Signal analysis device, signal analysis method, and program

Abstract

To provide a signal analyzer capable of estimating peak frequencies by generalization harmonic analysis.SOLUTION: The signal analyzer includes a control unit configured to perform the steps of: calculating a first peak frequency by executing generalization harmonic analysis while increasing the analysis frequency from the lower limit to the upper limit of a predetermined frequency range on waveform data to be evaluated; calculating a second peak frequency by executing generalization harmonic analysis while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range; and estimating the peak frequency of the waveform data by calculating the average of the first peak frequency and the second peak frequency.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a signal analysis device, a signal analysis method, and a program. [Background technology]

[0002] A technique has been proposed in which the vibrations of a device or the like to be evaluated are analyzed using a fast Fourier transform (FFT) and the state of the device or the like is evaluated from the peak frequency (for example, Patent Document 1). Patent Document 2 discloses an abnormality diagnosis device that performs an FFT analysis of the current of an electric motor and determines whether the motor is abnormal based on the spectral peak of the sideband wave component. Peak frequency analysis is also widely used in fields other than equipment monitoring and state evaluation.

[0003] Another known technique for frequency analysis is generalized harmonic analysis (GHA) (see, for example, Patent Document 3, Patent Document 4, and Non-Patent Document 1). To achieve high resolution in frequency analysis, zooming (Zoom FFT) is used in FFT, but zooming requires an extension of the measurement time for waveform data. In contrast, generalized harmonic analysis makes it possible to achieve high resolution in frequency analysis without extending the measurement time. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-45472 [Patent Document 2] Patent No. 6824493 [Patent Document 3] Japanese Patent Application Laid-Open No. 2002-328697 [Patent Document 4] Patent No. 5590547 [Non-patent literature]

[0005] [Non-Patent Document 1] Mikio Higashiyama, Tsunehiko Koike, "Signal analysis method with high frequency analysis accuracy", Transactions of the Acoustical Society of Japan, Vol. 54, No. 8, 1988, pp. 568-574 Summary of the Invention [Problem to be solved by the invention]

[0006] There is a need for a technique to estimate the peak frequency using generalized harmonic analysis.

[0007] The present disclosure provides a signal analysis device, a signal analysis method, and a program that can solve the above problems. [Means for solving the problem]

[0008] The signal analysis device of the present disclosure includes a control unit that performs generalized harmonic analysis on waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range to calculate a first peak frequency, performs generalized harmonic analysis on waveform data to be evaluated while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency, and calculates the average of the first peak frequency and the second peak frequency to estimate the peak frequency of the waveform data.

[0009] The signal analysis method of the present disclosure includes: 1. A computer-implemented signal analysis method, comprising: a step of calculating a first peak frequency by performing generalized harmonic analysis on the waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range; The computer performing generalized harmonic analysis while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency; The computer and estimating a peak frequency of the waveform data by calculating an average of the first peak frequency and the second peak frequency.

[0010] The program disclosed herein causes a computer to perform the following steps on waveform data to be evaluated: calculating a first peak frequency by performing generalized harmonic analysis while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range; calculating a second peak frequency by performing generalized harmonic analysis while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range; and estimating the peak frequency of the waveform data by calculating the average of the first peak frequency and the second peak frequency. [Effects of the Invention]

[0011] According to the above-described signal analysis device, signal analysis method, and program, the peak frequency can be estimated by generalized harmonic analysis. [Brief explanation of the drawings]

