Condition monitoring device and condition monitoring method

Abstract

The present invention provides a monitoring device that can identify data during periods when a rotating part is experiencing rotational fluctuations, and use that identified data to improve diagnostic accuracy. [Solution] The state monitoring device 20 includes a receiving unit 21 that receives sensor signals from a sensor 12 that monitors the state of the device to be monitored 10 and speed signals indicating the rotational speed of the rotating part, a storage device 24 that stores measurement data including sensor information and speed information corresponding to the sensor signals and speed signals received by the receiving unit 21, and a calculation device 23 that analyzes the measurement data. The calculation device 23 is configured to extract data from the measurement data that includes sections in which the rotational speed is fluctuating, and to perform tracking processing that analyzes the sensor information in the extracted data in accordance with the change in rotational speed.

Description

Technical Field

[0001] The present disclosure relates to a state monitoring device and a state monitoring method.

Background Art

[0002] For the fault diagnosis of various rotating devices, etc., a monitoring device that acquires data indicating the state of a device, such as vibration data and temperature data, is known. At that time, in order to estimate the cause of an abnormality, a method of performing frequency analysis of vibration data is used.

[0003] Since the rotational speeds of rotating shafts such as windmill shafts and wheels vary, the frequency of vibration to be analyzed also changes in a state monitoring device, and the frequency spectrum fluctuates. For this reason, various countermeasures have been studied against fluctuations in rotational speed.

[0004] For example, Japanese Patent Application Laid-Open No. 2015-34776 (Patent Document 1) discloses a method of fitting acquired data with a basis function having the rotational position of a rotating shaft as a variable. Thereby, frequency components related to the rotation of the rotating shaft can be extracted without being affected by changes in rotational speed.

[0005] Also, for example, Japanese Patent No. 6665062 (Patent Document 2) discloses a method that does not use a rotation sensor. In this method, the acquired data is divided, and the rotational speed is estimated for each divided data. The divided data is corrected based on the estimated rotational speed, and the corrected data is combined and analyzed. Thereby, the influence of rotational fluctuations can be reduced and accurate analysis can be performed.

Prior Art Documents

Patent Documents

[0006]

Patent Document 1

Patent Document 2

Summary of the Invention

[0007] Japanese Patent Publication No. 2015-34776 (Patent Document 1), etc., shows an example of a signal processing method for eliminating the effects of rotational speed fluctuations. Applying this technology can improve diagnostic accuracy, but this requires acquiring data with a certain degree of rotational fluctuation. However, there is no disclosed method for specifically acquiring data with rotational fluctuations, and there is a problem in that it is uncertain whether the effect of improving diagnostic accuracy can be enjoyed depending on the content of the acquired data.

[0008] The purpose of this disclosure is to provide a monitoring device that can identify data during periods when a rotating part is experiencing rotational fluctuations and use that identified data to improve diagnostic accuracy. [Means for solving the problem]

[0009] This disclosure relates to a condition monitoring device for monitoring the state of a device having a rotating part. The condition monitoring device includes a receiving unit that receives sensor signals from a sensor that monitors the state of the device and a speed signal indicating the rotational speed of the rotating part, a storage device that stores measurement data including sensor information and speed information corresponding to the sensor signal and speed signal received by the receiving unit, respectively, and a computing device that analyzes the measurement data. The computing device is configured to extract data from the measurement data that includes a section in which the fluctuation of the rotational speed exceeds a standard, and to perform tracking processing that analyzes the sensor information in the extracted data in accordance with the change in rotational speed. [Effects of the Invention]

[0010] According to the condition monitoring device disclosed herein, data from the period during which the rotational fluctuation of the rotating part exceeds a standard is identified and used for diagnosis, thereby improving diagnostic accuracy. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing the configuration of the condition monitoring device according to this embodiment. [Figure 2]It is a flowchart for explaining the data processing executed by the state monitoring device 20. [Figure 3] It is a waveform diagram for explaining the first determination method of rotational fluctuation. [Figure 4] It is a waveform diagram showing the determination result by the first determination method of rotational fluctuation. [Figure 5] It is a waveform diagram for explaining the second determination method of rotational fluctuation. [Figure 6] It is a waveform diagram showing the determination result by the second determination method of rotational fluctuation. [Figure 7] It is a waveform diagram for explaining the third determination method of rotational fluctuation. [Figure 8] It is a waveform diagram showing the determination result by the third determination method of rotational fluctuation. [Figure 9] It is a waveform diagram for explaining the fourth determination method of rotational fluctuation. [Figure 10] It is a waveform diagram showing the determination result by the fourth determination method of rotational fluctuation. [Figure 11] It is a waveform diagram for explaining the fifth determination method of rotational fluctuation. [Figure 12] It is a waveform diagram showing the determination result by the fifth determination method of rotational fluctuation. [Figure 13] It is a waveform diagram for explaining the sixth determination method of rotational fluctuation. [Figure 14] It is a waveform diagram showing the determination result by the sixth determination method of rotational fluctuation. [Figure 15] [[ID=(此处ID=39翻译有误,原文中ID=39的内容是关于未进行部分数据选择而执行频率分析时的频谱图说明,翻译时应保持一致)]]It is a graph showing the spectrum when frequency analysis is performed without performing partial data selection. [Figure 16] It is a graph showing the spectrum when data during the rotational fluctuation period is selected and frequency analysis is performed.

