Method for detecting abnormality and abnormality precursor of rotating machine

Abstract

A method for detecting an abnormality and an abnormality precursor of a rotating machine includes the following steps: a measurement step of measuring noise or vibration generated from the rotating machine; a frequency spectrum map creation step of creating a frequency spectrum map of three-dimensional data of time, frequency, and amplitude at each frequency by performing high-speed Fourier transform on the measured noise or vibration; and an abnormality determination step of comparing the created frequency spectrum map with a frequency spectrum map at a normal time, and determining whether an abnormality and an abnormality precursor have occurred in the rotating machine based on whether an abnormal signal appears in the created frequency spectrum map. In the detection method, the created frequency spectrum map is used to detect an abnormal signal that appears in proportion to a change in the rotational speed of the rotating machine when the amplitude at each frequency in the frequency spectrum map at the normal time is larger than the amplitude at each frequency.

Description

Technical Field

[0001] This invention relates to a method for detecting abnormalities and signs of abnormalities in rotating machinery. Background Technology

[0002] As rotating machinery, fans, turbines, compressors, generators, motors, etc., are known. This rotating machinery is mounted in various devices. For example, an air conditioner houses a fan motor as rotating machinery. In the fan motor (rotating machinery) of this air conditioner, it is important to check for abnormalities before serious malfunctions occur; therefore, regular mechanical diagnostic checks for fan motor abnormalities are performed. Furthermore, mechanical diagnostic checks for abnormalities are sometimes performed not only on fan motors in air conditioners, but also on other types of rotating machinery.

[0003] Mechanical diagnostics for checking fan motor malfunctions in air conditioners include measuring fan motor noise or vibration using a microphone or vibrometer, performing a high-speed Fourier transform (FFT) on the measured noise or vibration, and analyzing the frequency to determine the vibration corresponding to each frequency component. The malfunction of the air conditioner's fan motor is then checked based on whether vibrations exceeding a pre-set abnormal vibration threshold are present in the determined vibrations corresponding to these frequency components.

[0004] In addition, in previous mechanical diagnostics, there have been cases where abnormalities in the fan motor could not be detected.

[0005] In the fan motor driven and controlled by the inverter in the air conditioner, noise and vibration caused by the inverter's set carrier frequency (the frequency that determines the pulse width modulation period of the inverter in PWM control mode) is in the high-frequency range of 2kHz or higher. Therefore, even during normal operation, noise and vibration will be generated due to the set carrier frequency of the inverter.

[0006] Therefore, in the vibrations corresponding to the frequency components obtained through mechanical diagnostics, if the vibration caused by the fan motor malfunction and the vibration caused by the inverter's set carrier frequency occur at different frequencies, the fan motor malfunction can be detected. However, if the vibration caused by the fan motor malfunction and the vibration caused by the inverter's set carrier frequency occur at the same frequency, the vibration caused by the fan motor malfunction cannot be detected.

[0007] Figure 8 It is a graph showing the vibrations in a fan motor corresponding to the frequency components.

[0008] For example, when the inverter's set carrier frequency is 6kHz, as shown in the figure, among the vibrations corresponding to the frequency components obtained through mechanical diagnostics, the vibration caused by the inverter's set carrier frequency (represented by F in the graph) appears at approximately 6kHz. At this time, when the vibration caused by the fan motor malfunction (represented by E in the graph) also appears at approximately 6kHz, since the various vibrations overlap at the same frequency, the vibration caused by the fan motor malfunction cannot be detected, thus making it impossible to identify the fan motor malfunction.

[0009] As a result, there is a problem of serious malfunctions occurring in the fan motors of air conditioners. Summary of the Invention

[0010] The problem the invention aims to solve

[0011] The present invention was made in view of the following problem, and its object is to be able to reliably detect anomalies and signs of anomalies occurring in rotating machinery in order to prevent serious failures in rotating machinery.

[0012] Solution for solving the problem

[0013] This invention discloses a method for detecting abnormalities and potential abnormalities in rotating machinery. The method detects abnormalities and potential abnormalities in the rotating machinery based on noise or vibration measured by a measuring instrument. The method includes the following steps: a measurement step, in which noise or vibration generated by the rotating machinery is measured using a measuring instrument; a spectrum generation step, in which a high-speed Fourier transform is performed on the measured noise or vibration to generate a spectrum of three-dimensional data including time, frequency, and amplitude at each frequency; and an abnormality determination step, in which the generated spectrum is compared with a normal spectrum, and the presence of abnormal signals not present in the normal spectrum in the generated spectrum determines whether an abnormality or potential abnormality has occurred in the rotating machinery. In this method, the generated spectrum is used to detect abnormal signals where the amplitude at each frequency is larger than that in the normal spectrum when the rotational speed of the rotating machinery changes, and where the amplitude at each frequency is proportional to the change in rotational speed.

