Rotating device abnormality detection device, rotating device abnormality detection method, and rotating device abnormality detection program

Abstract

To solve the problem that a device, a method and a program for detecting the abnormality of a rotary rotation machine are required, which, if the inspection and maintenance of a rotary rotation machine, and the measures to detect damage of equipment are insufficient, may lead the rotary rotation machine to the deterioration, suddenly become inoperable, can be achieved by a simple mechanism and do not require advanced technique or experience in the data analysis.SOLUTION: Provided are a device, a method and a program for detecting the abnormality of a rotary rotation machine, which, by the correlation between the shaft rotation speed and vibration acceleration (effective value) or an index which determines an operational state of the rotary rotation machine obtained from the correlation between the shaft rotation speed and the vibration acceleration (effective value), can be achieved by a simple mechanism and do not require advanced technique or experience in the data analysis.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an abnormality detection device for a rotating device, an abnormality detection method for a rotating device, and an abnormality detection program for a rotating device. [Background technology]

[0002] Rotating equipment undergoes fatigue, wear, corrosion, cracks, and material deterioration over time, leading to damage. While such equipment damage occurs due to aging even under conditions assumed in the equipment's design, improper operation, such as overloading, and poor maintenance, such as poor installation foundations and poor or insufficient lubricating oil, can accelerate the deterioration that leads to equipment damage. Abnormalities in rotating equipment can lead to reduced functionality and breakdowns, resulting in reduced product quality, production stoppages, or accidents, so it is important to detect abnormalities early and take appropriate action.

[0003] A representative technology for detecting abnormalities in rotating equipment is vibration measurement and analysis technology. Generally, it is possible to measure vibrations in a frequency range of several to 20 kHz and identify the maintenance status and damage state from the measurement data. For example, technologies shown in Patent Document 1 and Patent Document 2 have been proposed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-101009 [Patent Document 2] Japanese Patent Application Publication No. 2018-36124 [Non-patent literature]

[0005] [Non-Patent Document 1] The Society of Plant Engineers of Japan, "Proceedings of the 2021 Spring Research Conference" (published June 3, 2021), pp. 38-41 [Non-patent document 2] The Society of Plant Engineers of Japan, "2021 Autumn Research Conference Proceedings" (published November 10, 2021), pp. 67-68 Summary of the Invention [Problem to be solved by the invention]

[0006] In large plants and the like, abnormalities in rotating equipment not only have an economic impact on the facility, but can also have a social impact and lead to serious accidents, so it is necessary to manage the operating conditions of rotating equipment within appropriate ranges and maintain it in good condition. For example, Patent Document 1 proposes a diagnostic method that can accurately identify the types of abnormalities in rotating equipment. Furthermore, Patent Document 2 proposes a condition monitoring device that can accurately analyze data on the state of rotating equipment obtained from sensors.

[0007] On the other hand, in so-called general industries with small and medium-sized manufacturing facilities, an abnormality in a rotating machine only has an economic impact on the facility, so inspection, maintenance, and management of the rotating machine at the level of a large plant, etc., would be excessive. In other words, adopting the diagnostic method shown in Patent Document 1 would be excessive for general industries, leading to increased management costs.

[0008] Furthermore, the manufacturing industry is facing a serious labor shortage, and general industries are facing a shortage of skilled workers at manufacturing sites. For this reason, even if general industries introduce condition monitoring devices capable of accurately analyzing data on the state of rotating equipment obtained from sensors, such as those shown in Patent Document 2, advanced skills and experience are required to interpret the analysis data, which is often not suited to the current situation of a shortage of skilled workers.

[0009] However, even in such general industries, insufficient inspection and maintenance or measures to detect equipment damage can lead to deterioration of rotating equipment, causing it to suddenly become inoperable. For this reason, general industries need a rotating equipment anomaly detection device, a rotating equipment anomaly detection method, and a rotating equipment anomaly detection program that can be implemented with a simple system and do not require advanced technology or experience for data analysis.

[0010] The inventors of the present invention have conducted vibration measurement and analysis of rotating equipment using a vibration sensor and found that the correlation between shaft rotation speed and vibration acceleration (effective value), or the correlation between shaft rotation speed and vibration velocity (effective value), can be used as an index to determine the operating state of rotating equipment. For details, please refer to Non-Patent Document 1 and Non-Patent Document 2.

[0011] The present invention aims to provide an abnormality detection device, an abnormality detection method, and an abnormality detection program for rotating equipment that can be realized with a simple mechanism using an index for determining the operating state of rotating equipment obtained from the correlation between shaft rotation speed and vibration acceleration (effective value) or the correlation between shaft rotation speed and vibration velocity (effective value), and that does not require advanced technology or experience in data analysis. [Means for solving the problem]

[0012] An abnormality detection device for a rotating device according to one aspect of the present invention comprises an input unit that receives and accumulates measurement data of vibration during operation of the rotating device that is the target of abnormality detection from an external device, a calculation unit that calculates an index value indicating the measured operating state of the rotating device from the data received and accumulated by the input unit, a determination unit that determines the operating state of the rotating device from the index value indicating the measured operating state of the rotating device calculated by the calculation unit, and an output unit that generates and outputs output information from the operating state determined by the determination unit and the index value indicating the operating state of the rotating device. Here, the measurement data of vibration during operation refers to measurement data containing information that determines the vibration during operation of the rotating device that is the target of abnormality detection. The same applies throughout the present application.

[0013] One aspect of the present invention is an abnormality detection device for rotating equipment, characterized in that the measurement data received and accumulated by the input unit from an external source is low-frequency vibration (approximately 100 Hz or less) during operation. In the method disclosed as the present invention, the correlation between shaft rotation speed and vibration acceleration (effective value) or the correlation between shaft rotation speed and vibration velocity (effective value) for obtaining an index for determining the operating state of the rotating equipment does not differ depending on the frequency range of the vibration measured. However, by limiting the frequency range in the abnormality detection device for rotating equipment to the low-frequency range, it is possible to obtain effects such as inexpensive vibration sensors and fewer computer resources required for analyzing the measurement data. Furthermore, measurement errors are small even when an unskilled person installs the sensor.