Embodiments for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals and their description will not be repeated.

[0013] [Basic Configuration of Condition Monitoring Device] FIG. 1 is a block diagram showing the configuration of a condition monitoring device according to the present embodiment. The condition monitoring device dealt with in the present embodiment and its surroundings will be described using FIG. 1.

[0014] The monitoring target device 10 is a device to be monitored by the condition monitoring device and has a rotating part. For example, the monitoring target device 10 is a wind power generation facility, a machine tool, a robot, a vehicle, etc. The wind power generation facility includes, for example, the rotating shaft of a windmill as the rotating part. The machine tool includes, for example, a rotating shaft for attaching a tool such as a drill as the rotating part. The vehicle includes, for example, the rotating shaft of a wheel as the rotating part. The monitoring target device 10 may be a device other than the above. Any device including a rotating shaft and a bearing (the type is not limited, such as a ball bearing or a roller bearing) may correspond to the monitoring target device 10.

[0015] The control device 11 is a device that controls the operation of the monitoring target device 10. For example, a SCADA (Supervisory Control And Data Acquisition) which is a remote monitoring and control system in a wind power generation facility, a PLC (Programmable Logic Controller) that controls a machine tool, an ECU (Electronic Control Unit) in a vehicle, etc. may correspond to the control device 11.

[0016] The sensor 12 is a sensor for monitoring the state of the monitoring target device 10. For example, a vibration sensor, an acoustic sensor, an AE sensor, a displacement sensor, a proximity sensor, a temperature sensor, a current sensor, a torque sensor, etc. may correspond to the sensor 12.

[0017] The status monitoring device 20 is a system that estimates, diagnoses, and detects the status of the monitored device 10 by acquiring or measuring signals from the sensor 12 and the control device 11, and processing and analyzing them. Furthermore, the monitored device 10 may be controlled based on the results of this processing, analysis, estimation, diagnosis, and detection.

[0018] The display / server 30 stores and displays data acquired or measured by the status monitoring device 20, as well as the results of processing, analyzing, estimating, diagnosing, and detecting the data.

[0019] The status monitoring device 20 includes a receiving unit 21, an interface (I / F) unit 22, a CPU (Central Processing Unit) 23, a storage device 24, and a transmitting unit 25. The storage device 24 includes ROM (Read Only Memory), RAM (Random Access Memory), a hard disk, etc.

[0020] The CPU 23 is an arithmetic unit that implements the state monitoring method, which will be described in detail later, by executing various programs stored in the ROM of the storage device 24. The RAM of the storage device 24 is used as a work area by the CPU 23. The ROM of the storage device 24 records a program that includes each step of the flowchart (described later) showing the procedure of the state monitoring method. The receiving unit 21 is a device for reading data from the control device 11. The transmitting unit 25 is a device for outputting the calculation results from the CPU 23.

[0021] Figure 2 is a flowchart illustrating the data processing performed by the status monitoring device 20.

[0022] First, in step S1, the status monitoring device 20 performs data measurement processing. In the data measurement processing, the CPU 23 acquires data from the sensor 12 and the control device 11 via the receiving unit 21, which includes signals indicating operating information and information detected by the sensors. The CPU 23 may continuously send the acquired data to step S2, or it may temporarily store the acquired data in a sufficiently large buffer memory (ring buffer) contained in the storage device 24 and send it to step S2 as needed or requested. Alternatively, the CPU 23 may temporarily store the acquired data in the storage device 24 or server 30, and read it from the storage device 24 or server 30 and send it to step S2 as needed or requested.

[0023] Next, in step S2, the status monitoring device 20 performs data selection processing. The CPU 23 determines whether the acquired data is from a period in which the rotational speed fluctuation exceeds a standard (hereinafter referred to as the rotational speed fluctuation period), and selects or extracts the data from the acquired data for the rotational speed fluctuation period and sends it to step S3. The rotational speed included in the data may be the rotational speed directly measured by a speed sensor, proximity sensor, etc., or it may be the rotational speed acquired or estimated from the control signal of the monitored device 10, or it may be the rotational speed estimated from an acceleration signal, current signal, etc.

