State diagnosis device, state diagnosis method, and abnormality indication estimation device for electric motor

Abstract

The present invention obtains a state diagnosis device for a motor, in which a unique peak value extraction unit extracts a peak value intensity of a sideband wave generated by a belt gear frequency from a peak value intensity of the sideband wave detected from a spectral waveform by a calculation unit, and a determination unit performs state diagnosis of the motor by comparing the extracted peak value intensity with a peak value intensity threshold value. Therefore, the number of parameters required by diagnosis can be reduced, and the time spent on diagnosis is shortened. A state diagnosis device (100) for an electric motor (5) is provided with: a current input unit (7) that inputs current data detected from the electric motor (5) by a current detector (4); an analysis unit (120) that performs frequency analysis on the current data input by the current input unit (7) and calculates an analysis result; a calculation unit (112) that detects a plurality of peak intensities of sideband waves of the power supply frequency of the motor (5) from the analysis result; a natural peak value extraction unit (131) that calculates the natural frequency of the belt gear unit (60) from the rated information of the motor (5), and extracts, from among the plurality of peak value intensities detected by the calculation unit (112), the peak value intensity of a sideband wave generated at a position deviating from the natural frequency with respect to the power supply frequency of the motor (5); and a determination unit (132) that diagnoses the motor (5) by comparing a peak intensity threshold value calculated on the basis of the normally operating motor (5) with the peak intensity extracted by the unique peak extraction unit (131).

Description

Technical Field

[0001] This disclosure relates to a condition diagnosis device, a condition diagnosis method, and an anomaly prediction device for electric motors. Background Art

[0002] Electric motors are indispensable in industrial plants, serving as the power source for production lines and machinery. Therefore, continuous and stable operation is always required.

[0003] Traditional motor condition diagnosis relies on inspectors visually assessing operation and listening to sounds during operation to identify any abnormalities. However, this method depends heavily on the inspector's experience or intuition, leading to significant accuracy discrepancies. Furthermore, given the high likelihood of sudden motor failures, there is a growing emphasis on continuous monitoring technologies.

[0004] In continuous monitoring technology for electric motors, there is a method for diagnosing abnormalities by detecting the current applied to the motor. For example, Patent Document 1 discloses an abnormality detection method for a rotating machinery system, which detects abnormalities in the motor by measuring the current applied to it and performing frequency analysis. Prior art literature Patent Literature

[0005] Patent Document 1: Japanese Patent Application Publication No. 2017-181437 Summary of the Invention Technical problem to be solved by the invention

[0006] In the aforementioned anomaly detection method for rotating machinery systems, a high-speed Fourier transform is performed on the detected motor current information. Characteristic frequencies are extracted from the obtained analysis results, and degradation parameters used to determine the degree of degradation in the rotating machinery system are calculated based on their peak values. On the other hand, the degradation parameters used for anomaly determination in rotating machinery systems require calculation of multiple parameters, including the effective value of the current, three-phase current balance, and current harmonic ratio.

[0007] This disclosure was made to solve the above-mentioned problems, and its purpose is to provide a motor condition diagnosis device, a condition diagnosis method, and an anomaly prediction device that can perform motor condition diagnosis by using the peak intensity of the sideband wave caused by the natural frequency of the geared section as a parameter, without the need to calculate multiple parameters. Technical means for solving technical problems

[0008] The electric motor condition diagnostic device disclosed herein includes: a current input unit that inputs current data detected from the electric motor by a current detector; an analysis unit that performs frequency analysis on the current data input from the current input unit and calculates the analysis results; a calculation unit that detects multiple peak intensities of sideband waves of the power supply frequency of the electric motor from the analysis results; an inherent peak extraction unit that calculates the inherent frequency of the geared section from the rated information of the electric motor and extracts the peak intensities of sideband waves generated at positions that deviate from the inherent frequency based on the power supply frequency of the electric motor from the multiple peak intensities detected in the calculation unit; and a judgment unit that compares a peak intensity threshold calculated from a normally operating electric motor with the peak intensity extracted from the inherent peak extraction unit to diagnose the electric motor.

[0009] Furthermore, the condition diagnosis method for an electric motor disclosed herein includes: a step of inputting current data detected from the electric motor by a current detector; a step of performing frequency analysis on the input current data and calculating the analysis result; a step of detecting multiple peak intensities of the sideband wave of the power supply frequency of the electric motor from the analysis result; a step of calculating the natural frequency of the geared portion from the rated information of the electric motor, and extracting the peak intensity of the sideband wave generated at a position corresponding to the natural frequency based on the power supply frequency of the electric motor from the detected multiple peak intensities; and a step of diagnosing the electric motor by comparing the peak intensity threshold calculated from a normally operating electric motor with the peak intensity.