[0012] [Figure 1] FIG. 1 illustrates an example of a signal analyzing device according to each embodiment. [Figure 2A] FIG. 10 is a diagram showing peak frequencies when frequency analysis is started from low frequencies. [Figure 2B] FIG. 10 is a diagram showing peak frequencies when frequency analysis starts from high frequencies. [Figure 3] FIG. 3 is a diagram illustrating a method for estimating a peak frequency according to the first embodiment. [Figure 4A] FIG. 3 is a first diagram illustrating how to improve the frequency resolution in the peak frequency estimation method according to the first embodiment. [Figure 4B] FIG. 2 is a second diagram illustrating how to improve the frequency resolution in the peak frequency estimation method according to the first embodiment. [Figure 4C] FIG. 10 is a third diagram illustrating how to improve the frequency resolution in the peak frequency estimation method according to the first embodiment. [Figure 4D] FIG. 4 is a fourth diagram illustrating how to improve the frequency resolution in the peak frequency estimation method according to the first embodiment. [Figure 5]5 is a flowchart illustrating an example of a peak frequency estimation process according to the first embodiment. [Figure 6] FIG. 10 is a diagram illustrating pre-processing of the peak frequency estimation process according to the second embodiment. [Figure 7] 10 is a flowchart illustrating an example of a peak frequency estimation process according to the second embodiment. [Figure 8] FIG. 2 is a diagram illustrating an example of a hardware configuration of a signal analyzing device according to each embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] The signal analysis device of the present disclosure will be described below with reference to Figures 1 to 8. In the following description, components having the same or similar functions will be assigned the same reference numerals, and redundant descriptions of those components may be omitted.

[0014] First Embodiment (composition) 1 shows an example of a signal analysis device according to each embodiment. The signal analysis device 10 estimates the peak frequency of waveform data to be evaluated using generalized harmonic analysis (GHA). The signal analysis device 10 includes a waveform data acquisition unit 11, a setting reception unit 12, a control unit 13, an output unit 14, and a storage unit 15. The waveform data acquisition unit 11 acquires time-series waveform data and records the acquired waveform data in the storage unit 15. The setting receiving unit 12 receives settings (target frequency range, resolution, etc.) necessary for analyzing peak frequencies, and records the settings in the storage unit 15. The control unit 13 controls the process of estimating the peak frequency of the waveform data to be evaluated. The control unit 13 has an analysis unit 131. Based on an instruction from the control unit 13, the analysis unit 131 performs frequency analysis on the time-series waveform data acquired by the waveform data acquisition unit 11 using GHA, and estimates the peak frequency based on the analysis result. The output unit 14 outputs the peak frequency estimated by the analysis unit 131 to a display device, an electronic file, or the like. The storage unit 15 stores waveform data, setting information required for analysis, data currently being analyzed by the analysis unit 131, and the like.

[0015] As disclosed in Non-Patent Document 1 and elsewhere, GHA sets an initial value ω1 of the analysis frequency and a target frequency range for waveform data x1 to be evaluated, and calculates waveform data approximating x1 using (A1·sin(ω1·t+φ1)) (t is time). Then, it calculates the amplitude A1 and phase φ1 that minimize the sum of squares of the residual waveform (differential waveform data) x2, expressed as x2 = {x1-A1·sin(ω1·t+φ1)}. Next, it calculates an analysis frequency ω2 that is shifted from the initial value ω1 by the resolution Δω. i+1 and the updated analysis frequency ω i+1 Using the same calculation for the residual waveform x2, we obtain the residual waveform x2 and A2·sin(ω i+1 Calculate the amplitude A2 and phase φ2 that minimize the sum of squares of the residual waveform x3, which is the difference between the analysis frequency ω i The same calculation is repeated while changing the resolution Δω over the target frequency range. In this way, when GHA is performed, the original waveform data x1 is converted into waveform data of different frequencies, A1·sin(ω1·t+φ1), A2·sin(ω i+1 Generally, when performing GHA, the lower limit of the target frequency range is set for the initial analysis frequency ω1, and the above process is performed while increasing the analysis frequency from low to high frequencies in increments of Δω.

