Frequency data analysis device, frequency data analysis method, and frequency data analysis program

Abstract

To provide a frequency data analysis device, a frequency data analysis method and a frequency data analysis program capable of extracting a peak value when abnormality occurs.SOLUTION: A processing server 10 sets a frequency range in which a frequency error rate with respect to a true value is within a frequency tolerance error rate α as a thinning section; based on a frequency spectrum containing a plurality of frequency components, detects a maximum value where signal strength is maximum in the thinning section; sets the detected maximum value as a median value of the thinning section; and records the set median value in a storage unit 140.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a frequency data analysis device, a frequency data analysis method, and a frequency data analysis program for analyzing a frequency spectrum including data of different frequencies. [Background technology]

[0002] Systems are being used to monitor machinery and equipment by analyzing the frequency of data acquired by sensors. In such systems, data transmitted from sensors via a network is typically stored in a database, and the target machinery and equipment is monitored based on the analysis results of the stored data. However, there is a problem in that the amount of data is large, which places a heavy load on the network and database.

[0003] In order to deal with the huge amount of data, for example, necessary waveforms are cut out from the original data to reduce the amount of data, but this can prevent proper capture of waveforms during abnormal times such as when peak values ​​suddenly increase. Alternatively, there are cases where the amount of data is reduced by acquiring only a predetermined number of peak values ​​in descending order after frequency analysis, but because the number of peaks that will occur when an abnormality occurs cannot be predicted in advance, acquiring only the predetermined number of peaks can result in the peak values ​​during the abnormality being missed.

[0004] Patent Document 1 discloses an invention of a data compression method in which time series data in a specific frequency range determined by comparing the time series data with a threshold value is analyzed. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 2017-122976 Summary of the Invention [Problem to be solved by the invention]

[0006] However, the invention described in Patent Document 1 does not target areas outside the specific frequency range, and therefore has the problem that peak values ​​occurring outside the range cannot be extracted.

[0007] In consideration of the above, an object of the present invention is to provide a frequency data analysis device, a frequency data analysis method, and a frequency data analysis program that are capable of extracting peak values ​​when an abnormality occurs. [Means for solving the problem]

[0008] To achieve the above objectives According to the first aspect The frequency data analysis device includes a thinning interval setting unit that sets frequency analysis data as a thinning interval for each predetermined frequency range, a maximum value detection unit that detects a maximum value that maximizes signal strength in the thinning interval from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data, and a median value setting unit that sets the maximum value to a median value of the thinning interval. the thinning-out interval setting unit sets a frequency allowable error rate α representing an allowable frequency error rate with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage) for a thinning-out start frequency F' obtained by adding a frequency resolution Δf of a detection unit that detects frequencies to a predetermined minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, a predetermined maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and the number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax): N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F' using the following formula: β = (2αNF' - 2.56R) / NF' to obtain a thinning frequency band F'β, and by setting the thinning interval to have the thinning start frequency F' as the lower limit and the frequency obtained by adding the thinning start frequency F' to the thinning frequency band F'β as the upper limit, a frequency range in which the error rate of the frequency with respect to the true value is within the frequency allowable error rate α is set as the thinning interval.

[0009] According to the first aspect According to the frequency data analyzer, the maximum value detected in a thinning section set for each predetermined frequency range of the frequency analysis data is set as the median value of the thinning section and recorded, thereby making it possible to extract the peak value when an abnormality occurs. In addition, it becomes possible to keep the error within an acceptable range when the maximum value detected in the thinning section is set as the median value of the thinning section, and the thinning section can be set as a frequency range in which the frequency error rate relative to the true value is within the frequency allowable error rate α.

[0014] According to the second aspect Frequency data analyzer In a first aspect, The thinning-out interval setting unit sets the upper limit to a new thinning-out start frequency F' and set the new thinning start frequency F' The highest monitoring frequency Fmax The new decimation start frequency is F' Based on the new decimated frequency band F'β and calculate the new thinning start frequency F'is the lower limit, and the new decimation start frequency F' The new thinned frequency band F'β a new thinning-out interval having an upper limit equal to the frequency obtained by adding the above-mentioned maximum value detection unit and detecting the maximum value in the new thinning-out interval from the frequency spectrum.

[0015] According to the second aspect According to the frequency data analyzing device, by successively updating the thinning interval, it is possible to perform data amount reduction processing for a wide frequency spectrum.