[0014] An abnormality detection device for rotating equipment according to one aspect of the present invention further includes an analysis unit that analyzes the value of an index that indicates the measured operating state of the rotating equipment calculated by the calculation unit, and generates and stores a judgment index for judging the operating state, and the judgment unit judges the operating state by evaluating the value of the index that indicates the measured operating state of the rotating equipment calculated by the calculation unit using the judgment index generated and stored by the analysis unit.

[0015] A rotating equipment abnormality detection device according to one aspect of the present invention is characterized in that an analysis unit analyzes the value of an index indicating the operating state of the rotating equipment by cluster analysis, and generates and stores a judgment index.

[0016] One aspect of the present invention is an abnormality detection device for rotating equipment, characterized in that the judgment unit judges the operating state by inputting the value of an index indicating the measured operating state of the rotating equipment calculated by the calculation unit into a trained learning model that has undergone machine learning to judge the operating state.

[0017] The abnormality detection device for a rotating device according to one aspect of the present invention is characterized in that the indicators indicating the operating state of the rotating device are the rotation speed of a rotating shaft of the rotating device and the effective value of vibration acceleration.

[0018] The rotating equipment abnormality detection device according to one aspect of the present invention is characterized in that the indicators indicating the operating state of the rotating equipment are the rotation speed and the effective value of the vibration velocity of the rotating shaft of the rotating equipment.

[0019] In one aspect of the present invention, an abnormality detection device for a rotating machine is characterized in that an analysis unit analyzes the value of an index indicating the operating state of the rotating machine by spectral analysis, and generates and stores a judgment index.

[0020] One aspect of the present invention is an abnormality detection device for rotating equipment, characterized in that the judgment unit judges the operating state by inputting the spectral pattern of the index value indicating the operating state of the rotating equipment into a trained learning model that has undergone machine learning to judge the operating state.

[0021] One aspect of the present invention is an abnormality detection device for rotating equipment, characterized in that when the judgment unit judges that the operating state of the rotating equipment is an unsafe operating state, it passes on possible countermeasures to the output unit, and the output information generated and output by the output unit includes the countermeasures.

[0022] One aspect of the present invention is a method for detecting an abnormality in a rotating device, which comprises an input step of receiving from the outside and accumulating measurement data of vibrations during operation of the rotating device that is the target of abnormality detection; a calculation step of calculating an index value indicating the operating state of the rotating device measured from the data received and accumulated in the input step; a determination step of determining the operating state of the rotating device from the index value indicating the operating state of the rotating device calculated in the calculation step; and an output step of generating and outputting output information from the operating state determined in the determination step and the index value indicating the operating state of the rotating device.

[0023] One aspect of the present invention is a rotating equipment abnormality detection program that causes a computer to execute the following steps: an input step for receiving and storing measurement data of vibrations during operation of a rotating equipment that is the target of abnormality detection from the outside; a calculation step for calculating an index value indicating the operating state of the rotating equipment measured from the data received and stored in the input step; a determination step for determining the operating state of the rotating equipment from the index value indicating the operating state of the rotating equipment calculated in the calculation step; and an output step for generating and outputting output information from the operating state determined in the determination step and the index value indicating the operating state of the rotating equipment. [Effects of the Invention]

[0024] The rotating equipment abnormality detection device, rotating equipment abnormality detection method, and rotating equipment abnormality detection program according to the present invention can detect abnormalities in rotating equipment using a simple mechanism, without the need for advanced technology or experience in data analysis, by using an index that determines the operating state of the rotating equipment, which is obtained from the correlation between the shaft rotation speed and vibration acceleration (effective value) or the correlation between the shaft rotation speed and vibration velocity (effective value). [Brief explanation of the drawings]

[0025] [Figure 1] FIG. 1 is a block diagram of a rotating machine abnormality detection device. [Figure 2] FIG. 2 is a hardware configuration diagram of a rotating equipment abnormality detection device. [Figure 3] 4 is a flowchart showing the operation of the rotating machine abnormality detection device. [Figure 4] This is a diagram of vibration measurement and analysis of rotating equipment using a vibration sensor. [Figure 5] This is a diagram of vibration measurement and analysis of rotating equipment using a vibration sensor. [Figure 6] FIG. 10 is a diagram illustrating a first analysis process. [Figure 7] FIG. 10 is a diagram illustrating a second analysis process. [Figure 8] 10 is a flowchart illustrating an output process related to a countermeasure output. DETAILED DESCRIPTION OF THE INVENTION

[0026] An embodiment of the present invention will be described with reference to the drawings. Duplicate descriptions will be omitted, and the same or corresponding parts in each drawing will be denoted by the same reference numerals.

[0027] In the embodiments of the present invention, the rotating equipment abnormality detection device, rotating equipment abnormality detection method, and rotating equipment abnormality detection program can detect abnormalities in rotating equipment using a simple mechanism and without advanced skills or experience in data analysis, by using an index that determines the operating state of the rotating equipment obtained from the correlation between the shaft rotation speed and vibration acceleration (effective value) or the correlation between the shaft rotation speed and vibration velocity (effective value).

[0028] In the embodiments of the present invention, in consideration of ease of explanation and understanding, an example will be described in which a vibration sensor installed in a rotating device that is the target of abnormality detection measures vibration acceleration, and the vibration acceleration measured at a fixed time period, for a fixed time unit (for example, every hour for 10 seconds, in 1 millisecond units) is input as measurement data of vibration during operation. Note that this is an example in consideration of ease of explanation and understanding, and the present invention is not limited to this. Any measurement data containing information that determines the vibration during operation of the rotating device that is the target of abnormality detection may be used, and for example, the amplitude, frequency, and phase of the vibration of the rotating device may be directly measured and input as measurement data of vibration during operation. [Example]

[0029] The rotating equipment abnormality detection device, rotating equipment abnormality detection method, and rotating equipment abnormality detection program disclosed as a first embodiment of the present invention analyze the value of an index indicating the operating state of the rotating equipment by cluster analysis, generate and evaluate the stored judgment index, and thereby judge the operating state. In the first embodiment, the explanation will be given assuming that the index for judging the operating state of the rotating equipment is obtained from the correlation between the shaft rotation speed and the vibration acceleration (effective value).

