JP5375701B2 - Method, apparatus and program for estimating rotational speed of rotating machine - Google Patents

Description

  The present invention relates to a rotational speed estimation method, apparatus, and program for a rotating machine that estimates the rotational speed of a rotating machine that changes in rotational speed.

  As a method for monitoring equipment deterioration of a rotating machine having a rotating part, vibration diagnosis by vibration analysis is used. When diagnosing vibration of a rotating machine whose rotational speed changes, for example, as disclosed in Patent Document 1, the rotational speed is measured simultaneously with the measurement of vibration, and the vibration value is corrected by the rotational speed. Eliminate fluctuations and determine vibration changes due to equipment deterioration.

Japanese Patent Laid-Open No. 7-218333

  Conventionally, when diagnosing vibration with a vibration diagnostic device, the rotating machine whose rotational speed varies, in addition to the signal of the vibration sensor installed in the rotary machine, the rotational speed is controlled from the control panel of the rotating machine. Information is extracted and entered. That is, the vibration diagnosis apparatus is connected to a vibration sensor installed in the rotating machine and a control panel in an electrical room or the like. However, depending on the installation location of the vibration diagnostic apparatus, it may be desirable to omit the wiring for inputting the rotational speed information. In particular, when the wiring distance between the control panel and the vibration diagnostic apparatus becomes long, the wiring cost becomes high.

  In addition, instead of taking in the information on the rotational speed from the control panel, a configuration in which a rotational speed sensor is installed in the rotating machine and connected to the vibration diagnosis apparatus is also conceivable. However, installing a rotation speed sensor in addition to the vibration sensor increases the cost. In addition, the rotation speed sensor may be attached with a target for detecting shaft rotation, and is generally more likely to fail than the vibration sensor or the like, and is often inferior in reliability. Furthermore, depending on the structure of the rotating machine, it may be difficult to install the rotation speed sensor.

  The present invention has been made in view of the above points, and an object of the present invention is to make it possible to estimate the rotational speed of a rotating machine whose rotational speed changes using a vibration signal of the rotating machine.