[0016] According to the third aspect Frequency data analyzer In a second aspect, The thinning-out interval setting unit sets the new thinning-out start frequency F' However, the highest monitoring frequency Fmax The new thinning interval is set until the number of thinning intervals exceeds the threshold.

[0017] According to the third aspect According to the frequency data analyzer, it is possible to perform a process of reducing the amount of data of the frequency spectrum from the lowest monitoring frequency to the highest monitoring frequency.

[0018] According to the fourth aspect Frequency data analyzer In any one of the first to third aspects, The thinning-out interval setting unit sets the thinning-out interval in the high frequency region to be wider than the thinning-out interval in the low frequency region so as to maintain the same frequency error rate over the entire frequency band.

[0019] According to the fourth aspect According to the frequency data analyzing device, by setting the thinning interval wider in the high frequency region than in the low frequency region, it is possible to maintain the same frequency error rate over the entire frequency band.

[0020] To achieve the above objectives According to the fifth aspect The frequency data analysis method sets the frequency analysis data as a thinning interval for each predetermined frequency range. No. 1 and detecting a maximum value in which the signal strength is greatest in the thinning section from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data. No. 2 and setting the maximum value to the median value of the thinning interval. Third Process and a frequency tolerance error rate α representing an error rate of a frequency that is allowable with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage) for a thinning-out start frequency F' obtained by adding a frequency resolution Δf of a detection unit that detects the frequency to a predetermined minimum monitoring frequency Fmin that is the lowest frequency to be monitored, a predetermined maximum monitoring frequency Fmax that is the highest frequency to be monitored, and a number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax): N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F' using the following formula: β = (2αNF' - 2.56R) / NF' to obtain a thinning frequency band F'β, and by setting the thinning interval to have the thinning start frequency F' as the lower limit and the frequency obtained by adding the thinning start frequency F' to the thinning frequency band F'β as the upper limit, a frequency range in which the error rate of the frequency with respect to the true value is within the frequency allowable error rate α is set as the thinning interval.

[0021] According to the fifth aspect According to the frequency data analysis method, the maximum value detected in a thinned-out section of frequency analysis data set for each predetermined frequency range is set as the median value of the thinned-out section and recorded, thereby making it possible to extract the peak value when an abnormality occurs. In addition, it becomes possible to keep the error within an acceptable range when the maximum value detected in the thinning section is set as the median value of the thinning section, and the thinning section can be set as a frequency range in which the frequency error rate relative to the true value is within the frequency allowable error rate α.

[0022] In order to achieve the above object, a frequency data analysis program according to a sixth aspect causes a computer to function as a thinning interval setting unit that sets frequency analysis data as a thinning interval for each predetermined frequency range, a maximum value detection unit that detects a maximum value that maximizes signal strength in the thinning interval from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data, and a median value setting unit that sets the maximum value to a median value of the thinning interval. a frequency tolerance error rate α representing an error rate of a frequency that is allowable with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage), for a thinning-out start frequency F' obtained by adding a frequency resolution Δf of a detection unit that detects frequencies to a predetermined minimum monitoring frequency Fmin that is the lowest frequency to be monitored; a predetermined maximum monitoring frequency Fmax that is the highest frequency to be monitored; and a number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax): N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F' using the following formula: β = (2αNF' - 2.56R) / NF' to obtain a thinning frequency band F'β, and by setting the thinning interval to have the thinning start frequency F' as the lower limit and the frequency obtained by adding the thinning start frequency F' to the thinning frequency band F'β as the upper limit, a frequency range in which the error rate of the frequency with respect to the true value is within the frequency allowable error rate α is set as the thinning interval.

[0023] According to the sixth aspect According to the frequency data analysis program, the maximum value detected in a thinned-out section of the frequency analysis data set for each specified frequency range is set as the median value of the thinned-out section and recorded, thereby making it possible to extract the peak value when an abnormality occurs. In addition, it becomes possible to keep the error within an acceptable range when the maximum value detected in the thinning section is set as the median value of the thinning section, and the thinning section can be set as a frequency range in which the frequency error rate relative to the true value is within the frequency allowable error rate α. [Effects of the Invention]

[0024] As described above, the frequency data analyzing device, the frequency data analyzing method, and the frequency data analyzing program according to the present invention can extract peak values ​​when an abnormality occurs. [Brief explanation of the drawings]