[0030] As an embodiment of the present invention, there is an embodiment in which an index for determining the operating state of a rotating device is calculated from the correlation between the shaft rotation speed and the vibration velocity (effective value), but the difference from Example 1 disclosed below is the vibration acceleration (effective value) and the vibration velocity (effective value), and in other respects, the configurations of the rotating device abnormality detection device, the rotating device abnormality detection method, and the rotating device abnormality detection program are the same. Therefore, a description of the embodiment in which an index for determining the operating state of a rotating device is calculated from the correlation between the shaft rotation speed and the vibration velocity (effective value) will be omitted.

[0031] First, the rotating equipment abnormality detection device will be described. FIG. 1 is a block diagram of the rotating equipment abnormality detection device. The rotating equipment abnormality detection device 1 may be configured as a standalone device, or may be incorporated into another device for use. The other device incorporating the calculation device 1 may be, for example, a rotating equipment itself, or an electrical appliance such as a personal computer, a smartphone, or a personal digital assistant.

[0032] Fig. 2 is a hardware configuration diagram of the rotating equipment abnormality detection device 1. As shown in Fig. 2, physically, it is configured as a computer including a central processing unit (CPU) 201, an input device 202, an output device 203, a main memory device (RAM / ROM) 204, and an auxiliary memory device 205.

[0033] Each function of the rotating equipment abnormality detection device 1 is realized by loading a rotating equipment abnormality detection program into a central processing unit (CPU) 201, a main memory device (RAM / ROM) 204, etc. shown in Figure 2, thereby operating an input device 202 and an output device 203 under the control of the central processing unit (CPU) 201, and reading and writing data from and to the main memory device (RAM / ROM) 204 and auxiliary memory device 205.

[0034] 1, the rotating equipment abnormality detection device 1 includes an input unit 101, a calculation unit 102, an analysis unit 103, a determination unit 104, and an output unit 105. In this embodiment, the rotating equipment abnormality detection device 1 receives input from the outside as measurement data of vibrations during operation of the rotating equipment via the input unit 101. The rotating equipment abnormality detection device 1 also outputs information to the outside via the output unit 105. The output information will be described later in the description of the output unit 105.

[0035] The function of each block of the rotating equipment abnormality detection device 1 will be described with reference to the block diagram of Figure 1. The detailed operation of each block will be described later.

[0036] The input unit 101 receives and stores measurement data of vibrations during operation of a rotating device that is the target of abnormality detection from the outside.

[0037] The calculation unit 102 calculates the value of an index that indicates the operating state of the rotating equipment measured from the data received and accumulated by the input unit 101.

[0038] The analysis unit 103 analyzes the value of the index that indicates the measured operating state of the rotating equipment calculated by the calculation unit 102, and generates and stores a determination index for determining the operating state.

[0039] The determination unit 104 determines the operating state by evaluating the value of the index that indicates the measured operating state of the rotating equipment, calculated by the calculation unit 102, using the determination index that the analysis unit 103 generates and holds.

[0040] The output unit 105 generates and outputs output information from the operating state determined by the determination unit 104 and the index value indicating the operating state of the rotating equipment.

[0041] Next, the operation of the rotating device abnormality detection device 1 according to this embodiment will be described. Fig. 3 is a flowchart showing the operation of the rotating device abnormality detection device 1 according to this embodiment. The operation of the rotating device abnormality detection device 1 according to this embodiment will be described according to the flowchart of Fig. 3.

[0042] In this embodiment, vibration acceleration measured for a rotating device that is the target of abnormality detection processing is input from the outside as measurement data of vibration during operation to the rotating device abnormality detection device 1. The rotating device abnormality detection device 1 starts operating after the data to be processed is input. Operation may start automatically after the information is input, or may start by an explicit command. When the rotating device abnormality detection device 1 starts operating, it performs the processing of the flowchart in Figure 3.

[0043] When the rotating equipment abnormality detection device 1 starts operating, the input unit 101 executes input processing (S301). In this embodiment, in the input processing (S301), vibration acceleration measured for the rotating equipment that is the target of abnormality detection is received from the outside as measurement data of vibration during operation and accumulated. The data received from the outside is also used in the analysis processing (S303) described below to generate a judgment index for judging the operating state. The judgment index is unique for each rotating equipment that is the target of abnormality detection and is generated from vibration data from past operations of the rotating equipment that is the target of abnormality detection. Therefore, in the input processing (S301), data is received from the outside and accumulated to generate a judgment index.

[0044] In this embodiment, the input processing (S301) may be, for example, sequential processing in which data is sequentially received and processed from a vibration sensor outside the device, or may be batch processing in which measurement data for a certain period of time is collectively received and processed.

[0045] When the input process (S301) is completed, the calculation unit 102 starts the calculation process (S302). In the calculation process (S302), the value of an index indicating the operating state of the rotating equipment measured from the data received and accumulated in the input process (S301) is calculated. In this embodiment, as described above, the index for determining the operating state of the rotating equipment is described as being obtained from the correlation between the shaft rotation speed and the vibration acceleration (effective value). Note that the effective value here refers to the square root of the mean square value of each instantaneous value within a certain period of time of the time-axis waveform, and the same applies throughout the present application.

[0046] In this embodiment, in the calculation process (S302), shaft rotation speed and vibration acceleration (effective value) are calculated as index values ​​indicating the operating state of the rotating equipment from the data received and accumulated in the input process (S301). In this embodiment, the shaft rotation speed is calculated from the power spectrum of each order of the frequency of the rotating equipment obtained by fast Fourier transform from the data received and accumulated in the input process (S301). Note that this is an example, and the method is not limited to this as long as the shaft rotation speed can be calculated. Calculations can also be made using other methods from the data received and accumulated in the input process (S301). Furthermore, values ​​can also be obtained by indirectly or directly measuring the shaft rotation speed and vibration acceleration (effective value). For example, the shaft rotation speed can be measured indirectly from the motor current of the rotating equipment, or directly using a tachometer.