[0014] The effects of the invention

[0015] According to the present invention, a measuring instrument is used to measure noise or vibration generated from rotating machinery. A spectrum diagram of three-dimensional data, including time, frequency, and amplitude at each frequency, is generated based on the noise or vibration generated from the rotating machinery. The presence of abnormal signals in the generated spectrum diagram is used to determine whether an abnormality or potential abnormality has occurred in the rotating machinery. Therefore, abnormalities and potential abnormalities in rotating machinery can be reliably detected, preventing serious malfunctions in the rotating machinery. Attached Figure Description

[0016] Figure 1 This is a block diagram of an air conditioner.

[0017] Figure 2 This is a block diagram of the equipment used in the method for detecting abnormalities and signs of abnormalities in rotating machinery according to the present invention.

[0018] Figure 3 This is a flowchart of a method for detecting abnormalities and warning signs of abnormalities in the fan motor of an air conditioner.

[0019] Figure 4 It is a spectrum diagram showing the time, frequency, and amplitude of the three-dimensional data produced.

[0020] Figure 5 It is shown schematically. Figure 4 The spectrum diagram.

[0021] Figure 6 It is a spectrum diagram showing the time, frequency, and amplitude of three-dimensional data at each frequency under normal conditions.

[0022] Figure 7 It is shown schematically. Figure 6 The spectrum diagram.

[0023] Figure 8 It is a graph showing the vibrations in a fan motor corresponding to the frequency components.

[0024] Explanation of reference numerals in the attached figures

[0025] 1: Air conditioner; 2: Fan motor; 3: Inlet; 4: Outlet; 5: Filter; 6: Fan; 7: Coil; 8: Humidifier; 11: Return air pipe; 12: Supply air pipe; 15: Measuring device; 16: Inverter; 17: Processing unit; 18: Display unit. Detailed Implementation

[0026] An embodiment of the method for detecting abnormalities and signs of abnormalities in rotating machinery according to the present invention will be described.

[0027] The method for detecting abnormalities and signs of abnormalities in rotating machinery involved in this embodiment is a method for detecting abnormalities and signs of abnormalities occurring in rotating machinery, such as blowers, turbines, compressors, generators, motors, etc.

[0028] The following description is given as an example and is directed to the fan motor 2 of the air conditioner 1, which is a rotating machine. That is, this embodiment is a method for detecting abnormalities and signs of abnormalities in the fan motor 2 of the air conditioner 1, but it can also be a method for detecting abnormalities and signs of abnormalities in other rotating machines.

[0029] Figure 1 This is a block diagram of air conditioner 1.

[0030] As shown in the figure, the air conditioner 1 has an intake 3 connected to a return air duct 11 and an outlet 4 connected to a supply air duct 12. Inside the air conditioner 1, there are, for example, a filter 5, coils (heating coil and cooling coil) 7 serving as heat exchangers, a humidifier 8, a fan 6, and a fan motor 2 arranged sequentially from the intake 3 to the outlet 4. The fan motor 2 is driven and controlled by an inverter (not shown). In this air conditioner 1, air entering from the intake 3 passes through the filter 5, coils 7, humidifier 8, and fan 6 and is blown out from the outlet 4.

[0031] Next, with reference to the accompanying drawings, the method for detecting abnormalities and signs of abnormalities in the fan motor 2 (rotating machinery) of the air conditioner 1 according to this embodiment will be described.

[0032] (Equipment Structure)

[0033] Figure 2 This is a block diagram of the equipment used in the detection method for abnormalities and signs of abnormalities in the fan motor 2 of the air conditioner 1.

[0034] The device used in the method for detecting abnormalities and signs of abnormalities in the fan motor 2 of the air conditioner 1 includes, as shown in the figure: a measuring device 15, which is disposed adjacent to the fan motor 2 of the air conditioner 1; an inverter 16, which drives and controls the fan motor 2; and a processing device 17, which receives measuring data from the measuring device 15 and receives signals from the inverter 16.

[0035] The measuring device 15 is a sound sensor or vibration sensor that measures noise or vibration. It starts or stops the measurement based on instructions from the processing device 17 and sends the measured noise or vibration value as measurement data to the processing device 17.