[0025] [Figure 1] 1 is a block diagram showing an example of the configuration of a frequency data analyzing device according to a first embodiment. [Figure 2] FIG. 2 is a block diagram showing an example of a specific configuration of a processing server according to the first embodiment. [Figure 3] FIG. 2 is a functional block diagram of a CPU of a processing server. [Figure 4] 4 is a flowchart showing an example of processing in each of a user terminal, a processing server, a measurement control unit, and a storage unit that configure the frequency data analyzing device according to the first embodiment. [Figure 5] 5 is a flowchart showing the measurement parameter calculation process in step S106 of FIG. 4. [Figure 6] 5 is a flowchart showing the data amount reduction process in step S120 of FIG. 4. [Figure 7] FIG. 10 is an explanatory diagram showing an outline of a thinning section. [Figure 8] FIG. 10A is an explanatory diagram illustrating a case where the maximum value is extracted in a thinning section, and FIG. 10B is an explanatory diagram illustrating a case where the extracted maximum value is set as the median value of the thinning section. [Figure 9] 10 is a schematic diagram comparing the number of data before thinning out before the data amount reduction process and the number of data after thinning out after the data amount reduction process. FIG. [Figure 10] 10A is an example of a frequency spectrum measured by a spectrum analyzer of data before the data volume reduction process, and FIG. 10B is an example of a frequency spectrum measured by a spectrum analyzer of data after the data volume reduction process. [Figure 11] FIG. 3 is a sequence diagram showing an example of processing performed by the frequency data analyzing device according to the first embodiment. [Figure 12] FIG. 10 is a block diagram showing an example of the configuration of a frequency data analyzer according to a second embodiment. [Figure 13]10 is a flowchart showing an example of processing performed by a frequency data analyzing device according to the second embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0026] [First embodiment] A frequency data analyzing device 100 according to this embodiment will be described below with reference to Fig. 1. The frequency data analyzing device 100 shown in Fig. 1 includes, on a network 62, a measurement control unit 120 that controls a sensor 122 that detects the state of an inspection object 124, a processing server 10 that analyzes data transmitted from the measurement control unit 120 via the network, a storage unit 140 that accumulates data analyzed by the processing server 10, and a user terminal 130 that can view the data accumulated in the storage unit 140. In this embodiment, the sensor 122 is taken as an example of a vibration sensor that detects vibrations of the inspection object 124, but it may also be a sensor that detects electrical characteristics of the inspection object 124, such as current or voltage.

[0027] The storage unit 140 is a data server equipped with a database, and the processing server 10 is a computer capable of executing advanced arithmetic processing at high speed. Each of the storage unit 140 and the processing server 10 may be a standalone server, or may be a cloud that can distribute the processing load. The storage unit 140 and the processing server 10 may be the same server. The user terminal 130 is not an essential component, and may be omitted if the processing server 10 is equipped with input devices such as a keyboard and a mouse, and an output device such as a display.

[0028] 2 is a block diagram showing an example of a specific configuration of the processing server 10 according to this embodiment. The processing server 10 is configured to include a computer 40. The computer 40 includes a CPU (Central Processing Unit) 42, a ROM (Read Only Memory) 44, a RAM (Random Access Memory) 46, and an input / output port 48. As an example, it is desirable that the computer 40 be a model capable of executing advanced arithmetic processing at high speed.

[0029] In the computer 40, the CPU 42, ROM 44, RAM 46, and input / output port 48 are connected to one another via various buses such as an address bus, a data bus, and a control bus. The input / output port 48 is connected to various input / output devices, such as a display 50, a mouse 52, a keyboard 54, a hard disk (HDD) 56, and a disk drive 60 that reads information from various disks (e.g., CD-ROM, DVD, etc.) 58.

[0030] Furthermore, a network 62 is connected to the input / output port 48, enabling information to be exchanged with various devices connected to the network 62. In this embodiment, the network 62 is connected to a measurement control unit 120, a user terminal 130, and a storage unit 140.

[0031] In this embodiment, the data analyzed by the processing server 10 is described as being stored in the memory unit 140, but the information in the memory unit 140 may also be stored in an external storage device such as the HDD 56 built into the computer 40 or an external hard disk.

[0032] A program for analyzing frequency data and the like is installed in the HDD 56 of the computer 40. In this embodiment, the CPU 42 executes the program to analyze the frequency data detected by the sensor 122 and acquired via the measurement control unit 120.

