sensor signal abnormality detection method provided. To realize the multi-sensor signal abnormality detection method according to claim 10,
Multi-sensor signal abnormality detection program.

Images