[0036] The processing device 17 is a computer that stores various programs and uses these programs for processing. Here, it receives measurement data from the measuring device 15, as well as operation start signal, operation stop signal and speed setting signal sent from the inverter 16 to the fan motor 2 that is driven and controlled by the inverter 16.

[0037] In the processing device 17, the existing program is used to perform a high-speed Fourier transform (FFT) and frequency analysis on the measurement data sent from the measuring device 15 to generate a spectrum diagram SPA of three-dimensional data of time, frequency and amplitude at each frequency, and the spectrum diagram SPA is stored.

[0038] In addition, the processing device 17 also generates and stores a spectrum diagram (SPB) of three-dimensional data of time, frequency, and amplitude at each frequency under normal conditions. This spectrum diagram (SPB) of three-dimensional data of time, frequency, and amplitude at each frequency under normal conditions is generated by measuring the fan motor 2 in advance by the measuring device 15 when there is no abnormality (normal condition), and by performing high-speed Fourier transform (FFT) and frequency analysis based on the measured data.

[0039] In addition, the processing device 17 includes a display unit 18. The display unit 18 displays a spectrum diagram SPA of the generated three-dimensional data, including time, frequency, and amplitude at each frequency. Furthermore, the display unit 18 can also display a spectrum diagram SPB of the normal three-dimensional data, including time, frequency, and amplitude at each frequency.

[0040] In the display unit 18, the generated spectrogram SPA and the normal spectrogram SPB are displayed side by side on two screens. In addition, the display unit 18 can switch the screen. Regarding the screen switching, it can switch from displaying two screens to displaying only one of the generated spectrogram SPA and the normal spectrogram SPB, or it can switch from displaying one screen to displaying two screens.

[0041] (Methods for detecting abnormalities and their early signs)

[0042] Figure 3 This is a flowchart of a method for detecting abnormalities and warning signs of abnormalities in the fan motor of an air conditioner.

[0043] As shown in the figure, the method for detecting abnormalities and signs of abnormality in the fan motor 2 of the air conditioner 1 includes the following steps: measurement step (S1), measuring the noise or vibration generated from the fan motor 2; spectrum diagram production step (S2), producing a spectrum diagram SPA of three-dimensional data of time, frequency and amplitude at each frequency; and abnormality determination step (S3), determining whether an abnormality or signs of abnormality has occurred in the fan motor 2.

[0044] In the measurement step (S1) for measuring the noise or vibration generated from the fan motor 2, the measuring device 15 arranged adjacent to the fan motor 2 is used to measure the noise or vibration generated from the fan motor 2.

[0045] That is, the inverter 16, which drives and controls the fan motor 2, sends a start signal (S1-1) for the operation of the fan motor 2 to the processing unit 17.

[0046] Based on the operation start signal, the processing device 17 sends a measurement start signal to the measuring device 15 (S1-2).

[0047] The measuring device 15 starts the measurement based on the measurement start signal (S1-3). In the measurement of the measuring device 15, the noise or vibration generated by the fan motor 2 is measured from the acceleration operation when the speed increases after the fan motor 2 is started until the steady-state operation when the speed is constant.

[0048] The noise or vibration value measured by the measuring device 15 is sent as measurement data to the processing device 17 (S1-4).

[0049] In the spectrum plotting process (S2) of creating a spectrum plot SPA of three-dimensional data of time, frequency and amplitude at each frequency, a high-speed Fourier transform (FFT) is performed on the noise or vibration measured by the measuring device 15 to create a spectrum plot SPA of three-dimensional data of time, frequency and amplitude at each frequency.

[0050] That is, in the processing device 17, the existing program is used to perform high-speed Fourier transform (FFT) and frequency analysis on the measurement data (noise value or vibration value) sent from the measuring device 15 to produce a spectrum diagram SPA (S2-1) of three-dimensional data of time, frequency and amplitude at each frequency.

[0051] Next, in the processing device 17, a spectrum diagram SPA (S2-2) of the three-dimensional data of the time, frequency and amplitude at each frequency is stored.

[0052] In the abnormality determination process (S3) to determine whether an abnormality or abnormality signs have occurred in the fan motor 2, the spectrum diagram SPA of the three-dimensional data of time, frequency and amplitude at each frequency produced in the previous process is compared with the spectrum diagram SPB of the three-dimensional data of time, frequency and amplitude at each frequency under normal conditions to determine whether an abnormality or abnormality signs have occurred in the fan motor 2.