[0010] Furthermore, the abnormal premonition prediction device for electric motors disclosed herein includes: a data acquisition unit that acquires current data and a diagnostic result of the electric motor associated with the current data prior to frequency analysis from the aforementioned electric motor condition diagnostic device; a completion learning model generation unit that learns from the current data acquired by the data acquisition unit based on the diagnostic result and generates a completion learning model; and a current data prediction unit that uses the completion learning model to predict abnormal premonitions of the electric motor from the newly acquired current data by the data acquisition unit and outputs the prediction result. Invention Effects

[0011] According to the motor condition diagnosis apparatus and method of this disclosure, the inherent peak extraction unit extracts the peak intensity of the sideband wave caused by the gear frequency from the peak intensity of the sideband wave detected by the calculation unit from the spectrum waveform, and the determination unit compares the extracted peak intensity with a peak intensity threshold to perform motor condition diagnosis. This reduces the number of parameters required for diagnosis and shortens the diagnosis time. Furthermore, according to the motor abnormality prediction apparatus of this disclosure, abnormality predictions are predicted from current data before the motor condition diagnosis apparatus detects an abnormality in the motor, and the presence or absence of abnormality predictions is output to a monitoring device, thereby enabling planned motor maintenance. Attached Figure Description

[0012] Figure 1 This is a system overall structure diagram of the motor condition diagnosis device involved in Implementation Method 1. Figure 2 This is a hardware structure diagram of the motor condition diagnosis device according to Implementation Method 1. Figure 3 This is a structural diagram of the arithmetic processing unit of the electric motor condition diagnosis device according to Embodiment 1. Figure 4 This is a diagram showing the connection relationship between the motor, geared part, and load device of the electric motor condition diagnosis device according to Embodiment 1. Figure 5 This is a flowchart illustrating the processing steps during the learning process of the motor condition diagnosis device according to Embodiment 1. Figure 6 This is a flowchart illustrating the processing steps during diagnosis by the condition diagnosis device for the electric motor according to Embodiment 1. Figure 7 This is a flowchart illustrating the frequency analysis processing steps of the motor condition diagnostic device according to Embodiment 1. Figure 8 This is a graph comparing the analysis results of the motor under normal conditions and when abnormalities occur. Figure 9 This is a structural diagram of the computational processing unit of the electric motor state diagnosis device according to Embodiment 2. Figure 10 This is a flowchart illustrating the processing steps during the learning process of the motor condition diagnosis device according to Embodiment 2. Figure 11 This is a flowchart illustrating the processing steps during diagnosis by the motor condition diagnosis device according to Embodiment 2. Figure 12 This is a system overall structure diagram of the motor condition diagnosis device involved in Implementation Method 3. Figure 13This is a system overall structure diagram of the abnormality prediction device for electric motors involved in Implementation Method 4. Figure 14 This is a flowchart illustrating the processing steps during the learning process of the abnormality prediction device for the electric motor according to Embodiment 4. Figure 15 This is a flowchart illustrating the processing steps when the abnormality prediction device for the electric motor according to Embodiment 4 is applied. Detailed Implementation

[0013] The embodiments will now be described in detail with reference to the accompanying drawings. Furthermore, the embodiments described below are examples. Additionally, the various embodiments can be appropriately combined for implementation.

[0014] Implementation method 1. Figure 1 This is a system overall structure diagram of the motor condition diagnosis device involved in Implementation Method 1. Figure 2 This is a hardware structure diagram of the motor condition diagnosis device according to Implementation Method 1. Figure 3 This is a structural diagram of the arithmetic processing unit of the electric motor state diagnosis device according to Embodiment 1. Figure 4 This is a diagram showing the connection relationship between the motor, geared part, and load device of the electric motor condition diagnosis device according to Embodiment 1. Figure 5 This is a flowchart illustrating the processing steps during the learning process of the motor condition diagnosis device according to Embodiment 1. Figure 6 This is a flowchart illustrating the processing steps during diagnosis by the condition diagnosis device for the electric motor according to Embodiment 1. Figure 7 This is a flowchart illustrating the frequency analysis processing steps of the motor condition diagnostic device according to Embodiment 1. Figure 8 This is a graph comparing the analysis results of the motor under normal conditions and when abnormalities occur.

[0015] like Figure 1 As shown, the main circuit 1 is connected to a wiring circuit breaker 2, an electromagnetic contactor 3, a current detector 4, and a motor 5. When a system malfunction occurs, the wiring circuit breaker 2 prevents overcurrent from flowing to the motor 5. The electromagnetic contactor 3 acts as a switch to connect / disconnect the motor 5. The current detector 4 detects the load current of the main circuit 1 and outputs it to the current input section 7 of the status diagnostic device 100. The motor 5 is connected to mechanical equipment 6, which constitutes a production line, etc., and is driven by a power conversion device, etc. (not shown). The motor 5 may be, for example, a three-phase induction motor, but is not limited to this.