[0047] When the calculation process (S302) is completed, the analysis unit 103 starts the analysis process (S303). The analysis process (S303) analyzes the values ​​of the indices indicating the measured operating state of the rotating equipment calculated by the calculation process (S302), and generates and stores a judgment index for determining the operating state. In this embodiment, the values ​​of the indices indicating the operating state of the rotating equipment are the shaft rotation speed and the vibration acceleration (effective value). In this embodiment, the analysis process (S303) analyzes the values ​​of the indices indicating the operating state of the rotating equipment by cluster analysis, and generates and stores a judgment index.

[0048] Prior to describing the analysis process (S303), we will outline the gist of Non-Patent Documents 1 and 2, which form the basis for the analysis process (S303). Figure 4 is a diagram of vibration measurement and analysis of a rotating device using a vibration sensor, conducted by the inventor of the present invention. Figure 4 plots shaft rotation speed and vibration acceleration (effective value) on a two-dimensional Cartesian coordinate plane, with the horizontal axis representing vibration acceleration (effective value) and the vertical axis representing shaft rotation speed. Note that in Figure 4 and Figure 5 (described below), the vertical axis represents double rotation speed, which is the most characteristic value in rotation speed measurements using spectrum analysis during experiments, and is therefore easy to explain. While the correlation with vibration acceleration (effective value) varies depending on the specific characteristics of the rotating device, shaft rotation speed remains an indicator for determining the operating state of the rotating device. The same is true for the correlation between vibration velocity (effective value) and shaft rotation speed.

[0049] The inventors of the present invention conducted vibration measurement and analysis of rotating equipment using a vibration sensor and found that the correlation between shaft rotation speed and vibration acceleration (effective value) can be used as an indicator for determining the operating state of the rotating equipment. In rotating equipment, as the load increases, the shaft rotation speed decreases. Conversely, as the load increases, the vibration acceleration of the rotating equipment increases. Therefore, a negative correlation can be found between shaft rotation speed and vibration acceleration (effective value). However, from the measurement results shown in Figure 4, they discovered the emergence of a data set in which the negative correlation between shaft rotation speed and vibration acceleration (effective value) disappears in the low shaft rotation speed range. This data set is plotted as "red operation" in Figure 4 with a circle. Furthermore, they found that rotating equipment whose shaft rotation speed and vibration acceleration (effective value) were included in this data set had either inappropriate operating conditions, inappropriate maintenance, or an abnormality in the rotating equipment itself, and that continued operation was in a dangerous operating state (hereinafter referred to as a "hazardous operating state"). Based on this, the analysis process (S303) analyzes the values ​​of the indices indicating the operating state of the rotating equipment (in this embodiment, the shaft rotation speed and vibration acceleration (effective value)), and generates and stores an index for determining a data set that represents a dangerous operating state.

[0050] FIG. 5 is a diagram showing vibration measurement and analysis of a rotating device using a vibration sensor, conducted by the inventor of the present invention. FIG. 5 plots shaft rotation speed and vibration velocity (effective value) on a two-dimensional Cartesian coordinate plane with vibration velocity (effective value) on the horizontal axis and shaft rotation speed on the vertical axis. Comparing FIG. 5 with FIG. 4, as shown in FIG. 5, a data set indicating a dangerous driving state is detected in the lower right (toward the fourth quadrant) between the shaft rotation speed and vibration velocity (effective value), as is the case between the shaft rotation speed and vibration acceleration (effective value) shown in FIG. 4. In FIG. 5, this data set is plotted as a red driving state with a circle. Thus, a similar tendency is observed in detecting a data set indicating a dangerous driving state between the shaft rotation speed and vibration acceleration (effective value) and between the shaft rotation speed and vibration velocity (effective value). Therefore, as described above, in the description of Example 1, a description of an embodiment in which an index for determining the operating state of a rotating device is obtained from the correlation between the shaft rotation speed and vibration velocity (effective value) will be omitted.

[0051] As described above, the analysis process (S303) analyzes the values ​​of the index indicating the operating state of the rotating equipment by cluster analysis, and generates and stores a judgment index. Any cluster analysis that can appropriately analyze the values ​​of the index indicating the operating state of the rotating equipment may be used, for example, an algorithm such as k-means with the number of clusters set to 2 may be used. This embodiment discloses an example of a cluster analysis using straight lines that can appropriately analyze the values ​​of the index indicating the operating state of the rotating equipment and also reduce the computer resources required for the analysis. However, the present invention is not limited to this example.

[0052] In this embodiment, the generation of a judgment index for judging the driving state performed by the analysis process (S303) will be described with reference to Figures 6 and 7. Regarding the analysis process (S303), a first analysis process and a second analysis process will be described, but these are merely examples, and the generation of a judgment index for judging the driving state is not limited to these processes. Furthermore, in the judgment process (S304) described later, the judgment may be based on either the judgment index from the first analysis process or the judgment index from the second analysis process, or may be based on both.

[0053] Fig. 6 is a diagram illustrating the first analysis process. Fig. 6 shows the shaft rotation speed and vibration acceleration (effective value), which are index values ​​indicating the operating state of the rotating equipment measured by the calculation process (S302), on a two-dimensional orthogonal coordinate plane with the vibration acceleration (effective value) on the horizontal axis and the shaft rotation speed on the vertical axis.

[0054] The data set of the dangerous driving state appears in the lower right (toward the fourth quadrant) in Figure 6. In the first analysis process, the area on the two-dimensional Cartesian coordinate plane where the data set of the dangerous driving state appears (hereinafter referred to as the dangerous area) is designated as a cluster of the dangerous area by two lines, one parallel to the horizontal axis and the other parallel to the vertical axis. In the first analysis process, the analysis process (S303) generates and stores the equations of the two lines that designate the cluster of the dangerous area as a judgment index for judging the driving state.

[0055] In the first analysis process, the measured vibration acceleration (effective value) is plotted along the line L in the distribution diagram of the vibration acceleration (effective value) and shaft rotation speed (Fig. 6). v The measured shaft rotation speed is on the right side of the line L h In the first analysis process, if the vehicle is located below the line L, it is determined that the vehicle is in a dangerous driving state. v The x-coordinate value α and the line L h The y-coordinate value β of the above is generated and stored as a judgment index for judging the driving state. The generation of the judgment index from a certain number of data measured in the past will be explained.