The rotational speed estimation method for a rotary machine according to the present invention is a rotational speed estimation method for a rotary machine that estimates the rotational speed of a rotary machine whose speed changes and the upper and lower limits of the speed change rate per time are known. Te, and measuring a vibration signal of the rotary machine by the vibration sensor, a frequency analysis step of frequency analyzing the vibration signal measured by the vibration sensor, a real number multiple of the peak spectrum of the rotational frequency of the rotating machine is observed A peak spectrum frequency extraction step for extracting a peak spectrum frequency in the frequency band from a result of frequency analysis performed in the frequency analysis step, and a peak spectrum frequency extracted in the peak spectrum frequency extraction step. Is divided by the real number times to calculate the rotational frequency, and the rotational speed for calculating the estimated rotational speed of the rotating machine Possess a step out, the in frequency analysis step, the vibration signal of the time series of a rotary machine frequency analysis with a predetermined time interval, with the peak spectral frequency extraction step, each timing of the frequency analysis The peak spectral frequency in the frequency band is extracted, and the rate of change in the number of revolutions in the predetermined time interval is calculated based on the peak spectral frequency in the frequency band at each timing extracted in the peak spectral frequency extracting step. The rotation speed change rate determining step for determining the effective rotation frequency when the known rotation speed change rate is within the upper and lower limit values is performed .
Another feature of the rotational speed estimation method for a rotating machine according to the present invention is that in the rotational speed calculation step, an average value or a median value of all rotational frequencies validated in the rotational speed change rate determination step. Is calculated as the estimated rotational speed, or the rotational speed is calculated from all the rotational frequencies validated in the rotational speed change rate determination step, and the average value of all the rotational speeds is calculated. Or it is in the point which calculates | requires a median value and makes it the said estimated rotation speed.
The rotational speed estimation device for a rotary machine of the present invention is a rotational speed estimation device for a rotary machine that estimates the rotational speed of a rotary machine whose rotational speed changes and whose upper and lower limit values of the rotational speed change rate per time are known. The frequency analysis means for analyzing the frequency of the vibration signal of the rotating machine measured by the vibration sensor, and a frequency band in which a peak spectrum that is a real number multiple of the rotation frequency of the rotating machine is preset, and the frequency analysis Based on the result of frequency analysis by means, the peak spectrum frequency extraction means for extracting the peak spectrum frequency in the frequency band and the peak spectrum frequency extracted by the peak spectrum frequency extraction means are divided by the real number times to calculate the rotation frequency. and, a speed calculating means for calculating an estimated rotational speed of the rotating machine, wherein the frequency analyzing means, the time series of the rotating machine A frequency analysis is performed on a moving signal at a predetermined time interval, and the peak spectrum frequency extraction unit extracts a peak spectrum frequency in the frequency band at each timing of performing the frequency analysis, and the peak spectrum frequency extraction unit Based on the extracted peak spectrum frequency in the frequency band at each timing, the rotational speed change rate at the predetermined time interval is obtained, and is effective when it is within the upper and lower limits of the known rotational speed change rate. A rotation speed change rate determination means for determining the rotation frequency is further provided .
The program of the present invention is a program for estimating the rotational speed of a rotating machine in which the rotational speed changes and the upper and lower limit values of the rotational speed change rate per time are known, and the rotating machine measured by a vibration sensor Frequency analysis processing for frequency analysis of the vibration signal of, and a frequency band in which a peak spectrum that is a real number multiple of the rotational frequency of the rotating machine is set in advance, and from the result of frequency analysis by the frequency analysis processing, the frequency A peak spectral frequency extraction process for extracting a peak spectral frequency in the band, and a rotational frequency is calculated by dividing the peak spectral frequency extracted by the peak spectral frequency extraction process by the real number multiple, and an estimated rotational speed of the rotating machine is calculated. to execute a speed calculating process of calculating in a computer, said at frequency analysis processing, vibration time series of said rotating machine The signal is frequency-analyzed at a predetermined time interval, and in the peak spectrum frequency extraction process, the peak spectrum frequency in the frequency band is extracted at each timing when the frequency analysis is performed, and is extracted by the peak spectrum frequency extraction process The rotation rate change rate at the predetermined time interval is obtained based on the peak spectrum frequency in the frequency band at each timing, and if the rotation rate is within the upper and lower limit values of the known rotation rate change rate, effective rotation the speed change rate determination process determines that the frequency Ru is performed further in the computer.

  According to the present invention, it is possible to estimate the rotational speed of a rotating machine whose rotational speed changes using the vibration signal of the rotating machine. Therefore, it is possible to estimate the number of revolutions with the vibration diagnostic device without taking in information on the number of revolutions from the control panel or installing a revolution number sensor in the rotating equipment, thereby reducing costs and improving reliability. Can be achieved.

It is a figure which shows the function structure of the radio vibration diagnostic apparatus which concerns on embodiment. It is a characteristic view which shows a vibration time series waveform, rotation speed change, and a rotation pulse waveform. It is a figure which shows the driving | running pattern of the 1st, 2nd winding drum. It is a characteristic view which shows the FFT waveform in which the harmonic vibration of the real number multiple of a rotation frequency has generate | occur | produced. It is a characteristic view which shows the FFT waveform of a gear vibration. It is a figure for demonstrating the frequency band in which the peak spectrum of the real number multiple of a rotation frequency is observed. It is a flowchart which shows the detailed process example by the radio | wireless vibration diagnostic apparatus which concerns on embodiment. It is a flowchart which shows the frequency-analysis process for measuring the several FFT waveform used for rotation speed estimation. It is a flowchart which shows a rotation speed estimation process. It is a figure which shows schematic structure of a carousel reel installation. It is a side view of a carousel reel installation. It is a top view of a carousel reel installation.