[0033] There are several methods for installing the program for analyzing frequency data in this embodiment into the computer 40. For example, the program can be stored on a CD-ROM, DVD, or the like together with a setup program, and the disk can be inserted into the disk drive 60, and the setup program can be executed by the CPU 42 to install the program into the HDD 46. Alternatively, the program can be installed into the HDD 46 by communicating with another information processing device connected to the computer 40 via a public telephone line or the network 62.

[0034] 3 is a functional block diagram of the CPU 42 of the processing server 10. Various functions realized by the CPU 42 of the processing server 10 executing a program related to frequency data analysis will be described. The program related to frequency data analysis has a measurement parameter calculation processing function for calculating the data acquisition conditions of the sensor 122, a frequency analysis function for performing frequency analysis of data using fast Fourier transform or the like, a data volume reduction processing function for reducing the amount of data, and an accumulated data analysis function for analyzing accumulated data. By the CPU 42 executing the machine learning program having these functions, the CPU 42 functions as a measurement parameter calculation processing unit 72, a frequency analysis unit 74, a data volume reduction processing unit 76, and an accumulated data analysis unit 78, as shown in FIG. 3.

[0035] 4 is a flowchart showing an example of processing in each of the user terminal 130, processing server 10, measurement control unit 120, and storage unit 140 that constitute the frequency data analysis device 100 according to this embodiment. In step S100, analysis information is input from the user terminal 130. The input analysis information includes a minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, a maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and a frequency allowable error rate α (0<α<1), which is an allowable frequency error rate. The minimum monitoring frequency Fmin, the maximum monitoring frequency Fmax, and the frequency allowable error rate α can each be set arbitrarily, but are set within a range that can be handled by the processing capacity, etc., of the processing server 10.

[0036] In step S102, the input analysis information is transmitted to the processing server 10. In step S104, the processing server 10 receives the analysis information.

[0037] In step S106, the measurement parameter calculation processing unit 72 of the processing server 10 calculates the measurement parameters.

[0038] FIG. 5 is a flowchart showing the measurement parameter calculation process in step S106 of FIG. 4. In step S200, a frequency range R is set. In this embodiment, the frequency range R is the maximum monitoring frequency Fmax input in step S100 of FIG. 4. In the thinning process that reduces the amount of data, the target data must have an accuracy equal to or greater than a predetermined allowable frequency error rate (frequency allowable error rate α). In this embodiment, the maximum frequency Fmax to be confirmed is set as the frequency range R, and the required number of sampling points N is determined using equation (1) described below.

[0039] In step S202, the number of sampling points N is calculated. The number of sampling points N is calculated using the following formula (1). In the following formula (1), 1.28 is a coefficient indicating the range in which fast Fourier transform is possible. The number of sampling points N is the quotient obtained by dividing the frequency range R multiplied by the coefficient by the product of the frequency tolerance rate α and the minimum monitoring frequency Fmin. N=1.28R / αFmin …(1)

[0040] In step S204, the number of sampling points N calculated by the above formula (1) is converted into an integer. Specifically, the calculated N is rounded up to convert it into an integer value.

[0041] In step S206, data acquisition conditions are set and the process returns. The data acquisition conditions set in step S206 are the frequency range R set in step S200 and the number of sampling points N converted to an integer in step S204.

[0042] In step S108 of FIG. 4, the processing server 10 transmits the calculated measurement parameters to the measurement control unit 120.

[0043] In step S110, the measurement control unit 120 receives the measurement parameters. In step S112, the measurement control unit 120 controls the sensor 122 in accordance with the frequency range R and the number of sampling points N included in the measurement parameters, and performs data measurement to acquire data. Then, in step S114, the measurement control unit 120 transmits the measurement data to the processing server 10.

[0044] In step S116, the processing server 10 receives the measurement data. Then, in step S118, the frequency analysis unit 74 of the processing server 10 performs frequency analysis to extract waveforms that constitute the measurement data. As an example, the frequency analysis in step S118 uses a fast Fourier transform or the like. The frequency analysis in step S118 results in a so-called frequency spectrum, in which the signals are classified by frequency.

[0045] In step S120, the data amount reduction processing unit 76 of the processing server 10 performs a data amount reduction process to reduce the amount of measurement data after frequency analysis.

[0046] Fig. 6 is a flowchart showing the data amount reduction process in step S120 in Fig. 4. In step S300, a frequency range R is set. As described above, the frequency range is the maximum frequency Fmax to be confirmed.