[0056] The measured data is v If it is on the left side of i , straight line L v If the x-coordinate value is on the right side of j In this case, Q is defined by the following equation (Equation 1). In equation (Equation 1), N i is the line L v The number of data to the left of N j is the line L v Indicates the number of data to the right of the

[0057]

number

[0058] Since the equation (Equation 1) is a downward convex curve, the line L is drawn at the position of α corresponding to the minimum value of Q. v Therefore, by partially differentiating equation (1) with respect to α, we obtain the following equation (2).

[0059]

number

[0060] For α corresponding to the minimum value of Q, the left side of equation (Equation 2) takes 0. Using k as a parameter, α is expressed by the following equation (Equation 3). Similarly, for the y coordinate value of the measured data, h The y coordinate value β of is also determined as in equation (3).

[0061]

number

[0062] The above is a description of the first analysis process for generating the determination index for determining the driving state performed by the analysis process (S303) in this embodiment.

[0063] Fig. 7 is a diagram illustrating the second analysis process. As in Fig. 6, the axis of abscissa indicates the vibration acceleration (effective value) and the axis of ordinate indicates the shaft rotation speed, which is an index value indicating the operating state of the rotating equipment measured by the calculation process (S302), and the vibration acceleration (effective value) are shown on a two-dimensional orthogonal coordinate plane with the axis of abscissa indicating the vibration acceleration (effective value) and the axis of abscissa indicating the shaft rotation speed.

[0064] The data set of the dangerous driving state appears in the lower right (towards the fourth quadrant) in FIG. 7. In the second analysis process, a cluster of the dangerous area is designated by a single straight line with a slope. In the second analysis process (S303), a single straight line L designating a cluster of the dangerous area is used as a judgment index for judging the driving state. s Generate and store the equation (Equation 4).

[0065]

number

[0066] In the second analysis, the line L s The slope a and intercept b of the above equation (Equation 4) are generated and stored as a judgment index for judging the driving state. The generation of the judgment index from a certain number of data measured in the past will be explained.

[0067] The measured data is s If the coordinate is above (x i ,y i ), straight line L s If it is below the coordinates of (x j ,y j ) In this case, Q is defined by the following equation (Equation 5). In equation (Equation 5), N u is the line L s The number of data points above N l is the line L s Indicates the number of data points below.

[0068]

number

[0069] In equation (5), (x0, y0) is (x i ,y i ) or (x j ,y j ) to the line L s The following equation (6) means the coordinates of the intersection point (the so-called foot of the perpendicular line) when a perpendicular line is dropped to (x i ,y i ) indicates (x0, y0) in the case of (x j ,y j ) can be calculated in the same way.

[0070]

number

[0071] Equation (5) is the line L s This means the sum of the centers of gravity of the data above and below the line L s This is a quantity that depends on the two constants a and b that define the equation (Equation 4). s If we give an appropriate value to the slope (a) of (e.g., a=1), Q becomes a quantity that depends only on b, and becomes a downward convex curve with respect to b. Therefore, the line L that takes b where Q is minimum is s Using this, the measurement data can be rationally separated into two clusters. To find such b, we simply substitute equation (6) into equation (5) and partially differentiate Q with respect to b, setting the result to 0, and the result is given by equation (7). Note that the slope is set to an appropriate value based on the distribution of the data.

[0072]

number

[0073] The above is a description of the second analysis process for generating the determination index for determining the driving state performed by the analysis process (S303) in this embodiment.

[0074] When the analysis process (S303) is completed, the determination unit 104 starts the determination process (S304). The determination process (S304) determines the operating state by evaluating the value of the index indicating the measured operating state of the rotating equipment calculated in the calculation process (S302) using the determination index generated and stored in the analysis process (S303). In this embodiment, the determination process (S304) evaluates the shaft rotation speed and vibration acceleration (effective value), which are the values ​​of the index indicating the measured operating state of the rotating equipment calculated in the calculation process (S302), using the determination index generated and stored in the analysis process (S303), to determine whether the value of the index indicating the operating state of the rotating equipment is in a dangerous region and the operating state of the rotating equipment is in a dangerous operating state. The determination index is an equation of two straight lines specifying a cluster of the dangerous region obtained by the first analysis process and / or an equation of one straight line specifying a cluster of the dangerous region obtained by the second analysis process.

[0075] The straight line L, which is the judgment index generated and stored in the first analysis process v The x-coordinate value α and the line L h When the determination based on the y coordinate value β is that (vibration acceleration (effective value))>α and (shaft rotation speed)<β, it is determined in the determination process (S304) that the driving state is unsafe.

[0076] The straight line L, which is the judgment index generated and stored in the second analysis process s The determination based on the slope a and intercept b of the equation is that if a×(vibration acceleration (effective value))+b<(shaft rotation speed), it is determined in the determination process (S304) that the driving state is unsafe.

[0077] When the determination process (S304) is completed, the output unit 105 starts the output process (S305). The output process (S305) generates and outputs output information from the operating state determined in the determination process (S304) and the index value indicating the operating state of the rotating equipment.

[0078] As output information, information useful for managing and maintaining the operating status of rotating equipment, such as shaft rotation speed, vibration acceleration (effective value), vibration velocity (effective value), etc. may be generated and output as output information. Furthermore, tables showing numerical values ​​such as maximum, minimum, and average values ​​during the measurement period for these data, graphs showing changes during the measurement period, etc. may be generated and output as output information.

[0079] Countermeasures that can be implemented based on the operating state determined in the determination process (S304) may be generated and output as output information. The output of countermeasures is an effective function for so-called general companies that have small and medium-sized manufacturing facilities, which is the implementation of the present invention, and will be described in detail below.