[0024] Next, in step S3, the status monitoring device 20 performs rotational speed tracking processing. Rotational speed tracking processing is a process that removes the effects of fluctuations in rotational speed, and is a process that resamples data acquired at normal fixed time intervals to time intervals corresponding to the rotational speed (or rotational angle). For example, as rotational speed tracking processing, the resampling processing described in Japanese Patent Publication No. 6665062 (Patent Document 2) can be used. However, this method is not limited to this method, and other resampling processing may be applied to rotational speed tracking processing.

[0025] Finally, in step S4, the condition monitoring device 20 performs analysis and diagnostic processing. In the analysis and diagnostic processing, the CPU 23 analyzes, diagnoses, and detects the state of the monitored device based on data from which the effects of rotational speed fluctuations have been reduced by the rotational speed tracking processing.

[0026] In the data processing described above, data for the rotational fluctuation period is selected in step S2. For example, there are six methods for determining whether the "rotational speed fluctuation exceeds the standard."

[0027] The first method is a determination method based on the difference between the minimum and maximum values ​​of the rotational speed. The second method is a determination based on an evaluation value (such as the standard deviation) that indicates the variation in rotational speed. The third method is a determination based on the minimum and maximum values ​​of the rotational speed. The fourth method is a determination based on the cumulative value of the derivative of the rotational speed. The fifth method is a determination based on the number of regions that the rotational speed has passed through, etc. The sixth method is a determination based on the value of the rotational speed after filtering. These determinations will be explained in order below.

[0028] [1. Judgment based on the difference between the minimum and maximum rotational speeds] Figure 3 is a waveform diagram illustrating the first method for determining rotational fluctuations. Figure 4 is a waveform diagram showing the determination result based on the first method for determining rotational fluctuations.

[0029] In Figure 3, waveform W1 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. Waveform W2 shows the speed change range (rpm), which is the difference between the maximum and minimum values ​​of the rotational speed over a certain period. For example, if the time from time t2 to t3 is a certain period Δt, the maximum value of the rotational speed during this period is 1970 rpm, and the minimum value is 1811 rpm. Therefore, the speed change range (rpm) at time t2 is the difference between 1970 - 1811 = 159. In other words, if the value of the speed change range (rpm) at time t is W2(t), then W2(t) is the maximum - minimum value of the rotational speed from time t until a certain period Δt has elapsed. Waveform W2 shows the results of performing the above calculation continuously for times t1 to t5.

[0030] Figure 4 shows how a threshold is set for waveform W2, and when the threshold is exceeded, it is determined that the rotational speed fluctuation exceeds a standard during the period Δt from the time when the threshold is exceeded. If waveform W2(t1s) exceeds the threshold at time t1s, data for a certain period Δt from time t1s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing. Similarly, if waveform W2(t4s) exceeds the threshold at time t4s, data for a certain period Δt from time t4s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing.

[0031] To simplify post-processing, data may be selected only from the point when waveform W2 exceeds the threshold, up to Δt, to reduce overlap in selected time periods. However, if overlap in selected time periods is acceptable, data may be selected from the point when waveform W2 exceeds the threshold, up to Δt.

[0032] Furthermore, while a fixed time interval Δt was used as the unit of the judgment interval, a fixed rotational speed ΔN may be used instead of a fixed time interval Δt. Also, while the difference between the maximum and minimum values ​​of the rotational speed in the judgment interval was used as the judgment target, the difference between the maximum and minimum values ​​obtained by performing calculations such as squaring on the rotational speed may also be used as the judgment target. Alternatively, instead of the difference between these maximum and minimum values, the value obtained by performing calculations such as squaring on the difference may be used as the judgment target. When any of the above judgment targets exceed a predetermined threshold, it can be determined that the fluctuation in rotational speed exceeds the standard.

[0033] [2. Judgment based on evaluation values ​​(such as standard deviation) that indicate the variation in rotational speed] Figure 5 is a waveform diagram illustrating the second method for determining rotational fluctuations. Figure 6 is a waveform diagram showing the determination results using the second method for determining rotational fluctuations.

[0034] In Figure 5, waveform W11 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. Waveform W12 shows the standard deviation value, which indicates the variation in rotational speed over a certain period. The standard deviation s is given by s = √(Σ(xi - xm)) 2 This is expressed as follows: Here, if n is the number of data points included in a certain period, then xi = x1, ..., xn, where xm represents the mean value. For example, if the time from time t12 to t13 is a certain period Δt, then the standard deviation s, which indicates the variation in rotational speed within this period, is 44.2. In other words, if the value of the velocity standard deviation s at time t is W12(t), then W12(t) is the standard deviation s of the rotational speed from time t until a certain period Δt has elapsed. Waveform W2 shows the results of performing the above calculation continuously from time t11 to t15.