[0016] In FIG. 2A, the waveform data to be evaluated is analyzed over the target frequency range from low to high frequencies with a resolution of r2. i2A shows the results of frequency analysis when GHA is performed by updating the analysis frequency from low frequencies. r2 is the resolution of the GHA (Δω as explained above). The vertical axis of Figure 2A shows the amplitude (dB), and the horizontal axis shows the frequency (Hz). The waveform data to be evaluated is time-series waveform data created by combining multiple known waveforms of various frequencies and amplitudes. In other words, the peak frequency of this waveform data is known, and the true value of the peak frequency is Pt as shown. Graph f2l in Figure 2A shows the results of frequency analysis by GHA when updating the analysis frequency from low frequencies, and graph f1 shows the results of frequency analysis by FFT. r1 represents the resolution of the FFT, and it can be seen that GHA is capable of frequency analysis with higher resolution. From low frequencies to the analysis frequency ω i When GHA is performed after updating, the estimated value of the peak frequency is Pl. When compared with the true value of the peak frequency Pt, the low frequency to the analysis frequency ω i It can be seen that when GHA is performed after updating the peak frequency Pl, the estimated peak frequency Pl shifts to the lower frequency side.

[0017] Figure 2B shows the same waveform data as in Figure 2A, with the analysis frequency ω i The graph f2h in Figure 2B shows the results of frequency analysis when GHA is performed after updating the frequency. The vertical axis of Figure 2B shows the amplitude (dB) and the horizontal axis shows the frequency (Hz). The graph f2h in Figure 2B shows the frequency distribution from the high frequency to the analysis frequency ω i The graph f1 shows the results of frequency analysis by GHA when updating is performed, and the graph f2 shows the results of frequency analysis by FFT. i The estimated value of the peak frequency when the update is performed is Ph. When compared with the true value of the peak frequency Pt, the high frequency to the analysis frequency ω i It can be seen that when GHA is performed after updating the peak frequency Ph shifts to the higher frequency side.

[0018] As explained above, the analysis frequency ω i When the update is performed from the low frequency, the estimated peak frequency Pl shifts to the low frequency side, and the analysis frequency ω iThe applicant has confirmed that when the update of the peak frequency is performed, the estimated value Ph of the peak frequency is shifted to the high frequency side. In order to correct this shift and estimate the peak frequency with high accuracy, in this embodiment, the analysis frequency ω i The update is performed twice, once from the low frequency and once from the high frequency, and the average value of the resulting peak frequency estimates Pl and Ph is taken as the peak frequency estimate Ps. i The peak frequency Pl estimated by updating and the high frequency to analysis frequency ω i The average value of the peak frequency Ph estimated by updating is used as the final peak frequency estimate Ps. i The estimation error on the low frequency side when updating and the error from the high frequency to the analysis frequency ω i When updating, the estimation error on the high frequency side is cancelled out, and the estimation error can be significantly reduced. This is shown in Figure 3.

[0019] 3 is a diagram illustrating a peak frequency estimation method according to the first embodiment. FIG. 3 shows an estimated value Ps obtained by performing the peak frequency estimation process of this embodiment on the waveform data to be evaluated, which is the same as that shown in FIG. 2A. As described above, the peak frequency Ps is calculated from the low frequency to the analysis frequency ω i The estimated peak frequency when GHA is performed with the updated value is Pl, and the peak frequency is the analysis frequency ω i When GHA is performed after updating Pl, the estimated value of the peak frequency is Ph. In this embodiment, the average of the estimated values ​​Pl and Ph is taken as the estimated value Ps. As shown in the figure, it is possible to estimate the estimated value Ps, which is closer to the true value Pt of the peak frequency than Pl and Ph.