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings.
First, a carousel reel facility will be described as a rotating machine to be subjected to vibration diagnosis in the present embodiment. FIG. 10 shows a schematic configuration of a carousel reel facility. FIG. 11 is a side view of the carousel reel facility, and FIG. 12 is a plan view. The carousel reel equipment continuously rewinds the transported strip by revolving two revolving winding drums 101 and 102 by a revolving speed increaser 103 and alternately switching their rewinding positions. .

  As shown in FIG. 10, a first motor 104 for rotating the first winding drum 101 and a second motor 105 for rotating the second winding drum 102 are connected to the speed reducer 106. To do. The output shaft 107 of the speed reducer 106 has a double structure having an outer and an inner. The output of the first motor 104 is transmitted to the shaft 107 (for example, the outer shaft) via the gear group 106 a in the speed reducer 106, and the first winding is transmitted via the gear group 103 a in the revolution speed increaser 103. Is transmitted to the drum 101. The output of the second motor 105 is transmitted to a shaft 107 (for example, an inner shaft) via a gear group 106 a in the speed reducer 106, and the second winding is transmitted via the gear group 103 a in the revolution speed increaser 103. Is transmitted to the drum 102. In addition, the number described in the gear represents the number of teeth.

  In such a carousel reel facility, as shown in FIGS. 11 and 12, a bearing case that supports the rotating shaft of the first winding drum 101 in order to perform vibration diagnosis of the speed increaser 103 that is a revolving part. A vibration acceleration sensor 108 is installed as a vibration sensor, and a vibration acceleration sensor 108 is installed as a vibration sensor in a bearing case that supports the rotating shaft of the second winding drum 102.

  Further, as shown in FIG. 11, the wireless vibration diagnostic device 100 is installed on the side surface of the revolving speed increaser 103 that revolves. The wireless vibration diagnostic apparatus 100 samples a signal from the vibration acceleration sensor 108, and details processing, which will be described later, executes processing such as frequency analysis and rotational speed estimation necessary for vibration diagnosis, and obtained as a result. Digital wireless transmission of data to external devices.

  In addition, a pair of power feeders 109a constituting a non-contact power feeding device are arranged around the revolution speed increaser 103 so as to face each other in the radial direction. And the power receiver 109b which comprises a non-contact electric power feeder is arrange | positioned in the appropriate place of the revolution gearbox 103, and this power receiver 109b connects to the radio | wireless vibration diagnostic apparatus 100. FIG. When the revolution speed increaser 103 revolves and the first winding drum 101 is in the winding position and when the winding drum 102 is in the winding position, the power receiver 109b approaches the power feeder 109a and wireless. Electric power is supplied to the vibration diagnostic apparatus 100.

  In addition, a pair of measurement sensors (proximity switches) 110 are disposed at appropriate positions of the revolution gearbox 103. Due to the relationship between the measurement sensor 110 and the target 111 arranged around the revolution gearbox 103, the radio vibration diagnostic apparatus 100 takes up the winding drum (strip winding) at the winding position among the winding drums 101 and 102. It is possible to recognize which drum is in the middle.

  As shown in FIG. 2, the carousel reel facility of the present embodiment has an operation pattern in which the rotation speeds of the winding drums 101 and 102 change from moment to moment. That is, during the winding, the number of rotations of the winding drum is gradually decreased (measurement section A), and after the end of winding, the number of rotations is rapidly decreased (measurement section B). As shown in FIG. 3, this operation pattern is repeated alternately by the first and second winding drums 101 and 102.

  Hereinafter, the configuration and processing of the wireless vibration diagnostic apparatus 100 according to the present embodiment will be described. In the present embodiment, the wireless vibration diagnostic apparatus 100 functions as a rotation speed estimation apparatus to which the present invention is applied.

  FIG. 1 shows a functional configuration of a wireless vibration diagnostic apparatus 100 according to the present embodiment. Reference numeral 1 denotes an input unit to which a signal from the vibration acceleration sensor 108 (vibration signal from a bearing of a carousel reel facility) is input. Reference numeral 2 denotes a frequency analysis unit, which performs frequency analysis on the signal from the vibration acceleration sensor 108 input to the input unit 1 to obtain an FFT waveform.