[0035] Figure 6 shows how a threshold is set for waveform W12, and when the threshold is exceeded, it is determined that the rotational speed fluctuation exceeds a standard during the period Δt from the time when the threshold is exceeded. At time t12s, if waveform W12(t12s) exceeds the threshold, data for a certain period Δt from time t12s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing. Similarly, at time t14s, if waveform W12(t14s) exceeds the threshold, data for a certain period Δt from time t14s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing.

[0036] To simplify post-processing, data may be selected only from the point when waveform W12 exceeds the threshold, up to Δt, to reduce overlap in selected time periods. However, if overlap in selected time periods is acceptable, data may be selected from the point when waveform W12 exceeds the threshold, up to Δt.

[0037] Furthermore, while a fixed time interval Δt was used as the unit of the judgment interval, a fixed rotational speed ΔN may be used instead of a fixed time interval Δt. Also, while the standard deviation s of the rotational speed in the judgment interval was used as the judgment target, other evaluation values ​​that indicate the variation in rotational speed in the relevant interval, such as a value obtained by performing calculations such as squaring the rotational speed, may also be used as the judgment target. When any of the above judgment targets exceed a predetermined threshold, it can be determined that the variation in rotational speed exceeds the standard.

[0038] [3. Judgment based on minimum and maximum rotational speed] Figure 7 is a waveform diagram illustrating the third method for determining rotational fluctuations. Figure 8 is a waveform diagram showing the determination results using the third method for determining rotational fluctuations.

[0039] In Figure 7, waveform W21 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. Waveform W22 shows the result of determining whether the rotational speed exceeded the upper limit during a certain period (True) or not (False). Waveform W23 shows the result of determining whether the rotational speed fell below the lower limit during a certain period (True) or not (False). Waveform W24 shows the result of AND processing of waveforms W22 and W23.

[0040] For example, if we define the period Δt as a fixed time interval from time t22 to t23, then during this period, the waveform W21 falls between the upper and lower limits. Therefore, the upper limit determination result is False, and the lower limit determination result is also False. Consequently, both the upper and lower limit determination results at t22 are False, and the result of their AND operation is also False.

[0041] Figure 8 shows how the system determines that the rotational speed fluctuation exceeds a certain threshold based on the waveform W24 determination result. When waveform W24 changes from False to True at time t22s, data is selected as the rotational fluctuation period for a certain period Δt from time t22s, and this selected data is used in subsequent processing. Similarly, when waveform W24 changes from False to True at time t24s, data is selected as the rotational fluctuation period for a certain period Δt from time t24s, and this selected data is used in subsequent processing.

[0042] In the waveforms described above, at the time when waveform W22, which indicates the upper limit determination result, switches from False to True (rising point), the rotation speed after a certain period Δt exceeds the upper limit. Also, at the rising point of change in waveform W23, which indicates the lower limit determination result, the rotation speed after a certain period Δt falls below the lower limit. On the other hand, at the falling point of change in waveform W23, the rotation speed at the left end of the certain period Δt (at the same time as the falling point) exceeds the lower limit, and there is no portion within the certain period Δt where the rotation speed falls below the lower limit. In other words, if a rotation speed value of 2 or more is observed during a certain period Δt starting from a certain time, and the observed rotation speed values ​​include a first rotation speed value and a second rotation speed value, and the first rotation speed value exceeds the upper limit and the second rotation speed value falls below the lower limit, then that certain period Δt is selected as the rotation fluctuation period.

[0043] To simplify post-processing, data may be selected only from the point when waveform W24 changes from False to True, reducing the overlap of selected time periods. However, if overlapping of selected time periods is acceptable, data may be selected from the point when waveform W24 is True, starting from that point.

[0044] Furthermore, while a fixed time interval Δt was used as the unit of the judgment interval, a fixed rotational speed ΔN may be used instead of a fixed time interval Δt. Also, while the judgment target was whether the rotational speed in the judgment interval exceeds the upper limit or falls below the lower limit, a similar judgment may be performed using a value obtained by squaring or other calculations of the rotational speed as the judgment target. When any of the above judgment targets exceed a predetermined threshold, it can be determined that the fluctuation in rotational speed exceeds the standard.

[0045] [4. Judgment based on the integral of the derivative of the rotational speed] Figure 9 is a waveform diagram illustrating the fourth method for determining rotational fluctuations. Figure 10 is a waveform diagram showing the determination results using the fourth method for determining rotational fluctuations.