[0020] Next, the relationship between the peak frequency estimation process and the resolution according to this embodiment will be described with reference to Figs. 4A to 4D. Figs. 4A to 4D show the estimation results of the estimated value Ps when the resolution is set to 0.025 Hz (5 times FFT), 0.01 Hz (12.5 times FFT), 0.005 Hz (25 times FFT), and 0.0025 Hz (50 times FFT), respectively. Referring to Figs. 4A to 4D, a typical GHA (from low frequencies to the analysis frequency ω iIt can be seen that the peak frequency Pl estimated by the method (method of updating Pt) does not reduce the estimation error (the difference between Pl and Pt) even if the resolution is made finer, but the estimated value Ps according to this embodiment can estimate the peak frequency with high accuracy regardless of the resolution. In other words, according to the method of this embodiment, if the peak frequency Ps is estimated with a resolution equal to or greater than a predetermined value, the peak frequency can be estimated with high accuracy.

[0021] (operation) Next, a flow of peak frequency estimation processing according to the first embodiment will be described with reference to Fig. 5. Fig. 5 is a flowchart showing an example of peak frequency estimation processing according to the first embodiment. First, the waveform data acquisition unit 11 acquires time-series waveform data to be evaluated (step S1). This waveform data is set as the first waveform data x1 to be evaluated. Next, the operator sets the target frequency range and resolution in the signal analysis device 10 (step S2). The setting reception unit 12 acquires the set information and records it in the storage unit 15. The target frequency range is the analysis frequency ω i The update range and resolution of the analysis frequency ω i Next, the control unit 13 updates the analysis frequency ω i The control unit 13 sets the lower limit of the target frequency range for the analysis frequency ω1, sets the value of the analysis frequency ω1 plus the resolution Δω for the analysis frequency ω2, sets the value of the analysis frequency ω2 plus the resolution Δω for the analysis frequency ω3, and so on. i until the upper limit of the frequency range of interest, the analysis frequency ω i (i=1 to end) in order from the low frequency side (step S3). i Then, the analysis unit 131 executes frequency analysis by GHA. First, the analysis unit 131 sets the variable i to an initial value of 1 (step S4). Next, the analysis unit 131 updates the waveform data x to be evaluated, which is calculated by the following equation (1): i and A i sin(ω i t+φ i) the amplitude A at which the sum of squares of the residual waveform, which is the difference between i , phase φ i is calculated (step S5). Residual waveform=x i -A i sin(ω i t+φ i ) ···(1) The analysis unit 131 calculates the amplitude A i and phase φ i Analyze the frequency ω i and record it in the storage unit 15 in association with the above.

[0022] Next, the analysis unit 131 calculates the residual waveform calculated above as the waveform data x i+1 (step S6). x i+1 =x i -A i sin(ω i t+φ i ) ···(2) Next, the analysis unit 131 adds 1 to i (step S7). Next, the analysis unit 131 determines whether i exceeds an upper limit value (end) (step S8). If i is equal to or less than the upper limit value (step S8; No), the analysis unit 131 repeats the processing from step S5. If i exceeds the upper limit value (step S8; Yes), the analysis unit 131 estimates the peak frequency Pl (step S9). The analysis unit 131 selects the amplitude A i Select the largest A i The corresponding analysis frequency ω i The specified analysis frequency ω i The analysis unit 131 stores the estimated value Pl of the peak frequency in the storage unit 15.

[0023] Steps S3 to S9 are performed by adjusting the analysis frequency ω from low to high frequencies. i After this process is completed, the analysis frequency ω is updated from high to low frequencies. iand executes GHA (steps S10 to S16 described below). First, the control unit 13 updates the analysis frequency ω i (i=1 to end), values ​​are set in order from high frequency (step S10). Specifically, the control unit 13 sets the upper limit value of the target frequency range for the analysis frequency ω1, sets the value obtained by subtracting the resolution Δω from the analysis frequency ω1 for the analysis frequency ω2, sets the value obtained by subtracting the resolution Δω from the analysis frequency ω2 for the analysis frequency ω3, and so on. i until the lower limit of the frequency range of interest, i For (i=1 to end), values ​​are set in order from the high frequency side.