  Reference numeral 3 denotes a rotation speed estimation unit that calculates an estimated rotation speed of the equipment. FIG. 4 is a characteristic diagram showing an FFT waveform in which the frequency is on the x-axis (in FIG. 4, the axis extending horizontally on the paper surface), the rotational speed is on the y-axis (the axis extending in front and back in FIG. 4), and the amplitude is on the z-axis. is there. In the vibration signal of the rotating machine, harmonic vibrations that are real times (order times) of the rotational frequency of the rotating machine may occur. Also in the example shown in FIG. 4, it can be seen that harmonic vibrations that are a real number times the rotational frequency are generated. The rotational speed estimation unit 3 utilizes the phenomenon that harmonic vibrations that are a real number multiple of the rotational frequency are generated in the vibration signal of the rotating machine, and determines the rotational frequency from the vibration characteristics and the order of harmonic vibration that can be known in advance. And the estimated number of revolutions of the equipment is calculated.

  A diagnostic unit 4 corrects the vibration value obtained from the vibration acceleration sensor 108 with the rotation number estimated by the rotation number estimation unit 3, and then determines a vibration change due to equipment deterioration. Reference numeral 5 denotes an output unit that digitally transmits the diagnosis result of the diagnosis unit 4 and the rotation number estimated by the rotation number estimation unit 3 to an external device. As another hardware configuration, in the device installed in the carousel reel main body, the FFT data is digitally transmitted as far as frequency analysis 2 and is connected to a communication network or other external device such as a computer, the rotational speed estimation unit 3, the diagnosis unit 4. It is good also as a structure with the output part 5. FIG.

  Next, an outline of the rotational speed estimation method in the present embodiment will be described. First, by measuring the absolute value of the vibration value obtained from the vibration acceleration sensor 108 and the threshold value of the change rate of the vibration value, the measured vibration value is measured in the measurement section A (rotation state of 300 rpm or more as shown in FIG. 2) and measurement section B ( As shown in FIG. 2, it is estimated whether the rotation state is less than 300 rpm. Then, an FFT waveform, which is frequency analysis data, is obtained for the measurement section A. 5A and 5B show examples of FFT waveforms in the measurement section A. FIG. 5A and 5B, the horizontal axis represents frequency and the vertical axis represents amplitude. The analysis conditions are a 2 kHz analysis range, 400 lines, 95% overlap of FFT averaging, and a frequency resolution of 5 Hz.

  The basic idea is that a frequency band B in which a peak spectrum S that is a real number N times the rotational frequency of the equipment is observed is set in advance, and the peak spectrum frequency S in the frequency band B is extracted. Then, the rotational frequency fr is calculated by dividing the peak spectral frequency S by the real number N, and the estimated rotational speed of the equipment is calculated.

If the peak spectrum which varies in proportion to the rotation frequency is several types, the frequency band B _i of the peak spectrum is observed in the real number N _i of the rotational frequency is set in advance respectively in each frequency band B _i The peak spectral frequency S _i is extracted. Then, to calculate the rotation frequency fr _i by dividing the peak spectral frequency S _i a real number multiple N _i, and calculates the estimated rotational speed of the equipment.

  For example, as shown in FIG. In this example, the fluctuation range of the rotation speed of the equipment is 300 to 500 rpm set as the operation condition (see FIG. 2), and the range B in which the spectrum of the rotation frequency appears is 5.0 to 8.3 Hz.

Based on the range B in which the spectrum of the rotation frequency appears, the frequency band in which the first gear meshing frequency appears is defined as B ₁ (order N ₁ = 44), and the frequency band in which the third gear meshing frequency appears. Is defined as B ₂ (order N ₂ = 44 × 3). Although the excitation source is unknown, a peak spectrum may appear at a constant real number multiple with respect to the rotation frequency, and its frequency band is defined as B ₃ (in the example of FIG. 6, the order N ₃ = 335. 5). In this example, the peak spectrum that can theoretically be generated, for example, the second, fourth, and fifth orders of the gear meshing frequency are always smaller than the third order of the actually generated gear meshing frequency. Are confirmed in advance, and those frequency bands are not set.