[0080] Countermeasures that can be implemented when rotating equipment is in an abnormal operating state can be broadly divided into three categories. First, compliance with operating conditions. Specifically, this includes keeping the load within the range allowable for the rotating equipment and operating in a manner that does not exceed environmental conditions. In the explanation of this embodiment, this is referred to as the first countermeasure. Second, appropriate maintenance. Specifically, this includes daily inspection and maintenance such as maintaining the appropriate quality and amount of lubricating oil, properly maintaining the alignment between the motor (the rotation drive source) and the rotating equipment, properly maintaining the condition of belts, etc., and securing the rotating equipment to a sturdy base. In the explanation of this embodiment, this is referred to as the second countermeasure. Third, early detection and repair of equipment abnormalities. Specifically, this includes detecting damage to the rotating equipment due to aging deterioration as early as possible and repairing it. In the explanation of this embodiment, this is referred to as the third countermeasure. In so-called general companies with small to medium-sized manufacturing facilities, it is often difficult to distinguish between these two issues smoothly, and when an abnormality occurs in rotating equipment, response is delayed or incorrect responses are taken, which often leads to the abnormality being prolonged or the equipment suddenly becoming inoperable.

[0081] 8 is a flowchart illustrating the output process (S305) for outputting countermeasures in this embodiment. The countermeasures may be output as standard output information, or may be output at any time by issuing an external output instruction when an abnormality occurs in the rotating equipment. In this example, for ease of understanding, the explanation will be given assuming that the driver, sensing that an abnormality has occurred in the rotating equipment, issues an instruction to output countermeasures.

[0082] When the output process (S305) for outputting countermeasures starts, the process of S801 determines whether the determination process (S304) determined that the operating state is unsafe. If the determination process (S304) determined that the operating state is not unsafe, it is possible that the cause of the abnormality in the rotating equipment is that the operating conditions are outside of the appropriate range. Therefore, the process of S802 outputs a countermeasure (first countermeasure) for complying with the operating conditions. The specific countermeasure differs depending on the rotating equipment, but for example, the output countermeasure may be to check whether the load is within the range allowed for the rotating equipment and make improvements if there are any problems, or to check whether the environmental conditions are exceeded and make improvements if there are any problems.

[0083] If the determination process (S304) determines that an unsafe operating state exists, it is possible that the cause of the abnormality in the rotating equipment is due to insufficient maintenance. Therefore, in the process of S803, appropriate maintenance countermeasures (second countermeasures) are output. Specific countermeasures differ depending on the rotating equipment, but output countermeasures include, for example, checking the quality and amount of lubricating oil and maintaining it at an appropriate level, checking and maintaining appropriate alignment between the motor (the rotation drive source) and the rotating equipment, checking and maintaining appropriate conditions for belts, etc., and checking that the rotating equipment is fixed to its base.

[0084] If the determination process (S304) determines that the vehicle is in an unsafe driving state, the countermeasures (second countermeasures) output because the vehicle is in an unsafe driving state are implemented, and then the present invention is implemented again to determine whether the vehicle is in an unsafe driving state. If the determination process (S304) still determines that the vehicle is in an unsafe driving state despite the implementation of all countermeasures, this may be due to an equipment abnormality, such as damage to equipment due to aging, and a request is made for a detailed inspection or repair by an expert (third countermeasure).

[0085] In this way, by generating and outputting as output information the countermeasures that can be implemented based on the operating state determined in the determination process (S304), so-called general companies with small to medium-sized manufacturing facilities can quickly identify the cause of abnormalities in rotating equipment and take action without relying on experienced personnel.

[0086] The above is a description of the rotating machine abnormality detection device.

[0087] Next, a rotating equipment abnormality detection program for causing a computer to function as a rotating equipment abnormality detection device will be described. The configuration of the computer is as shown in FIG.

[0088] The rotating equipment abnormality detection program includes a main module, an input / output module, and an arithmetic processing module. The main module is the part that controls the overall processing. The input / output module causes a computer to acquire input information such as image data and outputs calculated information to the computer as numerical values ​​or images. The arithmetic processing module includes an input module, a calculation module, an analysis module, a determination module, and an output module. The functions realized by executing the main module, the input / output module, and the arithmetic processing module are similar to the functions of the input unit 101, the calculation unit 102, the analysis unit 103, the determination unit 104, and the output unit 105 of the rotating equipment abnormality detection device 1, respectively.

[0089] The rotating device abnormality detection program may be provided by a storage medium such as a ROM or a semiconductor memory. The calculation program may also be provided via a network.

[0090] The above is the description of the first embodiment. [Example]

[0091] The rotating equipment abnormality detection device, rotating equipment abnormality detection method, and rotating equipment abnormality detection program disclosed as Example 2 of the present invention determine the operating state by inputting the value of an index indicating the operating state of the rotating equipment into a trained learning model that has undergone machine learning to determine the operating state.

[0092] In Example 2, similarly to Example 1, the indicators indicating the operating state of the rotating equipment will be described as the shaft rotation speed and the vibration acceleration (effective value). Also, for the same reasons as in Example 1, the description of the embodiment in which the indicators indicating the operating state of the rotating equipment are the shaft rotation speed and the vibration velocity (effective value) will be omitted.

[0093] The rotating equipment abnormality detection device of this embodiment will be described in comparison with Embodiment 1. In the block diagram of the rotating equipment abnormality detection device of Embodiment 1 shown in Fig. 1, the input unit 101, the calculation unit 102, and the output unit 105 are the same as those of this embodiment in terms of function and operation, and therefore their description will be omitted.

[0094] The rotating equipment anomaly detection device of this embodiment does not include an analysis unit 103. Instead, it includes a trained learning model that has undergone machine learning to determine the operating state from the values ​​of indicators that indicate the operating state of the rotating equipment (in this embodiment, the shaft rotation speed and the vibration acceleration (effective value)). The machine learning may be either unsupervised learning or supervised learning. The learning model may be provided from outside or may be generated inside the rotating equipment anomaly detection device.

[0095] In the case of unsupervised learning, for example, values ​​of indicators indicating the operating state of the rotating equipment (in this embodiment, the shaft rotation speed and vibration acceleration (effective value)) are given as features, and a learning model is generated by unsupervised learning with the operating state (in this embodiment, dangerous driving state and non-dangerous driving state) as the objective variable.