[0024] Next, the control unit 13 instructs the analysis unit 131 to execute GHA. Then, the analysis unit 131 executes frequency analysis by GHA. First, the analysis unit 131 sets the variable i to an initial value of 1 (step S11). Next, the analysis unit 131 calculates the waveform data x of the evaluation target calculated by the above formula (1). i The amplitude A at which the sum of squares of the residual waveform from i , phase φ i (Step S12). The analysis unit 131 calculates the calculated amplitude A i and phase φ i Analyze the frequency ω i Then, the analysis unit 131 stores the residual waveform in the storage unit 15 in association with the waveform data x i+1 (Step S13). Next, the analysis unit 131 adds 1 to i (Step S14). Next, the analysis unit 131 determines whether i exceeds an upper limit value (end) (Step S15). If i is equal to or less than the upper limit value (Step S15; No), the analysis unit 131 repeats the processing from Step S12. If i exceeds the upper limit value (Step S15; Yes), the analysis unit 131 estimates the peak frequency Ph (Step S16). The analysis unit 131 selects the amplitude A from the frequency analysis results recorded in the storage unit 15. i Select the largest A i The corresponding analysis frequency ω i The specified analysis frequency ωi is the estimated value Ph of the peak frequency. The analysis unit 131 records the estimated value Ph of the peak frequency in the storage unit 15.

[0025] Next, the analysis unit 131 reads out the peak frequency estimates Pl and Ph from the storage unit 15 and calculates their average (step S17). The analysis unit 131 records the calculated average in the storage unit 15 as the final peak frequency estimate Ps. The output unit 14 may also read out the peak frequency estimate Ps from the storage unit 15 and output it to a display device or the like. In step S17, the average of the estimates Pl and Ph is calculated, but the embodiment is not limited to this. For example, the analysis unit 131 may calculate a weighted average of the estimates Pl and Ph to calculate the final peak frequency estimate Ps. In addition, in the flowchart of FIG. 5, the analysis frequency ω is first calculated from low frequencies to high frequencies. i After that, the analysis frequency ω i We have explained the process of performing GHA while updating the analysis frequency ω i After that, the analysis frequency ω is updated from low to high frequencies. i The processing order may be such that GHA is performed while updating the above.

[0026] As described above, according to this embodiment, by using the GHA, it is possible to estimate the peak frequency without extending the measurement time of the waveform data. i The peak frequency can be estimated with high accuracy by performing two GHAs, one that updates the frequency from low frequencies and one that updates the frequency from high frequencies, and then calculating the average (including weighted average) of the resulting peak frequency estimates Pl and Ph. The peak frequency estimation method of this embodiment can be implemented simply by performing GHA twice using general GHA logic and modifying the software logic so that the result is used to estimate the final peak frequency.

[0027] Second Embodiment A signal analysis device according to a second embodiment of the present invention will be described below with reference to FIGS. 6 and 7. FIG. 6 is a diagram illustrating preprocessing for peak frequency estimation according to the second embodiment. In the first embodiment, a method for accurately estimating the peak frequency of waveform data was described. In the second embodiment, a method for accurately estimating the frequency of sideband waves in waveform data will be described. A sideband wave is a wave with large amplitude observed near the frequency of a fundamental wave. For example, when observing the current of a motor-driven device such as a fan or pump, waveforms with large amplitude are observed on both sides of the power supply frequency (50 kHz or 60 kHz). In FIG. 6, X1 is the fundamental wave, and X2 and X3 are sideband waves. Anomalies in fans and pumps may appear in the sideband waves or harmonic components of the sideband waves in the current waveform data. Therefore, when monitoring these devices for abnormalities based on the sideband waves, it is necessary to accurately estimate the frequency of the sideband waves. When estimating the frequency of the sideband waves, the frequency analysis using the GHA described in the first embodiment can be used, but two problems exist. That is, (1) because the amplitude of the fundamental wave is overwhelmingly larger than that of the sideband waves, the fundamental wave affects the estimation of the sideband waves, resulting in reduced estimation accuracy. (2) When the target frequency range is wide, the amount of calculation required for GHA becomes enormous, resulting in increased calculation load and calculation time. Therefore, in the second embodiment, to address issue (1), (preprocessing 1) removes the fundamental wave component from the time-series waveform data before executing GHA. In the example of FIG. 6, the fundamental wave X1 is removed from the waveform data to be evaluated. The GHA logic can be used to remove the fundamental wave X1. Furthermore, to address issue (2), (preprocessing 2) limits the target frequency range to the range in which sideband waves appear and executes GHA. These steps enable accurate and efficient estimation of the frequency of the sideband waves. For example, if the target is a fan or a pump, the frequency of the sideband waves can be estimated by engineering calculations based on its shape, size, structure, etc.