The FFT waveform is measured j times at a predetermined time interval (for example, every fixed time t seconds), and the peak spectral frequency S _ij of the frequency band B _i at each timing is extracted and divided by the real number N _i. To calculate the rotation frequency fr _ij . For example, the j-th measurement data is stored in the buffer memory at the first, eighth and fifteenth FFT 16-time averaging, and j = 1 to 3.

Here, in order to simplify the description, j = 1 and 2 will be described. The FFT waveform measurement interval (FFT averaging shift time) t at the first and eighth FFT 16-time averaging is
t = (400/2000) × 0.05 × (8-1) = 0.07 sec
It is.

Peak spectrum frequencies S ₁₁ = 250 Hz, S ₂₁ = 810 Hz, S ₁₂ = 245 Hz, and S ₂₂ = 805 Hz were extracted in the first and eighth FFT 16-time averaging. Therefore, calculating the rotation frequency fr _ij by dividing a peak spectral frequency S _ij a real number multiple _{N i, fr 11 = 5.68Hz,} fr 21 = 6.14Hz, fr 12 = 5.57Hz, fr 22 = 6. 10 Hz.

Next, the rotational speed change rate ΔREV _ij = (S _{i (j−1)} −S _ij ) · 60 / N _i / t (=
(Fr _{i (j−1)} −fr _ij ) · 60 / t) is calculated. And the threshold value (upper / lower limit value) of the rotational speed change rate of the said equipment is grasped | ascertained beforehand from equipment specification or operation conditions, and rotational speed change rate determination is performed. Rotational speed change rate ΔREV _ij is
ΔREV ₁₂ = (250-245) /44/0.07=1.623 rpm / s
ΔREV ₂₂ = (810−805) / (44 × 3) /0.07=0.541 rpm / s
It is. Then, the threshold of the speed change ratio that is preset in the equipment is the upper limit 4 rpm / s, the lower limit 1 rpm / s (see Figure 2), the speed change rate DerutaREV ₁₂ is within upper and lower limit However, the rotational speed change rate ΔREV ₂₂ is out of the lower limit value. In this case, the rotational frequency fr _21, fr ₂₂ estimated from the peak spectral frequency S _21, S ₂₂ is a parameter of the speed change ratio DerutaREV ₂₂ discards.

Next, an average value of all the rotation frequencies fr _ij remaining without being discarded in the rotation speed change rate determination is obtained and set as the rotation frequency of the equipment. Here, 5.63 Hz which is an average value of fr ₁₁ = 5.68 Hz and fr ₁₂ = 5.57 Hz is obtained as the rotation frequency of the equipment. And if the rotation frequency of the said installation is obtained, the estimated rotation speed of the said installation can be calculated by multiplying 60. Here, the rotation speed is calculated after _obtaining the average value of all the rotation frequencies fr _ij remaining without being discarded, but the rotation speed is calculated from all the rotation frequencies fr _ij remaining without being discarded. And you may make it obtain | require the average value of all the rotation speeds. Further, although the average value is obtained, the median value may be obtained.

In the above example, j = 1 and 2 have been described for the sake of simplicity. However, when j = 3 is considered, for example, as described above, the rotational speed change rate ΔREV ₂₂ deviates from the upper and lower limit values. It is also assumed that the number change rate ΔREV ₂₃ is within the upper and lower limit values. In this case, only the rotational frequency fr ₂₁ is discarded, and the rotational frequencies fr ₂₂ and fr ₂₃ remain. That is, if the rotational speed change rate ΔREV ₂₃ is within the upper and lower limit values, both parameters (rotational frequencies fr ₂₂ , fr ₂₃ ) used for the calculation are considered to be effective, and the rotational speed change rate ΔREV outside the upper and lower limit values. _It is considered that there is a problem in one (rotation frequency fr ₂₁ ) among the parameters (rotation frequencies fr ₂₁ , fr ₂₂ ) used in the calculation of ₂₂ .