[0046] In Figure 9, waveform W31 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. Waveform W32 shows the square of the derivative of the rotational speed over a certain period Δt. Waveform W33 shows the cumulative value of waveform W32 over a certain period Δt. For example, the value of waveform W33 at t32 is the cumulative value of waveform W32 over a certain period Δt from time t32 to t33. Waveform W33 shows the results of performing the above calculations continuously from time t31 to t35.

[0047] Figure 10 shows how a threshold is set for waveform W33, and when the threshold is exceeded, it is determined that the rotational speed fluctuation exceeds a standard during the period Δt from that time. At time t32s, if waveform W33(t32s) exceeds the threshold, data for a certain period Δt from time t32s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing. Similarly, at time t34s, if waveform W33(t34s) exceeds the threshold, data for a certain period Δt from time t34s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing.

[0048] To simplify post-processing, data may be selected only from the point when waveform W33 exceeds the threshold, up to Δt, to reduce overlap in selected time periods. However, if overlap in selected time periods is acceptable, data may be selected from the point when waveform W33 exceeds the threshold, up to Δt.

[0049] Furthermore, while a fixed time interval Δt was used as the unit of the judgment interval, a fixed rotational speed ΔN may be used instead of a fixed time interval Δt. Also, while the square of the derivative of the rotational speed in the judgment interval was used as the judgment target, the derivative of the rotational speed itself or the sum of values ​​obtained by calculations other than squaring may also be used as the judgment target. When any of the above judgment targets exceed a predetermined threshold, it can be determined that the fluctuation of the rotational speed exceeds the standard.

[0050] [5. Determination based on the number of regions through which the rotation speed has passed.] Figure 11 is a waveform diagram illustrating the fifth method for determining rotational fluctuations. Figure 12 is a waveform diagram showing the determination results using the fifth method for determining rotational fluctuations.

[0051] In Figure 11, waveform W41 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. The rotational speed range is divided into 100 rpm intervals by the boundaries shown in Lv.0 to Lv.5. Points P1 to P14 that pass through the boundaries are shown on waveform W41.

[0052] Waveform W42 shows the number of times the rotation speed transitions from one divided region to an adjacent region during a certain period Δt. This number indicates how many points P1 to P14 are included during that period Δt. For example, as shown in Figure 11, if the period Δt is t42 to t43, then the period Δt includes two points, P4 and P5. Therefore, at time t42, the waveform W42 shows a level change count N=2. If the period Δt is shifted slightly backward, the period Δt will include three points, P4, P5, and P6. Therefore, the waveform W42 changes from a level change count N=2 to N=3 a little after time t42. Waveform W42 shows the results of continuously performing the above counting process from time t41 to t45.

[0053] Figure 12 shows how a threshold is set for waveform W42, and when the threshold is exceeded, it is determined that the rotational speed fluctuation exceeds the standard during the period Δt from the time of exceeding the threshold. At time t42s, if waveform W42(t42s) exceeds the threshold, data for a certain period Δt from time t42s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing. Similarly, at time t44s, if waveform W42(t44s) exceeds the threshold, data for a certain period Δt from time t44s is selected as the rotational fluctuation period, and the selected data is used for subsequent processing.

[0054] To simplify post-processing, data may be selected only from the point when waveform W42 exceeds the threshold, up to Δt, to reduce overlap in selected time periods. However, if overlap in selected time periods is acceptable, data may be selected from the point when waveform W42 exceeds the threshold, up to Δt.

[0055] Furthermore, while the number of transitions from one region to another within a certain period of time was used as the unit for determination, the number of regions traversed may also be used as the unit for determination. Also, while a certain period of time Δt was used as the unit for the determination interval, a certain number of rotations ΔN may be used instead of a certain period of time Δt as the unit for the determination interval. When counting the number of times moving between regions, an unconventional counting method may be adopted, such as not counting movements to adjacent regions. When any of the above determination targets exceed a predetermined threshold, it can be determined that the change in rotational speed exceeds the standard.

[0056] [6. Judgment based on the filtered value of the rotation speed] Figure 13 is a waveform diagram illustrating the sixth method for determining rotational fluctuations. Figure 14 is a waveform diagram showing the determination results using the sixth method for determining rotational fluctuations.

[0057] In Figure 13, waveform W51 shows the change in rotational speed (rpm) of the rotating part observed by the sensor. Waveform W52 shows the result of applying a high-pass filter to the rotational speed signal.

[0058] Waveform W53 shows the value obtained by calculating the root mean square (RMS) of the value of waveform W52 over a certain period Δt. For example, the value of waveform W53 at t52 is the RMS of the value of waveform W52 over a certain period Δt from time t52 to t53. Waveform W53 shows the results of performing the above calculation continuously from time t51 to t55.