[0028] (operation) Next, a description will be given of the flow of a method for estimating the frequency of a sideband wave according to the second embodiment. Fig. 7 is a flowchart showing an example of a peak frequency estimation process according to the second embodiment. First, the waveform data acquisition unit 11 acquires time-series waveform data to be evaluated (step S21). Next, an operator sets the frequency of the fundamental wave, a limited frequency range, and a resolution in the signal analysis device 10 (step S22). The setting reception unit 12 acquires the various set information and records it in the storage unit 15. For example, the limited frequency range is the range in which the sideband wave X3 appears, and the resolution is the range in which the analysis frequency ω i The frequency of the fundamental wave X1 is, for example, the power supply frequency. The frequency of the fundamental wave X1 may be determined by separately performing an FFT on the waveform data to be evaluated, and the operator may select a value closest to the power supply frequency from the frequencies analyzed by the FFT and set the value in the signal analyzing device 10. Next, the control unit 13 sets the analysis frequency ω0 for preprocessing 1 to the frequency of the fundamental wave X1 set in step S22 (step S23). Next, the control unit 13 instructs the analysis unit 131 to perform GHA on the waveform data to be evaluated acquired in step S21 only once, with ω0 as the analysis frequency. Then, the analysis unit 131 calculates the amplitude A and phase φ that minimize the sum of squares of the residual waveform, which is the difference between the waveform data x to be evaluated and A·sin(ω·t+φ) calculated using the following equation (3) (step S24). Residual waveform=xA sin(ωt+φ) (3) The analysis unit 131 records the calculated amplitude A and phase φ in the storage unit 15 in association with the analysis frequency ω0.

[0029] Next, the analysis unit 131 sets the residual waveform as the first waveform data x1 to be evaluated using the following equation (4) (step S25). x1=xA·sin(ω0·t+φ) ···(4) The analysis unit 131 records the waveform data x1 to be first evaluated in the storage unit 15. By steps S13 to S15, the fundamental wave X1 can be removed from the waveform data to be evaluated (pre-processing 1).

[0030] Next, the control unit 13 sets the limited frequency range set in step S12 as the analysis frequency ω iThe frequency analysis range is then set as the update range (i.e., target frequency range) (step S26) (preprocessing 2). This limits the frequency analysis range to the range in which sideband waves appear, reducing the amount of calculation. It also prevents the detection of frequencies of waves other than the target sideband wave X3 (for example, sideband wave X2).

[0031] Next, the control unit 13 executes steps S3 to S17 of the flowchart in FIG. 5 using the residual waveform x1 obtained after removing the fundamental wave X1 as the waveform data to be evaluated and the limited frequency range as the target frequency range (step S27). The fundamental wave X1 is removed, the target frequency range is limited to the frequency band in which sidebands appear, and the peak frequency is estimated by the method of the first embodiment. That is, the analysis frequency ω i The GHA is performed twice, once starting from the low frequency and once starting from the high frequency, and the peak frequency estimates Pl and Ph obtained by each GHA are averaged to calculate the frequency of the desired sideband. This makes it possible to estimate the frequency of the desired sideband (e.g., sideband X3).