  7 to 9 show detailed processing examples by the wireless vibration diagnostic apparatus 100 according to the present embodiment. Here, management is performed by dividing into three stages according to the frequency range. Considering only the gear meshing frequency described above, it is sufficient to target the middle frequency band (for example, 3 Hz to 10 kHz), but other factors cause a peak spectrum that is a real number multiple of the rotational frequency in the low frequency band and the high frequency band. It is also possible. Further, the vibration diagnosis itself may use vibration values not only in the middle frequency band but also in the low frequency band and the high frequency band. Therefore, here, processing for handling a low to high frequency band will be described.

When the data collection timer is ON (step S1) and the measurement sensor 110 confirms that the winding drum 101 or the winding drum 102 is in the winding position (step S2), the process proceeds to step S3. In step S3, the signal (vibration value V ₁ ) of the vibration acceleration sensor 108 of the winding drum at the winding position is sampled. As shown in FIG. 12, when a plurality of vibration acceleration sensors 108 are installed on one winding drum, the signals of the plurality of vibration acceleration sensors 108 are sequentially sampled.

  Next, sampling and frequency analysis are performed to obtain an FFT waveform (step S4). Details of the frequency analysis processing in step S4 will be described later.

Next, in order to determine the section, the signal (vibration value V ₂ ) of the vibration acceleration sensor 108 of the winding drum at the winding position is sampled again (step S5). If the vibration values V ₁ and V ₂ are larger than the preset threshold value V _a (step S6), and if the vibration value change rate | V ₁ −V ₂ | is smaller than the preset threshold value α (step S6). S7), assuming that the vibration value is in the measurement section A, the process proceeds to step S8. In step S8, the estimated rotational speed by using the results of frequency analysis of the vibration value V _1. Details of the rotation speed estimation process in step S8 will be described later. When it is determined that the vibration value is in the measurement section B (steps S6 and S7), and when the rotation speed is not calculated in step S8 (step S9), the process proceeds to step S10 to perform a retry process.

When the rotational speed is calculated in step S8, the vibration value V ₁ is corrected with the rotational speed estimated in step S8, and then a vibration change due to equipment deterioration is determined. For example, the rotational speed correction calculation of the average vibration value is performed, and the trend management graph is plotted (steps S11 and S12). Also, the FFT abnormal frequency occupancy calculation is performed, and the abnormality probability calculation is output (steps S13 and S14). In addition, what is necessary is just to utilize a well-known thing about vibration diagnosis itself, The content is not limited.

FIG. 8 is a flowchart showing the frequency analysis processing in step S4. After the anti-aliasing filter processing is performed on the signal (vibration value V ₁ ) of the vibration acceleration sensor 108 (step S15), A / D conversion is performed (step S16).

  In order to obtain an FFT waveform in a low frequency band, first, integration processing is performed to convert the speed (step S17). Thereafter, resampling is performed to a frequency range necessary for frequency analysis of the low frequency band (step S18), frequency analysis (for example, FFT 16 times averaging) is performed (steps S19 and S20), and an FFT waveform is obtained (step S21). At this time, the FFT waveform used for averaging, for example, the first, eighth, and fifteenth FFT waveforms of FFT 16 times averaging is stored in the buffer memory 50.

  Further, in order to obtain an FFT waveform in the intermediate frequency band, resampling is performed to a frequency range necessary for frequency analysis in the intermediate frequency band (step S22), and frequency analysis (for example, FFT 16 times averaging) is performed (steps S23 and S24). The FFT waveform is obtained (step S25). At this time, the FFT waveform used for averaging, for example, the first, eighth, and fifteenth FFT waveforms of FFT 16 times averaging is stored in the buffer memory 50.

  In order to obtain an FFT waveform in a high frequency band, first, after high-pass filter processing (step S26), envelope processing is performed (step S27). Thereafter, resampling is performed to a frequency range necessary for frequency analysis of the high frequency band (step S28), frequency analysis (for example, FFT 16 times averaging) is performed (steps S29 and S30), and an FFT waveform is obtained (step S31). At this time, the FFT waveform used for averaging, for example, the first, eighth, and fifteenth FFT waveforms of FFT 16 times averaging is stored in the buffer memory 50.