[0047] In step S302, the thinning-out start frequency F' is set. The thinning-out start frequency F' is calculated by the following equation (2). As shown in equation (2), the thinning-out start frequency F' is obtained by adding the frequency resolution Δf to the minimum monitoring frequency Fmin. The frequency resolution Δf relates to the range in which fast Fourier transform is possible, and is, for example, 2.56. F'=Fmin + Δf …(2)

[0048] In step S304, a thinning rate β is calculated. The thinning rate β is calculated using the following equation (3). As will be described later, the thinning rate β is determined so that the error rate of the frequency relative to the true value is within the frequency allowable error rate α. β=(2αNF'- 2.56R) / NF' …(3)

[0049] In step S306, the thinning-out interval is calculated. Fig. 7 is an explanatory diagram showing an outline of the thinning-out interval F'β. As shown in Fig. 7, the thinning-out interval F'β is a predetermined frequency range from the thinning-out start frequency F' set in step S302 to F'+F'β.

[0050] In step S308, the maximum value of the signal strength within the thinning-out section F'β is extracted. Fig. 7 shows a case where the maximum value is detected at the thinning-out start frequency F'. Furthermore, it is assumed that the detected maximum value is actually a true value that should be detected at frequency F. In such a case, the frequency at which the maximum value is detected has a measurement and analysis error of |F-F'| with respect to the frequency at which the true value is detected.

[0051] In step S310, the frequency value of the extracted maximum value is set to the median value of the thinning section F'β. FIG. 7 shows a state in which the extraction point, which is the extracted maximum value, is set to the frequency of F' + F'β / 2, which is the median value of the thinning section. Since the maximum value is set as the extraction point to the median value of the thinning section F'β, an error of |F - F'| + F'β / 2 occurs in the frequency of the extraction point with respect to the frequency at which the true value is detected. Therefore, the error rate of the frequency of the extraction point with respect to the frequency at which the true value is detected is as shown in equation (4) below. In this embodiment, the thinning rate β is set so that the error rate shown in equation (4) below is within the frequency allowable error rate α. (|F-F'| + F'β / 2) / F …(4)

[0052] In step S310, the extraction point set at the median value of the thinning section F'β is recorded as thinned data.

[0053] Fig. 8(A) is an explanatory diagram showing a case where the maximum value is extracted in the thinning section, and Fig. 8(B) is an explanatory diagram showing a case where the extracted maximum value is set as the median value of the thinning section. As shown in Fig. 8(B), in this embodiment, the amount of data is reduced by deleting data other than the thinned data set as the median value.

[0054] In this embodiment, the thinning interval is set wider in the high frequency region than in the low frequency region to maintain the same frequency error rate across the entire frequency band. In this embodiment, the entire frequency band is, for example, the frequencies from the minimum monitoring frequency Fmin to the maximum monitoring frequency Fmax. As shown in FIGS. 8(A) and 8(B), the thinning start frequency F2' in the high frequency region is greater than the thinning start frequency F1' in the low frequency region, and therefore the thinning rate β2 in the high frequency region, calculated using the above-described formula (3), is greater than the thinning rate β1 in the low frequency region. As a result, F1'β1≦F2'β2, and therefore the thinning interval can be set wider in the high frequency region than in the low frequency region.

[0055] In step S312, the thinning-out start frequency F' is updated. The new thinning-out start frequency F' is F'+F'β shown in FIG.

[0056] In step S314, it is determined whether the updated thinning-out start frequency F' is equal to or less than the maximum monitoring frequency Fmax. If the updated thinning-out start frequency F' is equal to or less than the maximum monitoring frequency Fmax in step S314, the procedure proceeds to step S304, and the procedure from calculation of the thinning-out rate β to the thinning-out start frequency F' is repeated. If the updated thinning-out start frequency F' is not equal to or less than the maximum monitoring frequency Fmax in step S314, the process returns.

[0057] Fig. 9 is a schematic diagram comparing the number of data before decimation before the data volume reduction process and the number of data after decimation after the data volume reduction process. In Fig. 9, as an example, when the frequency tolerance rate α is set in the range of 1 to 5%, the data volume reduction process can reduce the data volume by nearly 99%.

[0058] Figure 10(A) is an example of the frequency spectrum of data before data reduction processing, measured using a spectrum analyzer, and Figure 10(B) is an example of the frequency spectrum of data after data reduction processing, measured using a spectrum analyzer. In Figure 10(B), the data reduction processing, which has reduced the data volume by nearly 99%, allows for prominent peak values ​​to be observed, reducing the computational load in the subsequent analysis of stored data.