[0032] As described above, according to this embodiment, preprocessing 1 can reduce estimation errors due to large fundamental components when estimating the optimal values ​​of amplitude Ai and phase φi that minimize the error with the residual waveform. Furthermore, preprocessing 2 narrows the target frequency range, thereby reducing the number of times the analysis frequency is updated and preventing the amount of calculation from becoming excessive. These features enable efficient and highly accurate estimation of the sideband frequencies. Note that the process described in the flowchart of FIG. 7 is merely an example and is not limiting. For example, the process of setting a limited frequency range as the target frequency range can be omitted. Furthermore, the number of fundamentals to be removed is not limited to one. If there are multiple waveforms with large amplitudes that affect the estimation of the peak frequency, the frequencies of those waveforms may be set to ω0 and steps S23 to S24 may be performed multiple times (the processes are performed on the residual waveform from the second time onward). The waveform data after removing the multiple fundamentals may then be used as the initial waveform to be evaluated, and steps S3 to S17 may then be performed. Furthermore, the detection target is not limited to sidebands, and the peak frequency within a certain frequency range can be accurately estimated.

[0033] FIG. 8 is a diagram illustrating an example of a hardware configuration of a signal analyzing device according to each embodiment. The computer 900 includes a CPU 901, a main storage device 902, an auxiliary storage device 903, an input / output interface 904, and a communication interface 905. The above-described signal analysis device 10, the web application system 20, the terminal 23, the application support system 30, and the OCR system 40 are each implemented in the computer 900. The above-described processes are stored in the auxiliary storage device 903 in the form of a program. The CPU 901 reads the program from the auxiliary storage device 903, loads it into the main storage device 902, and executes the above-described processes in accordance with the program. The CPU 901 also allocates a storage area in the main storage device 902 in accordance with the program. The CPU 901 also allocates a storage area in the auxiliary storage device 903 for storing data being processed in accordance with the program.

[0034] In at least one embodiment, the auxiliary storage device 903 is an example of a non-transitory tangible medium. Other examples of non-transitory tangible media include a magnetic disk, a magneto-optical disk, a CD-ROM, a DVD-ROM, and a semiconductor memory connected via the input / output interface 904. Furthermore, when this program is distributed to the computer 900 via a communication line, the computer 900 that receives the program may load the program into the main storage device 902 and execute the above-described processing. The program may also be for realizing part of the above-described functions. Furthermore, the program may be a so-called differential file (differential program) that realizes the above-described functions in combination with another program already stored in the auxiliary storage device 903.

[0035] In addition, the components in the above-described embodiments can be replaced with well-known components as appropriate without departing from the spirit of the present invention. Furthermore, the technical scope of the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the spirit of the present invention.

[0036] <Additional Notes> The signal analyzing device, the signal analyzing method, and the program according to each embodiment can be understood, for example, as follows.

[0037] (1) A signal analysis device 10 according to a first aspect includes a control unit (control unit 13, analysis unit 131) that performs generalized harmonic analysis (GHA) on waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range to calculate a first peak frequency, performs generalized harmonic analysis on the waveform data while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency, and calculates the average (average value, weighted average value) of the first peak frequency and the second peak frequency, thereby estimating the peak frequency of the waveform data. This allows the peak frequency to be estimated with high accuracy by generalized harmonic analysis.

[0038] (2) A signal analysis device 10 according to a second aspect is the signal analysis device 10 of (1), in which the control unit (control unit 13, analysis unit 131) generates second waveform data by removing a fundamental wave (e.g., a power supply frequency of 50 kHz, 60 kHz, etc.) from the waveform data, calculates the first peak frequency and the second peak frequency for the second waveform data, and estimates the peak frequency (frequency of a sideband wave) of the second waveform data by calculating the average (average value, weighted average value) of the first peak frequency and the second peak frequency. By removing the fundamental wave, which has a large effect on the GHA, the frequency of the sideband waves can be estimated accurately using generalized harmonic analysis.