[0059] Figure 14 shows how a threshold is set for waveform W53, and when the threshold is exceeded, it is determined that the rotational speed fluctuation exceeds a standard during the period Δt from that time. At time t52s, if waveform W53(t52s) exceeds the threshold, data for a certain period Δt from time t52s is selected as the rotational fluctuation period and used in subsequent processing. Similarly, at time t54s, if waveform W53(t54s) exceeds the threshold, data for a certain period Δt from time t54s is selected as the rotational fluctuation period and used in subsequent processing.

[0060] To simplify post-processing, data may be selected only from the point when waveform W53 exceeds the threshold, up to Δt, to reduce overlap in selected time periods. However, if overlap in selected time periods is acceptable, data may be selected from the point when waveform W53 exceeds the threshold, up to Δt.

[0061] Furthermore, while a fixed time interval Δt was used as the unit of the judgment interval, a fixed rotational speed ΔN may be used instead of a fixed time interval Δt. Also, while the RMS value was calculated from the value after high-pass filtering, it may be calculated from the value after other filtering processes such as band-pass filtering. Filtering may be performed using digital filters such as FIR or IIR, or analog filters such as switched-capacitor or RC filters. In addition, while the RMS value of the rotational speed after filtering in the judgment interval was used as the judgment target, a value that evaluates the magnitude of the fluctuating value, such as the sum of squares, may be used as the judgment target instead of the RMS value. When any of the above judgment targets exceed a preset threshold, it can be determined that the fluctuation in rotational speed exceeds the standard.

[0062] Furthermore, if we define the data length (such as frequency analysis) used in step S4 of Figure 2 as the "window length," then in the filtering process, it is desirable to attenuate frequency components below the reciprocal of eight times the window length, or at least avoid emphasizing them. This is because extremely low-frequency components interfere with this judgment method. For example, if RMS is used as the evaluation value, a large amount of low-frequency components will cause the RMS to constantly exceed the threshold, making judgment impossible.

[0063] [How to select data for the rotational fluctuation period] The above explains how to determine whether the rotational speed fluctuation exceeds the standard. Now, let's explain how to select the data determined using this method.

[0064] [Selection Method 1] The following describes a method for processing continuous or sufficiently long measurement data in short data intervals relative to the data length to be analyzed. Note that Selection Method 1 includes so-called real-time measurements. Furthermore, Selection Method 1 includes cases where the data is separated but there is no dead time between those intervals.

[0065] A portion of the measurement data is extracted, and it is determined whether the data in that extracted portion is exhibiting rotational fluctuations. For the determination method, any one of the six methods described in Figures 3 to 14 can be used. The data extraction interval is sequentially slid over time, and the measurement data from the interval determined to exhibit rotational fluctuations is extracted as data for analysis (data selection in S2).

[0066] Furthermore, an upper limit, lower limit, or both of the data length to be extracted may be determined based on time or rotation speed. If an upper limit is set, data exceeding the upper limit will be discarded or separated into other data. If a lower limit is set, data below the lower limit will be discarded.

[0067] Furthermore, some of the extracted data may include data where the rotational speed fluctuation is determined not to exceed a certain threshold. For example, a section where the rotational speed fluctuation is determined not to exceed a certain threshold may be included in the section where the rotational speed fluctuation is determined to exceed a certain threshold. Also, sections where the rotational speed fluctuation is determined not to exceed a certain threshold may be included before and after the section where the rotational speed fluctuation is determined to exceed a certain threshold.

[0068] [Selection Method 2] For all or part of a pre-defined set of measurement data, one of the six methods described in Figures 3 to 14 is used to determine whether the rotational speed fluctuation exceeds a certain threshold. If it is determined that the rotational speed fluctuation exceeds the threshold, all or part of that divided data is extracted as data for analysis (data selection in S2).

[0069] [Derivations of the two selection methods above] The data extracted for analysis using selection method 1 or selection method 2 may be ranked according to the values ​​used for evaluation, and data within a certain range of ranks may be ultimately extracted (selected). Furthermore, weighting may be applied to the ranking based on data length or data volume. Additionally, weighting may be applied to the ranking based on the evaluation values ​​of data that are close in time.

[0070] Once data has been extracted that has been determined to exceed a certain threshold for rotational speed fluctuation, it is acceptable to refrain from extracting further data for a certain period of time. For example, in selection method 1, if the period is set to be the same as or longer than Δt as shown in Figure 3, it is possible to avoid selecting data at overlapping time points.

[0071] This embodiment explains the effects of selecting data for the rotational fluctuation period using the condition monitoring device.

[0072] Figure 15 is a graph showing the spectrum when frequency analysis is performed without partial data selection. Figure 16 is a graph showing the spectrum when frequency analysis is performed by selecting data from the rotational fluctuation period.