  FIG. 9 is a flowchart showing the rotation speed estimation process in step S8. In the rotational speed estimation process, the FFT waveforms (first, eighth, and fifteenth FFT waveforms of FFT 16 times averaging) stored in the buffer memory 50 in the frequency analysis process of step S4 are used.

The frequency band B _i where the peak spectrum of the real number N _i times the rotational frequency of the equipment is observed is read (step S32). The frequency band B _i is preset as shown in FIG. 6 and is held in the wireless vibration diagnostic apparatus 100. Then, the peak spectrum frequency S _ij in each frequency band B _i is extracted (step S33). The processes in steps S32 and S33 are repeated for all frequency bands i and measurement times j.

Then, to calculate the rotation frequency fr _ij by dividing a peak spectral frequency S _ij in real number N _i (step S36).

Next, using the rotational frequency fr _ij calculated in step S36, the rotational speed change rate ΔREV _ij = (S _{i (j-1)} -S _ij ) · 60 / N _i / t = (fr _{i (j-1 )} -Fr _ij )
Calculate 60 / t (step S37). The rotation as compared to the threshold value of the rotational speed change rate which is previously set in the equipment (the upper limit DerutaREV _L, the lower limit value ΔREV _U) (step S38), the rotational speed changing rate DerutaREV _ij deviated from the upper and lower limit values The frequency fr _ij is discarded (step S39). In this case, as described above, when the two rotational speed change rates ΔREV _ij are calculated using a certain rotational frequency fr _ij ,
If both of them are out of the upper and lower limit values, the rotation frequency fr _ij is discarded, but if one of them is within the upper and lower limit values, the rotation frequency fr _ij is made effective.

Next, it is determined whether or not all the rotation frequencies fr _ij calculated in step S36 have been discarded (step S40). As a result, if all the rotation frequencies fr _ij are not discarded, the estimated rotation speed of the equipment is calculated from all the rotation frequencies (effective rotation frequencies) fr _ij that remain without being discarded (step S41). That is, an average value or a median value of all the effective rotation frequencies fr _ij is obtained to calculate the rotation number, and set as the estimated rotation number of the equipment. Alternatively, the number of rotations is calculated from all the effective rotation frequencies fr _ij, and the average value or median value of all the rotation numbers is obtained and used as the estimated rotation number of the equipment. On the other hand, if all the rotation frequencies fr _ij have been discarded, it is assumed that the estimated rotation speed of the equipment cannot be calculated in the current loop (step S42).

  As described above, the rotational speed of a rotating machine (carousel reel equipment) whose rotational speed changes can be estimated using the vibration signal of the rotating machine. Therefore, it is not necessary to take in information on the number of revolutions from the control panel or install a revolution number sensor in the rotating device, and it is possible to reduce costs and improve reliability. In particular, since wiring with the control panel is not required, the degree of freedom of the installation location of the vibration diagnostic device is greatly increased, and the wireless vibration diagnostic device 100 is installed in the revolution gearbox 103 that revolves as in this embodiment. Layout is also possible.

  Specifically, the above-described wireless vibration diagnostic apparatus 100 is based on a central processing unit that executes arithmetic processing, a CPU (computation program), various application programs, ROM (read only memory) that stores data, and the like. A RAM (Random Access Memory), which is a work area used when the CPU performs processing, and a storage unit such as a hard disk for storing calculation results and the like.