[0039] (3) The signal analysis device 10 according to the third aspect is the signal analysis device 10 of (2), wherein the control unit (control unit 13, analysis unit 131) sets the frequency of the fundamental wave as the analysis frequency, performs generalized harmonic analysis, and generates the second waveform data by using the residual waveform data obtained by performing the generalized harmonic analysis as the second waveform data. This allows the fundamental frequency component to be removed using the calculation method of generalized harmonic analysis.

[0040] (4) A signal analysis device 10 according to a fourth aspect is the signal analysis device 10 of (2) to (3), in which the control unit calculates the first peak frequency and the second peak frequency for a predetermined local frequency range that includes the peak frequency of the second waveform data, and estimates the peak frequency of the second waveform data by calculating the average (average value, weighted average value) of the first peak frequency and the second peak frequency. This reduces the amount of calculation required for generalized harmonic analysis, and allows the peak frequency of the second waveform data (for example, the frequency of a sideband) to be estimated efficiently.

[0041] (5) A signal analysis method according to a fifth aspect includes the steps of: calculating a first peak frequency by performing generalized harmonic analysis on waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range; calculating a second peak frequency by performing generalized harmonic analysis on waveform data while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range; and estimating the peak frequency of the waveform data by calculating the average of the first peak frequency and the second peak frequency.

[0042] (6) A program according to a sixth aspect causes a computer to execute the following steps: calculating a first peak frequency by performing generalized harmonic analysis on waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range; calculating a second peak frequency by performing generalized harmonic analysis on waveform data to be evaluated while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range; and estimating the peak frequency of the waveform data by calculating the average of the first peak frequency and the second peak frequency. [Explanation of symbols]

[0043] 10 Signal analysis device, 11 Waveform data acquisition unit, 12 Setting reception unit, 13 Control unit, 131 Analysis unit, 14 Output unit, 15 Storage unit, 900 Computer, 901 CPU, 902 Main storage unit, 903 Auxiliary storage unit, 904 Input / output interface, 905 Communication interface

Claims

1. a control unit that performs generalized harmonic analysis on waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range to calculate a first peak frequency, performs generalized harmonic analysis on waveform data to be evaluated while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency, and calculates an average of the first peak frequency and the second peak frequency, thereby estimating the peak frequency of the waveform data; A signal analysis device comprising:

2. the control unit generates second waveform data by removing a fundamental wave from the waveform data, calculates the first peak frequency and the second peak frequency for the second waveform data, and estimates the peak frequency of the second waveform data by calculating an average of the first peak frequency and the second peak frequency. The signal analyzer according to claim 1 .

3. the control unit sets the frequency of the fundamental wave as an analysis frequency, performs generalized harmonic analysis, and defines residual waveform data obtained by performing the generalized harmonic analysis as the second waveform data, thereby generating the second waveform data. The signal analyzer according to claim 2 .

4. the control unit calculates the first peak frequency and the second peak frequency for a predetermined local frequency range including the peak frequency of the second waveform data, and estimates the peak frequency of the second waveform data by calculating an average of the first peak frequency and the second peak frequency.

4. The signal analyzing device according to claim 2 or 3.

5. A computer-implemented signal analysis method, comprising: a step in which the computer performs generalized harmonic analysis on the waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range, thereby calculating a first peak frequency; The computer performs generalized harmonic analysis while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency; the computer estimating a peak frequency of the waveform data by calculating an average of the first peak frequency and the second peak frequency; A signal analysis method comprising:

6. On the computer, a step of calculating a first peak frequency by performing generalized harmonic analysis on the waveform data to be evaluated while increasing an analysis frequency from the lower limit to the upper limit of a predetermined frequency range; performing generalized harmonic analysis while decreasing the analysis frequency from the upper limit to the lower limit of the frequency range to calculate a second peak frequency; estimating a peak frequency of the waveform data by calculating an average of the first peak frequency and the second peak frequency; A program that executes the following.

Images