[0096] In the case of supervised learning, for example, a learning model is generated by learning data on risky and non-risky driving states identified by an expert.

[0097] In the rotating equipment abnormality detection device of this embodiment, the judgment unit 104 judges the operating state by inputting the values ​​of the indices (in this embodiment, the shaft rotation speed and vibration acceleration (effective value)) that indicate the measured operating state of the rotating equipment calculated by the calculation unit 102 into a trained learning model that has undergone machine learning to judge the operating state.

[0098] The hardware configuration of the rotating equipment abnormality detection device of this embodiment is the same as the hardware configuration of the rotating equipment abnormality detection device of embodiment 1 shown in Fig. 2. A learning model is stored in a main storage device (RAM / ROM) 204 or an auxiliary storage device 205. Furthermore, like the rotating equipment abnormality detection device of embodiment 1, the rotating equipment abnormality detection device of this embodiment may not only be configured as a standalone device, but may also be incorporated into another device for use. The other device incorporating the calculation device 1 may be, for example, a rotating equipment main body, or an electrical appliance such as a personal computer, a smartphone, or a personal digital assistant.

[0099] The rotating equipment abnormality detection program for causing a computer in this embodiment to function as a rotating equipment abnormality detection device differs from the rotating equipment abnormality detection program of Example 1 in that it is provided with a trained learning model that has undergone machine learning to determine the operating state from the values ​​of indicators that indicate the operating state of the rotating equipment (in this embodiment, shaft rotation speed and vibration acceleration (effective value)), and the function realized by executing the judgment module determines the operating state by inputting the values ​​of indicators that indicate the measured operating state of the rotating equipment calculated by the calculation module (in this embodiment, shaft rotation speed and vibration acceleration (effective value)) into a trained learning model that has undergone machine learning to determine the operating state.

[0100] This concludes the description of the second embodiment. [Example]

[0101] The rotating equipment abnormality detection device, rotating equipment abnormality detection method, and rotating equipment abnormality detection program disclosed as Example 3 of the present invention determine the operating state by using the results of analyzing the values ​​of indicators indicating the operating state of the rotating equipment by spectral analysis.

[0102] In Example 3, similarly to Example 1, the indicators indicating the operating state of the rotating equipment will be described as the shaft rotation speed and the vibration acceleration (effective value). Also, for the same reason as Example 1, the description of the embodiment in which the indicators indicating the operating state of the rotating equipment are the shaft rotation speed and the vibration velocity (effective value) will be omitted.

[0103] The rotating equipment abnormality detection device of this embodiment will be described in comparison with Embodiment 1. In the block diagram of the rotating equipment abnormality detection device of Embodiment 1 shown in Fig. 1, the input unit 101, the calculation unit 102, and the output unit 105 are the same as those of this embodiment in terms of function and operation, and therefore their description will be omitted.

[0104] In the rotating equipment abnormality detection device of this embodiment, the analysis unit 103 performs envelope analysis on the measured data using fast Fourier transform on the shaft rotation speed and vibration acceleration (effective value), which are indices that indicate the operating state of the rotating equipment, thereby identifying a spectrum that is characteristic of a dangerous operating state and storing it as a judgment index.

[0105] In the rotating machine abnormality detection device of this embodiment, when a spectrum characteristic of a dangerous operating state stored in the analysis unit 103 occurs, the determination unit 104 determines that the operating state is dangerous.

[0106] The rotating equipment anomaly detection device of this embodiment may be provided with a learning model for determining the driving state, which is generated by deep learning of a spectrum characteristic of the dangerous driving state identified by the analysis unit 103, when the determination unit 104 determines that the driving state is a dangerous driving state, or a spectrum pattern when an expert determines that the driving state is a dangerous driving state, and this learning model may be used for determination. The learning model may be provided externally or may be generated internally in the rotating equipment anomaly detection device. The determination unit 104 determines the driving state by inputting the measured spectral pattern into a trained learning model that has undergone machine learning to determine the driving state.

[0107] The hardware configuration of the rotating equipment abnormality detection device of this embodiment is the same as the hardware configuration of the rotating equipment abnormality detection device of embodiment 1 shown in Fig. 2. A learning model is stored in a main storage device (RAM / ROM) 204 or an auxiliary storage device 205. Furthermore, like the rotating equipment abnormality detection device of embodiment 1, the rotating equipment abnormality detection device of this embodiment may not only be configured as a standalone device, but may also be incorporated into another device for use. The other device incorporating the calculation device 1 may be, for example, a rotating equipment main body, or an electrical appliance such as a personal computer, a smartphone, or a personal digital assistant.

[0108] The rotating equipment abnormality detection program in this embodiment, which causes a computer to function as a rotating equipment abnormality detection device, differs from the rotating equipment abnormality detection program of Example 1 in that the function realized by executing the analysis module performs envelope analysis using fast Fourier transform on the values ​​of the shaft rotation speed and vibration acceleration (effective value), which are indicators of the operating state of the rotating equipment, identifies a spectrum characteristic of a dangerous operating state and stores it as a judgment indicator, is provided with a learning model for judging the operating state by deep learning the spectral pattern, and the function realized by executing the judgment module judges that a dangerous operating state exists based on a spectrum or spectral pattern characteristic of a dangerous operating state.