[0059] 4, the data after the data volume reduction process described above is analyzed by the stored data analysis unit 78 of the processing server 10. As an example, if the sensor 122 is a vibration sensor, the analysis of the stored data in step S124 is an analysis of the vibration frequency. In this embodiment, the data volume reduction process in step S120 reduces the data volume by extracting the maximum value in the thinning section as the median value of the thinning section, thereby reducing the calculation load in analyzing the stored data.

[0060] In step S126, the data is stored in storage unit 140. After the data is stored in storage unit 140 in step S126, the processing shown in FIG.

[0061] 11 is a sequence diagram showing an example of processing of the frequency data analysis device 100 according to this embodiment. In step S1, analysis information is input from the user terminal 130, and in step S2, the input analysis information is transmitted to the processing server 10.

[0062] In step S3, the processing server 10 calculates measurement parameters based on the analysis information transmitted from the user terminal 130. Then, in step S4, the calculated measurement parameters are transmitted to the measurement control unit 120.

[0063] In step S5, the measurement control unit 120 outputs a measurement control signal according to the received measurement parameters to the sensor 122. Then, in step S6, the sensor 122 performs data measurement according to the received measurement control signal, and in step S7, the sensor 122 transmits data detected by the measurement to the measurement control unit 120.

[0064] In step S8, the measurement control unit 120 receives data from the sensor 122, and in step S9, the measurement control unit 120 transmits the received data to the processing server 10.

[0065] In step S10, the processing server 10 receives the data. In step S11, the processing server 10 performs frequency analysis using a fast Fourier transform or the like.

[0066] In step S12, the processing server 10 performs a data amount reduction process using thinning sections. In step S13, the processing server 10 analyzes the stored data. Then, in step S14, the processing server 10 transmits the analyzed data to the storage unit 140.

[0067] In step S15, the storage unit 140 stores the data received from the processing server 10.

[0068] In step S16, the user terminal 130 sends a data viewing request to the storage unit 140.

[0069] In step S17, the storage unit 140 transmits data in response to the data viewing request from the user terminal 130, and in step S18, the user terminal 130 receives the data transmitted from the storage unit 140.

[0070] As described above, in this embodiment, the thinning-out interval is set so that the error rate of the frequency of the signal detected in the frequency spectrum between the minimum monitoring frequency Fmin and the maximum monitoring frequency Fmax relative to the true value is within the frequency allowable error rate α, which is an allowable frequency error rate. Then, the maximum value at which the signal strength is greatest in the thinning-out interval is detected, and the detected maximum value is set as the median value of the thinning-out interval, thereby making it possible to extract the peak value when an abnormality occurs and reducing the amount of data to be used for data analysis.

[0071] By using the maximum value of the thinning section as the representative value, the remaining data can be reduced, making it possible to reduce the amount of data by nearly 99% from the original data. As a result, as shown in Figure 10(B), peak values ​​can be clearly observed, reducing the calculation load in the subsequent data analysis.

[0072] The setting of the thinning interval is repeated until the thinning start frequency F' at which the thinning interval begins exceeds the maximum monitoring frequency Fmax, so that data analysis of the frequency spectrum between the minimum monitoring frequency Fmin and the maximum monitoring frequency Fmax can be performed while suppressing the computational load.

[0073] In this embodiment, the target data is, for example, the frequency spectrum of vibrations detected by a vibration sensor, but is not limited to this. If the original data is data that exhibits a frequency spectrum, it is possible to extract peak values ​​when an abnormality occurs, even for data analysis of electromagnetic waves, currents, voltages, sound waves, etc., and it is also possible to reduce the amount of data used for data analysis.

[0074] [Second embodiment] Next, a second embodiment of the present invention will be described. As shown in Fig. 12, a frequency data analyzer 200 according to this embodiment differs from the first embodiment in that a single control and arithmetic device 220 has the functions of the processing server 10, measurement control unit 120, user terminal 130, and storage unit 140 in the first embodiment. However, since the sensor 122 and the inspection object 124 are the same as those in the first embodiment, they are denoted by the same reference numerals as in the first embodiment and detailed description thereof will be omitted.