1: Input unit 2: Frequency analysis unit 3: Rotational speed estimation unit 4: Diagnosis unit 5: Output unit

Claims (4)

A rotational speed estimation method for a rotary machine that estimates the rotational speed of a rotary machine whose rotational speed changes and the upper and lower limits of the rotational speed change rate per time are known ,
Measuring a vibration signal of the rotating machine with a vibration sensor;
A frequency analysis step for frequency analysis of a vibration signal measured by the vibration sensor ;
A peak spectrum frequency for extracting a peak spectrum frequency in the frequency band from a result of performing a frequency analysis in the frequency analysis step in advance by setting a frequency band in which a peak spectrum that is a real number multiple of the rotational frequency of the rotating machine is observed. An extraction step;
The peak spectral frequencies extracted by the peak spectral frequency extracting step calculates the rotation frequency by dividing the real number, possess a speed calculating step of calculating an estimated rotational speed of the rotating machine,
In the frequency analysis step, frequency analysis is performed for a vibration signal of a time series of the rotating machine at a predetermined time interval,
In the peak spectrum frequency extraction step, a peak spectrum frequency in the frequency band at each timing for performing the frequency analysis is extracted,
Based on the peak spectral frequency in the frequency band at each timing extracted in the peak spectral frequency extraction step, the rotational speed change rate at the predetermined time interval is obtained, and the upper and lower limit values of the known rotational speed change rate If it is within the range, a rotation speed change rate determination step for determining an effective rotation frequency is performed . In the rotation speed calculation step, an average value or median value of all rotation frequencies validated in the rotation speed change rate determination step is obtained to calculate the rotation speed, and set as the estimated rotation speed, or the rotation speed each calculated rotational speed from all rotational frequencies that are effective in changing rate determining step of claim 1, which the average value or the median of all of the rotational speed, characterized in that said estimated rotation speed The rotational speed estimation method of the rotary machine as described in 2. A rotational speed estimation device for a rotating machine that estimates the rotational speed of a rotating machine whose rotational speed changes and the upper and lower limit values of the rotational speed change rate per time are known ,
Frequency analysis means for frequency analysis of the vibration signal of the rotating machine measured by a vibration sensor;
A peak spectrum frequency for extracting a peak spectrum frequency in the frequency band from a result of frequency analysis by the frequency analysis means in advance by setting a frequency band in which a peak spectrum of a real number times the rotational frequency of the rotating machine is observed Extraction means;
A rotation frequency calculating means for calculating a rotation frequency by dividing the peak spectrum frequency extracted by the peak spectrum frequency extraction means by the real number multiple, and calculating an estimated rotation speed of the rotating machine ,
The frequency analysis means performs frequency analysis on a time series vibration signal of the rotating machine at a predetermined time interval,
The peak spectrum frequency extraction means extracts a peak spectrum frequency in the frequency band at each timing of performing the frequency analysis,
Based on the peak spectrum frequency in the frequency band at each timing extracted by the peak spectrum frequency extracting means, the rotation rate change rate at the predetermined time interval is obtained, and the upper and lower limit values of the known rotation rate change rate A rotation speed estimation device for a rotating machine, further comprising a rotation speed change rate determination means for determining that the rotation frequency is an effective rotation frequency . A program for estimating the rotational speed of a rotating machine whose rotational speed changes and whose upper and lower limits of the rotational speed change rate per time are known ,
Frequency analysis processing for frequency analysis of the vibration signal of the rotating machine measured by the vibration sensor;
A peak spectrum frequency for extracting a peak spectrum frequency in the frequency band from a result of performing a frequency analysis in the frequency analysis process in advance by setting a frequency band in which a peak spectrum that is a real number multiple of the rotational frequency of the rotating machine is observed. Extraction process,
Dividing the peak spectral frequency extracted in the peak spectral frequency extraction process by the real number multiple to calculate a rotational frequency, and causing the computer to execute a rotational speed calculation process for calculating an estimated rotational speed of the rotating machine ,
In the frequency analysis process, the frequency analysis of the vibration signal of the time series of the rotating machine is performed at a predetermined time interval,
In the peak spectrum frequency extraction process, the peak spectrum frequency in the frequency band at each timing for performing the frequency analysis is extracted,
Based on the peak spectrum frequency in the frequency band at each timing extracted by the peak spectrum frequency extraction process, the rotation rate change rate at the predetermined time interval is obtained, and the upper and lower limit values of the known rotation rate change rate If there within, yet because of the program cause the computer to execute a rotational speed change rate determination process for determining an effective rotational frequency.

Images