[0109] This concludes the description of the third embodiment. [Explanation of symbols]

[0110] 1. Rotating equipment abnormality detection device 101 Input section 102 Calculation Unit 103 Analysis Department 104 Judgment section 105 Output section 201 Central processing unit (CPU) 202 Input Device 203 Output Device 204 Main memory (RAM / ROM) 205 Auxiliary Memory Device

Claims

1. An abnormality detection device for a rotating device, an input unit that receives and stores measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of abnormality detection; a calculation unit that calculates an index value indicating an operating state of the rotating equipment from the data received and accumulated by the input unit; an analysis unit that analyzes the value of an index that indicates an operating state of the rotating equipment calculated by the calculation unit, and generates and stores a judgment index that determines the operating state of the rotating equipment; a determination unit that determines the operating state of the rotating equipment by evaluating the value of an index that indicates the operating state of the rotating equipment, calculated by the calculation unit, using the determination index that is generated and held by the analysis unit; and an output unit that generates output information from the operating state of the rotating device determined by the determination unit and an index value indicating the operating state of the rotating device, and outputs the output information; Equipped with the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, the determination index is an index for each of the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determination unit determines that the operating state of the rotating device is abnormal when the rotation speed of the rotating shaft of the rotating device is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index. An abnormality detection device for a rotating machine,

2. An abnormality detection device for a rotating device, an input unit that receives and stores measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of abnormality detection; a calculation unit that calculates an index value indicating an operating state of the rotating equipment from the data received and accumulated by the input unit; an analysis unit that analyzes the value of an index that indicates an operating state of the rotating equipment calculated by the calculation unit, and generates and stores a judgment index that determines the operating state of the rotating equipment; a determination unit that determines the operating state of the rotating equipment by evaluating the value of an index that indicates the operating state of the rotating equipment, calculated by the calculation unit, using the determination index that is generated and held by the analysis unit; and an output unit that generates output information from the operating state of the rotating device determined by the determination unit and an index value indicating the operating state of the rotating device, and outputs the output information; Equipped with the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, The judgment index is a function indicating a correlation between the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determination unit determines that the operating state of the rotating device is abnormal when the index is in a region divided by the function that is the determination index, where the rotation speed of the rotating shaft of the rotating device is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index. An abnormality detection device for a rotating machine,

3. the determination unit determines the operating state of the rotating equipment by inputting the value of the index that indicates the measured operating state of the rotating equipment, calculated by the calculation unit, into a trained learning model that has undergone machine learning for determining the operating state of the rotating equipment.

3. The rotating equipment abnormality detection device according to claim 1 or 2.

4. when the determination unit determines that the operating state of the rotating equipment is a dangerous operating state, the determination unit transfers executable countermeasures to the output unit; The output information generated and output by the output unit includes the countermeasure.

3. The rotating equipment abnormality detection device according to claim 1 or 2.

5. A method for detecting an abnormality in a rotating device, comprising: an input step of receiving and accumulating measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of anomaly detection; a calculation step of calculating an index value indicating an operating state of the rotating equipment from the data received and accumulated in the input step; an analysis step of analyzing the value of the index indicating the operating state of the rotating equipment calculated in the calculation step, and generating and storing a judgment index for judging the operating state of the rotating equipment; a determination step of determining an operating state of the rotating equipment by evaluating a value of an index indicating the operating state of the rotating equipment calculated in the calculation step using the determination index generated and stored in the analysis step; an output step of generating and outputting output information from the operating state of the rotating device determined in the determination step and a value of an index indicating the operating state of the rotating device; Equipped with the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, the determination index is an index for each of the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determining step determines that the operating state of the rotating device is abnormal when the rotation speed of the rotating shaft of the rotating device is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index; A method for detecting abnormalities in a rotating device, comprising:

6. A method for detecting an abnormality in a rotating device, comprising: an input step of receiving and accumulating measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of anomaly detection; a calculation step of calculating an index value indicating an operating state of the rotating equipment from the data received and accumulated in the input step; an analysis step of analyzing the value of the index indicating the operating state of the rotating equipment calculated in the calculation step, and generating and storing a judgment index for judging the operating state of the rotating equipment; a determination step of determining an operating state of the rotating equipment by evaluating a value of an index indicating the operating state of the rotating equipment calculated in the calculation step using the determination index generated and stored in the analysis step; an output step of generating and outputting output information from the operating state of the rotating device determined in the determination step and a value of an index indicating the operating state of the rotating device; Equipped with the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, The judgment index is a function indicating a correlation between the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determination step determines that the operating state of the rotating equipment is abnormal when the index is in a region divided by the function that is the determination index, where the rotation speed of the rotating shaft of the rotating equipment is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index. A method for detecting abnormalities in a rotating device, comprising:

7. A rotating equipment abnormality detection program, an input step of receiving and accumulating measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of anomaly detection; a calculation step of calculating an index value indicating an operating state of the rotating equipment from the data received and accumulated in the input step; an analysis step of analyzing the value of the index indicating the operating state of the rotating equipment calculated in the calculation step, and generating and storing a judgment index for judging the operating state of the rotating equipment; a determination step of determining an operating state of the rotating equipment by evaluating a value of an index indicating the operating state of the rotating equipment calculated in the calculation step using the determination index generated and stored in the analysis step; an output step of generating and outputting output information from the operating state of the rotating device determined in the determination step and a value of an index indicating the operating state of the rotating device; on the computer, the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, the determination index is an index for each of the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determining step determines that the operating state of the rotating device is abnormal when the rotation speed of the rotating shaft of the rotating device is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index; A rotating equipment abnormality detection program characterized by

8. A rotating equipment abnormality detection program, an input step of receiving and accumulating measurement data of low-frequency vibrations during operation of the rotating equipment that is the target of anomaly detection; a calculation step of calculating an index value indicating an operating state of the rotating equipment from the data received and accumulated in the input step; an analysis step of analyzing the value of the index indicating the operating state of the rotating equipment calculated in the calculation step, and generating and storing a judgment index for judging the operating state of the rotating equipment; a determination step of determining an operating state of the rotating equipment by evaluating a value of an index indicating the operating state of the rotating equipment calculated in the calculation step using the determination index generated and stored in the analysis step; an output step of generating and outputting output information from the operating state of the rotating device determined in the determination step and a value of an index indicating the operating state of the rotating device; on the computer, the index indicating the operating state of the rotating device is a rotation speed of a rotating shaft of the rotating device and an effective value of vibration acceleration, The judgment index is a function indicating a correlation between the rotation speed of the rotating shaft of the rotating device and the effective value of vibration acceleration, the determination step determines that the operating state of the rotating equipment is abnormal when the index is in a region divided by the function that is the determination index, where the rotation speed of the rotating shaft of the rotating equipment is smaller than the determination index and the effective value of the vibration acceleration is larger than the determination index. A rotating equipment abnormality detection program characterized by

Images