[0075] 13 is a flowchart showing an example of processing by the frequency data analyzer 200 according to this embodiment. In step S400, analysis information is input. As in the first embodiment, the input analysis information includes a minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, a maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and a frequency allowable error rate α (0<α<1), which is an allowable frequency error rate. The minimum monitoring frequency Fmin, the maximum monitoring frequency Fmax, and the frequency allowable error rate α can each be set arbitrarily, but are set within a range that can be handled by the processing capacity, etc., of the control and arithmetic device 220.

[0076] In step S402, measurement parameters are calculated in the same manner as in step S106 in Fig. 4. The measurement parameters to be calculated are the frequency range R and the number of sampling points N, as in the first embodiment.

[0077] In step S404, the sensor 122 is controlled in accordance with the frequency range R and the number of sampling points N included in the measurement parameters, and data measurement is performed to acquire data.

[0078] In step S406, a frequency analysis is performed to extract waveforms that constitute the measurement data. The frequency analysis in step S406 uses fast Fourier transform or the like, as in the first embodiment.

[0079] In step S408, similarly to step S120 in FIG. 4, a data amount reduction process is performed to reduce the amount of measurement data after frequency analysis.

[0080] In step S410, the accumulated data is analyzed. For example, if the sensor 122 is a vibration sensor, the analysis of the accumulated data in step S410 is an analysis of the vibration frequency. In this embodiment, the data amount reduction process in step S408 reduces the amount of data by extracting the maximum value in the thinning interval as the median value of the thinning interval, thereby reducing the calculation load in analyzing the accumulated data.

[0081] In step S412, the data is stored in the storage device of the control and arithmetic device 220, and the process ends.

[0082] As described above, in this embodiment, the control and arithmetic device 220, which has the functions of the processing server 10, measurement control unit 120, user terminal 130, and memory unit 140 in the first embodiment, can set measurement parameters, control the sensor 122, reduce data volume, analyze accumulated data, and accumulate data.

[0083] In this embodiment, too, by using the maximum value of the thinning section as the representative value, it is possible to reduce the amount of data other than the maximum value, thereby reducing the amount of data by nearly 99% from the original data. As a result, as shown in Figure 10(B), peak values ​​can be clearly observed, and the calculation load in the subsequent data analysis can be reduced.

[0084] In addition, in this embodiment, it is possible to extract peak values ​​when an abnormality occurs not only for data analysis of vibrations but also for data analysis of electromagnetic waves, current, voltage, sound waves, etc., and it is also possible to reduce the amount of data used for data analysis.

[0085] The "thinning section setting unit," "maximum value detection unit," and "median value setting unit" described in the claims correspond to the "data amount reduction processing unit 76" described in the detailed description of the invention in the specification. Also, the "recording unit" described in the claims corresponds to the "storage unit 140" described in the detailed description of the invention in the specification.

[0086] In the above embodiments, the processing performed by the CPU after reading the software (program) may be performed by various processors other than the CPU. Examples of such processors include programmable logic devices (PLDs) whose circuit configuration can be changed after fabrication, such as field-programmable gate arrays (FPGAs), and dedicated electrical circuits, such as application-specific integrated circuits (ASICs), which are processors with circuit configurations specifically designed to perform specific processing. The processing may be performed by one of these processors, or by a combination of two or more processors of the same or different types (e.g., multiple FPGAs, or a combination of a CPU and an FPGA). The hardware structure of these processors is, more specifically, an electrical circuit that combines circuit elements such as semiconductor devices.

[0087] In addition, in each of the above embodiments, the program is described as being stored (installed) in advance in the disk drive 60 or the like, but this is not limiting. The program may be provided in a form stored in a non-transitory storage medium such as a CD-ROM (Compact Disk Read Only Memory), a DVD-ROM (Digital Versatile Disk Read Only Memory), or a USB (Universal Serial Bus) memory. The program may also be downloaded from an external device via a network.