[0053] That is, in the processing device 17, the spectrum diagram SPA of the three-dimensional data of time, frequency and amplitude at each frequency and the spectrum diagram SPB of the three-dimensional data of time, frequency and amplitude at each frequency in the pre-stored normal time are displayed side by side on the display unit 18 on two screens (S3-1).

[0054] Next, the generated spectrum SPA displayed on display unit 18 is compared with the normal spectrum SPB (S3-2). This comparison is performed by the work supervisor. That is, the work supervisor observes and compares the generated spectrum SPA displayed on display unit 18 with the normal spectrum SPB.

[0055] When comparing the generated spectrogram SPA with the normal spectrogram SPB, the work supervisor checks whether the generated spectrogram SPA contains any abnormal signals K that are not present in the normal spectrogram SPB (refer to...). Figure 4 The system uses a frequency spectrum analyzer (S3-3) to determine whether an abnormality or abnormal sign has occurred in the fan motor 2. In other words, if an abnormal signal K appears in the generated frequency spectrum analyzer SPA (if an abnormal signal K is detected), it is determined that an abnormality or abnormal sign has occurred in the fan motor 2; if no abnormal signal K appears, it is determined that no abnormality or abnormal sign has occurred in the fan motor 2.

[0056] The abnormal signal K appearing in the generated spectrum diagram SPA is illustrated with reference to the attached diagram. Furthermore, the set carrier frequency of the inverter 16 in this case is set to 6 kHz.

[0057] Figure 4 This is a spectral graph (SPA) showing the generated three-dimensional data of time, frequency, and amplitude at each frequency. Figure 5 It is shown schematically. Figure 4 The spectrum of SPA. Figure 6 This is a spectrum plot (SPB) of three-dimensional data showing time, frequency, and amplitude at each frequency under normal conditions. Figure 7 It is shown schematically. Figure 6 The spectrum of SPB.

[0058] exist Figure 4 , Figure 5The spectrum plot SPA of the three-dimensional data produced, including time, frequency, and amplitude at each frequency, is shown. Figure 6 , Figure 7 In the spectrum graph SPB showing the normal time, frequency, and amplitude at each frequency in a three-dimensional dataset, the horizontal axis represents time, the vertical axis represents frequency, and the color of each location indicates the magnitude of the amplitude at each frequency. The colors of these locations are, for example, blue (B), green (G), yellow (Y), and red (R) in the graph. Smaller amplitudes are represented by blue (B), and as the amplitude increases, they change to green (G), yellow (Y), and red (R). By representing the amplitude at each frequency with color in this way, it is easy to identify whether an abnormal signal K has occurred; that is, the detection of abnormal signal K becomes easier.

[0059] Additionally, regarding the spectrograms SPA and SPB, in Figure 5 and Figure 7 In the schematic diagram, a boundary line appears clearly as the color changes from blue (B) to green (G), from green (G) to yellow (Y), and from yellow (Y) to red (R). However, in reality, the boundary line does not appear clearly; rather, the color changes gradually.

[0060] In addition, in these spectrum diagrams SPA and SPB, the amplitude at each frequency is represented by color, but instead of color, the amplitude at each frequency can be numerically represented, with the amplitude at each frequency being represented by a numerical value at each location.

[0061] The abnormal signal K is as follows: When the fan motor 2 starts to run and accelerates, that is, when the speed of the fan motor 2 increases (changes), the amplitude of each frequency in the spectrum diagram SPA is larger than that in the normal spectrum diagram SPB, and the amplitude of each frequency is proportional to the increase in the speed of the fan motor 2.

[0062] That is, in the event of an abnormality or abnormal signs occurring in fan motor 2, such as Figure 4 , Figure 5 As shown in the generated spectrum, when the speed of fan motor 2 increases, the amplitude at each frequency (yellow (Y) and red (R)) that is larger than the normal amplitude at each frequency in the spectrum SPB appears proportionally to the increase in the speed of fan motor 2. That is, an abnormal signal K appears in the generated spectrum. On the other hand, as... Figure 6 , Figure 7 As shown, in the spectrum diagram under normal conditions, the amplitudes at each frequency that are larger than those in the spectrum diagram SPB under normal conditions (yellow (Y) and red (R)) do not appear proportionally to the increase in the speed of the fan motor 2.

[0063] In a spectrum diagram produced in this way, when the amplitude at each frequency is larger than that in the normal spectrum diagram SPB, and the increase in the speed of the fan motor 2 is proportional (when the abnormal signal K appears), it can be determined that an abnormality or abnormal sign has occurred in the fan motor 2.