[0073] In Figure 15, six peaks are observed in the vibration of the rotating part: X1-X3 and Y1-Y3. Here, X1-X3 are rotation-synchronous components and represent peaks that may be characteristic frequency peaks indicating damage. On the other hand, Y1-Y3 are non-rotation-synchronous components and represent peaks that do not have the potential to be characteristic frequency peaks indicating damage.

[0074] In Figure 16, as a result of resampling that takes rotational speed into account, the rotation-synchronous components, peaks X1-X3, become sharper, while the non-rotation-synchronous components, Y1-Y3, become less sharp. Therefore, it becomes easier to separate the rotation-synchronous components from other components (such as power supply noise), facilitating early detection of damage to rotating parts.

[0075] As described above, in this embodiment, data during the rotational fluctuation period is selectively acquired. When rotational tracking processing is applied to the measurement data during the rotational fluctuation period, in frequency analysis, peaks originating from parts other than the rotating part are dulled, while peaks originating from the rotating part are emphasized. This makes it possible to distinguish between vibrations from the rotating part and other vibrations, thereby improving diagnostic accuracy. By specifically measuring data during the rotational fluctuation period, the effect of improving diagnostic accuracy through rotational tracking processing can be maximized.

[0076] [Note] Those skilled in the art will understand that the exemplary embodiments described above are specific examples of the following embodiments.

[0077] (1) This disclosure relates to a condition monitoring device 20 for monitoring the state of a device 10 having a rotating part. The condition monitoring device 20 shown in Figure 1 includes a receiving unit 21 that receives sensor signals from a sensor 12 that monitors the state of the device 10 and speed signals indicating the rotational speed of the rotating part, a storage device 24 that stores measurement data including sensor information and speed information corresponding to the sensor signals and speed signals received by the receiving unit 21, respectively, and a computing device (CPU) 23 that analyzes the measurement data. The computing device 23 is configured to extract data from the measurement data that includes a section in which the fluctuation of the rotational speed exceeds a standard, and to perform tracking processing that analyzes the sensor information in the extracted data in accordance with the change in rotational speed.

[0078] (2) In the condition monitoring device 20 described in (1), the calculation device 23 determines whether the fluctuation in rotational speed exceeds a standard over a certain period of time or over an elapsed period of time, as shown in Figures 3 and 4, by using the difference between the maximum and minimum values ​​obtained from the rotational speed over a certain period of time or over an elapsed period of time while the rotating part rotates for a certain number of rotations.

[0079] (3) In the condition monitoring device 20 described in (1), the calculation device 23 determines whether the fluctuation in rotational speed exceeds a standard over a certain period of time or over an elapsed period of time, as shown in Figures 5 and 6, by using an evaluation value (such as a standard deviation) that shows the variation in the value obtained from the rotational speed over a certain period of time or over an elapsed period of time while the rotating part rotates for a certain number of rotations.

[0080] (4) In the state monitoring device 20 described in (1), the calculation device 23 determines whether the fluctuation in rotational speed over a certain period of time or over an elapsed period of time exceeds a standard, based on whether the value obtained from the rotational speed over a certain period of time or over an elapsed period of time during which the rotating part rotates for a certain number of rotations includes a value that is below a predetermined lower limit and also includes a value that is above a predetermined upper limit.

[0081] (5) In the condition monitoring device 20 described in (1), the calculation device 23 determines whether the fluctuation in rotational speed exceeds a standard over a certain period of time or over an elapsed period of time, using the integral value obtained from the differential value of the rotational speed over a certain period of time or over an elapsed period of time while the rotating part rotates for a certain number of rotations, as shown in Figures 9 and 10.

[0082] (6) In the condition monitoring device 20 described in (1), the range of rotational speeds that the speed information can indicate is divided in advance into a plurality of rotational speed regions, and the calculation device 23 is configured to determine which of the plurality of rotational speed regions the rotational speed of the rotating part belongs to, as shown in Figures 11 and 12, and the calculation device 23 determines whether the fluctuation of the rotational speed exceeds a standard during a certain period of time or during the elapsed time while the rotating part rotates for a certain number of rotations, using the number of times the rotational speed transitions between the plurality of rotational speed regions or the number of regions the rotational speed passes through among the plurality of rotational speed regions.

[0083] (7) In the state monitoring device 20 described in (1), the calculation device 23 determines whether the fluctuation in rotational speed exceeds a standard over a certain period of time or over an elapsed period of time, as shown in Figures 13 and 14, by using an evaluation value which is obtained by evaluating the magnitude of the value after filtering the rotational speed over a certain period of time or over an elapsed period of time during which the rotating part rotates for a certain number of rotations. The filtering process includes a process that emphasizes or attenuates some frequency components.