[0088] (Additional note 1) Memory and at least one processor coupled to said memory; Including, The processor: a frequency range in which the error rate of the frequency relative to the true value is within a frequency tolerance rate is set as a thinning interval, a maximum value in which the signal strength is greatest in the thinning interval is detected from a frequency spectrum including a plurality of frequency components, the maximum value is set as a median value of the thinning interval, and the median value is recorded; An information processing device configured as follows. [Explanation of symbols]

[0089] 10 Processing Server 40 Computer 42 CPU 44 ROM 46 RAM 48 input / output ports 50 displays 52 Mouse 54 keyboard 60 disk drives 62 Network 72 Measurement parameter calculation processing unit 74 Frequency analysis section 76 Data volume reduction processing unit 78 Accumulated Data Analysis Department 100 Frequency data analyzer 120 Measurement control section 122 sensors 124 Inspection object 130 User terminals 140 Storage section 200 Frequency Data Analysis Device 220 Control and arithmetic unit

Claims

1. a thinning interval setting unit that sets the frequency analysis data as a thinning interval for each predetermined frequency range; a maximum value detection unit that detects a maximum value that is the maximum signal strength in the thinning section from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data; a median setting unit that sets the maximum value to a median value of the thinning interval; Including, The thinning-out interval setting unit sets a thinning-out start frequency F' obtained by adding a frequency resolution Δf of a detection unit that detects frequencies to a predetermined minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, and calculates a frequency allowable error rate α that represents an allowable frequency error rate with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage), a predetermined maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and the number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax), N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F′ using the following formula: β=(2αNF'- 2.56R) / NF' to obtain a thinning frequency band F'β, and by setting the thinning interval to have the thinning start frequency F' as a lower limit and a frequency obtained by adding the thinning start frequency F' to the thinning frequency band F'β as an upper limit, the frequency range in which the error rate of the frequency from the true value is within the frequency allowable error rate α is set as the thinning interval.

2. The thinning-out interval setting unit sets the upper limit to a new thinning-out start frequency F', and when the new thinning-out start frequency F' is equal to or less than the maximum monitored frequency Fmax, calculates a new thinning-out frequency band F'β based on the new thinning-out start frequency F', and sets a new thinning-out interval whose lower limit is the new thinning-out start frequency F' and whose upper limit is the frequency obtained by adding the new thinning-out frequency band F'β to the new thinning-out start frequency F'; 2. The frequency data analyzing device according to claim 1, wherein the maximum value detecting section detects a maximum value in the new thinning interval from the frequency spectrum.

3. The frequency data analysis device according to claim 2, wherein the thinning-out interval setting unit sets the new thinning-out interval until the new thinning-out start frequency F' exceeds the monitored maximum frequency Fmax.

4. A frequency data analysis device as described in any one of claims 1 to 3, wherein the thinning interval setting unit sets the thinning interval in the high frequency region to be wider than the thinning interval in the low frequency region so as to maintain the same frequency error rate across the entire frequency band.

5. A first step of setting frequency analysis data as thinning intervals for each predetermined frequency range; a second step of detecting a maximum value that is the maximum signal strength in the thinning section from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data; a third step of setting the maximum value to a median value of the thinning section; A frequency data analysis method in which a computer executes a process including: The first step involves calculating a frequency allowable error rate α representing an allowable frequency error rate with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage) for a thinning-out start frequency F′ obtained by adding a frequency resolution Δf of a detection unit that detects the frequency to a predetermined minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, a predetermined maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and the number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax): N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F′ using the following formula: β=(2αNF'- 2.56R) / NF' to obtain a thinning frequency band F'β, and to set the thinning interval to a frequency range in which an error rate of the frequency from the true value is within the frequency allowable error rate α by multiplying the frequency band F'β by the thinning start frequency F' and the frequency band F'β is set as an upper limit.

6. A computer, a maximum value detection unit that detects a maximum value that is the maximum signal strength in the thinning interval from a frequency spectrum including a plurality of frequency components obtained from the frequency analysis data; and a median value setting unit that sets the maximum value to a median value of the thinning interval, The thinning-out interval setting unit sets a thinning-out start frequency F' obtained by adding a frequency resolution Δf of a detection unit that detects frequencies to a predetermined minimum monitoring frequency Fmin, which is the lowest frequency to be monitored, and calculates a frequency allowable error rate α that represents an allowable frequency error rate with respect to a true value (a value calculated by multiplying a value obtained by dividing the error by the true value by a percentage), a predetermined maximum monitoring frequency Fmax, which is the highest frequency to be monitored, and the number of sampling points N calculated by the following formula (where R is a frequency range, and R=Fmax), N=1.28R / αFmin and a thinning rate β calculated based on the thinning start frequency F′ using the following formula: β=(2αNF'- 2.56R) / NF' to obtain a thinning frequency band F'β, and to set the thinning interval to a frequency range in which an error rate of the frequency from a true value is within the frequency allowable error rate α by multiplying the frequency band F'β by the thinning start frequency F' and the frequency band F'β is set as an upper limit.

Images