[0064] As described above, a spectrum diagram SPA of three-dimensional data, including time, frequency, and amplitude at each frequency, is generated based on the noise or vibration generated by the fan motor 2 as measured by the measuring device 15. The presence of an abnormal signal K in the generated spectrum diagram SPA determines whether an abnormality or potential abnormality has occurred in the fan motor 2. Therefore, abnormalities and potential abnormalities in the fan motor 2 can be reliably detected, preventing serious malfunctions.

[0065] In addition, the abnormality and abnormal signs of the fan motor 2 can be detected when the air conditioner 1 starts to operate, that is, when the fan motor 2 starts to operate. However, the abnormality and abnormal signs of the fan motor 2 can also be detected daily, or as needed, such as once every 1 to 2 weeks.

[0066] Alternatively, the detection of abnormalities and abnormal signs can be performed when the fan motor 2 stops running, instead of when the fan motor 2 starts running.

[0067] That is, when the processing device 17 receives a stop signal from the inverter 16, it sends a measurement start signal to the measuring device 15 based on the stop signal, thereby starting the measurement using the measuring device 15. In the measurement using the measuring device 15, the noise or vibration generated by the fan motor 2 is measured from the steady-state operation of the fan motor 2 with a constant speed until the rotation stops when the speed decreases.

[0068] Subsequently, similar to the above embodiment, a high-speed Fourier transform is performed on the noise or vibration measured by the measuring device 15 to generate a spectrum diagram SPA of three-dimensional data, including time, frequency, and amplitude at each frequency. The generated spectrum diagram SPA is compared with the normal spectrum diagram SPB. Based on whether an abnormal signal K appears in the generated spectrum diagram SPA, it is determined whether an abnormality or abnormality precursor has occurred in the fan motor 2. The abnormal signal K at this time is as follows: when the fan motor 2 stops running, that is, when the speed of the fan motor 2 decreases (changes), the amplitude at each frequency in the generated spectrum diagram SPA that is larger than that in the normal spectrum diagram SPB appears proportionally to the decrease in the speed of the fan motor 2.

[0069] In addition, the detection of abnormalities and abnormal signs is not limited to when the fan motor 2 starts or stops running, but can also be performed when the speed of the fan motor 2 changes (when the inverter 16 sends a speed setting signal to the fan motor 2).

[0070] Furthermore, in the above embodiment, during the anomaly determination process (S3), the comparison between the generated spectrum diagram SPA and the normal spectrum diagram SPB is performed by the person in charge of the operation. However, this is not a limitation; for example, artificial intelligence (AI) technology can be used to compare the generated spectrum diagram SPA with the normal spectrum diagram SPB to determine whether an anomaly or an abnormality has occurred in the fan motor 2. Thus, unlike the case where the comparison is performed by the person in charge of the operation, the detection of anomalies and abnormality signs can be performed continuously.

Claims

1. A method for detecting abnormalities and early warning signs of rotating machinery, comprising detecting abnormalities and early warning signs of abnormalities occurring in the rotating machinery based on noise or vibration measured by a measuring instrument measuring noise or vibration generated from the rotating machinery, the method being characterized by comprising the following steps: The measurement process involves using a measuring instrument to measure noise or vibration generated from rotating machinery. The spectrum generation process involves performing a high-speed Fourier transform on the measured noise or vibration to generate a three-dimensional spectrum of time, frequency, and amplitude at each frequency; and The anomaly detection process involves comparing the generated spectrum diagram with the spectrum diagram under normal conditions. Based on whether any abnormal signals not present in the normal spectrum diagram appear in the generated spectrum diagram, it is determined whether an anomaly or any signs of an anomaly have occurred in the rotating machinery. In the method for detecting abnormalities and precursors of rotating machinery, a generated spectrum is used to detect abnormal signals that occur at the start or stop of operation of the rotating machinery when its rotational speed changes, with amplitudes at each frequency exceeding those in the normal state. The amplitude at each frequency is proportional to the change in the rotational speed of the machinery. The start of operation of the rotating machinery is the period from when the rotating machinery starts and the speed begins to increase until it reaches a steady state with a constant speed. The stop of operation of the rotating machinery is the period from when the rotating machinery responds to a stop signal and the speed begins to decrease from a steady state with a constant speed until it stops rotating.

2. The method for detecting abnormalities and early signs of abnormalities in rotating machinery according to claim 1, characterized in that, In a spectrum of three-dimensional data including time, frequency, and amplitude at each frequency, the magnitude of the amplitude at each frequency is indicated by color differences.

Images