[0084] (8) In other aspects, this disclosure relates to a condition monitoring method for monitoring the condition of a device 10 having a rotating part. The condition monitoring method includes the steps of: receiving a sensor signal from a sensor 12 that monitors the condition of the device 10 and a speed signal indicating the rotational speed of the rotating part; storing measurement data including sensor information and speed information corresponding to the received sensor signal and speed signal, respectively (S1); extracting data from the measurement data that includes a section in which the fluctuation of the rotational speed exceeds a standard (S2); and performing a tracking process (S3) that analyzes the sensor information in the extracted data in accordance with the change in rotational speed.

[0085] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the present invention is indicated by the claims rather than by the description of the embodiments above, and all modifications within the meaning and scope equivalent to the claims are intended to be included. [Explanation of Symbols]

[0086] 10. Device to be monitored, 11. Control device, 12. Sensor, 20. Status monitoring device, 21. Receiving unit, 22. Interface unit, 23. Processing unit, 24. Storage device, 25. Transmitter, 30. Display / server.

Claims

1. A condition monitoring device that monitors the state of a device to be monitored that has a rotating part, A receiving unit that receives a sensor signal from a sensor that monitors the status of the device to be monitored and a speed signal indicating the rotational speed of the rotating part, A storage device that stores measurement data including sensor information and speed information corresponding to the sensor signal and speed signal received by the receiving unit, The system includes a computing device that analyzes the aforementioned measurement data, The calculation unit is configured to extract data from the measurement data that includes a section in which the rotational speed fluctuation exceeds a standard, and to perform tracking processing that analyzes the sensor information in the extracted data in accordance with the change in rotational speed, thereby providing a condition monitoring device.

2. The state monitoring device according to claim 1, wherein the calculation device determines whether the fluctuation in rotational speed exceeds a standard during the elapsed time while the rotating part rotates for a certain period of time or for a certain number of rotations, by using the difference between the maximum and minimum values ​​obtained from the rotational speed during the elapsed time while the rotating part rotates for a certain period of time or for a certain number of rotations.

3. The state monitoring device according to claim 1, wherein the calculation device determines whether the fluctuation in rotational speed exceeds a standard during the elapsed time while the rotating part rotates for a certain period of time or a certain number of rotations, using an evaluation value that shows the variation in the value obtained from the rotational speed during the elapsed time while the rotating part rotates for a certain period of time or a certain number of rotations.

4. The state monitoring device according to claim 1, wherein the calculation device determines whether the fluctuation in rotation speed during the certain period of time or the elapsed time exceeds a standard, based on whether the value obtained from the rotation speed during the elapsed time while the rotating part rotates for a certain number of rotations includes a value below a predetermined lower limit and a value above a predetermined upper limit.

5. The state monitoring device according to claim 1, wherein the calculation device determines whether the fluctuation in rotational speed exceeds a standard during the elapsed time while the rotating part rotates for a certain period of time or a certain number of rotations, by using the integrated value of the derivative of the rotational speed during the elapsed time while the rotating part rotates for a certain period of time or a certain number of rotations.

6. The range of rotational speeds that the aforementioned speed information can indicate is divided in advance into multiple rotational speed regions. The calculation device is configured to determine which of the plurality of rotation speed ranges the rotation speed of the rotating part belongs to. The state monitoring device according to claim 1, wherein the calculation device determines whether the fluctuation of the rotation speed exceeds a standard during the elapsed time, or during the time during which the rotating part rotates for a certain period of time or for a certain number of rotational speeds, using the number of times the rotation speed transitions between the plurality of rotational speed regions or the number of regions the rotation speed passes through among the plurality of rotational speed regions.

7. The calculation device uses an evaluation value, which is obtained by evaluating the magnitude of the value after filtering the rotational speed during a certain period of time or the elapsed time during which the rotating part rotates for a certain number of rotations, to determine whether the fluctuation in rotational speed during that certain period of time or the elapsed time exceeds a standard. The state monitoring device according to claim 1, wherein the filtering process includes a process to enhance or attenuate some frequency components.

8. A condition monitoring method for monitoring the state of a device having a rotating part, The steps include receiving a sensor signal from a sensor that monitors the status of the device to be monitored and a speed signal indicating the rotational speed of the rotating part, A step of storing measurement data including sensor information and speed information corresponding to the received sensor signal and speed signal, respectively. The steps include: extracting data from the aforementioned measurement data that includes a section in which the rotational speed fluctuation exceeds a standard; A condition monitoring method comprising the step of performing a tracking process that analyzes the sensor information in the extracted data in accordance with changes in rotational speed.

Images