[0016] The status diagnostic device 100 includes a current input unit 7, an arithmetic processing unit 8, a rated information storage unit 9, a rated information setting unit 10, a display unit 11, a contactor drive unit 12, an output unit 13, and a communication unit 14. The current input unit 7 inputs the current data detected by the current detector 4 and outputs it to the arithmetic processing unit 8. The arithmetic processing unit 8 analyzes the current data input by the current input unit 7, calculates parameters for diagnosing the motor 5, and performs abnormality diagnosis of the motor 5. Details of the arithmetic processing unit 8 will be described later.

[0017] Figure 2 This is a hardware structure diagram of the status diagnostic device 100 disclosed herein. The status diagnostic device 100 is configured to include a processor 20 such as a CPU (Central Processing Unit), a memory 30 such as RAM (Random Access Memory), a display 40, and an input interface (I / F) 50. The arithmetic processing unit 8 is implemented by the processor 20 executing the program stored in the memory 30. However, for example, it can also be implemented by multiple processors 20 working together.

[0018] The rated information setting unit 10 pre-inputs rated information of the motor 5, such as power frequency, rated output, rated current, number of poles, and rated speed. Rated information refers to basic information recorded in the manufacturer's catalog, instruction manual, or similar documents for the motor 5. In this embodiment, an example is shown where only one motor 5 is being diagnosed; however, when diagnosing multiple motors 5, the rated information of each motor 5 is pre-input into the rated information setting unit 10.

[0019] The rated information storage unit 9 stores the rated information input to the rated information setting unit 10, and outputs the rated information to the arithmetic processing unit 8 as needed. Additionally, Figure 1 An example is shown where a rated information storage unit 9 and a rated information setting unit 10 are respectively provided, but the example is not limited thereto. The rated information storage unit 9 and the rated information setting unit 10 can be provided together to store the rated information of the motor 5 and configured to output the rated information to the arithmetic processing unit 8 when analyzing the current data input from the current input unit 7.

[0020] Furthermore, the rated information storage unit 9 and the rated information setting unit 10 are not limited to a structure included in the status diagnostic device 100, but may also be provided externally. In this case, for example, the rated information of the motor 5 can be stored in an external server or the like, and the rated information can be input to the arithmetic processing unit 8 when diagnosing the motor 5.

[0021] The display unit 11, contactor drive unit 12, output unit 13, and communication unit 14 are connected to the arithmetic processing unit 8. When an abnormality is detected in the arithmetic processing unit 8, the display unit 11 displays the detected current data and shows the abnormal status, alarm, etc. In addition, warnings are not limited to displays; they can also be issued via alarms or other means.

[0022] When an abnormality is detected in the arithmetic processing unit 8, the contactor drive unit 12 outputs a control signal for switching the electromagnetic contactor 3. Furthermore, the output unit 13 outputs abnormal status, alarms, etc., to the production line management department, etc. The communication unit 14 transmits the abnormal status of the motor 5, etc., via a network to a monitoring device 200, such as a PC or tablet terminal installed externally, which acts as the host of the overall status diagnostic device 100. Additionally, the method for transmitting data from the communication unit 14 to the monitoring device 200 can be wired or wireless.

[0023] Next, we will use Figure 3 Explain the structure of the arithmetic processing unit 8. For example... Figure 3 As shown, the processing unit 8 includes a current variation calculation unit 110, an analysis interval determination unit 111, a calculation unit 112, a reference value storage unit 113, a diagnostic result storage unit 114, an analysis unit 120, and a diagnostic unit 130. Furthermore, the analysis unit 120 includes a frequency analysis unit 121, a peak detection calculation unit 122, a rotating frequency band extraction unit 123, a frequency axis conversion unit 124, and an averaging processing unit 125. Furthermore, the diagnostic unit 130 includes an inherent peak extraction unit 131 and a determination unit 132.

[0024] The current variation calculation unit 110 calculates whether there is any change in the current data input from the current input unit 7 and determines whether the current data is in a stable state. Specifically, it performs statistical variation analysis of the current data. For example, it determines whether the motor 5 is in a stable state by using analysis methods such as Mahalanobis distance.

[0025] Furthermore, the current variation calculation unit 110 pre-acquires and stores an interval determination threshold from the motor 5 operating normally, as a threshold for determining whether the current data is in a stable state. For example, current data from multiple motors 5 can be pre-acquired, and the interval determination threshold can be selected within a range smaller than the standard deviation of the acquired current data. Alternatively, current data of the motor 5 being diagnosed can be collected over a certain period, and the interval determination threshold can be set based on the standard deviation of the collected current data. Additionally, the interval determination threshold can be stored in the rating information storage unit 9 along with the rated information of the motor 5, instead of in the current variation calculation unit 110.

[0026] The analysis interval determination unit 111 compares the interval determination threshold pre-calculated by the normally operating motor with the calculation result of the current variation calculation unit 110 to determine the analysis interval. That is, the analysis interval determination unit 111 extracts the interval in a stable state from the current data input by the current input unit 7 to determine the analysis interval executed by the analysis unit 120. Specifically, the interval in a stable state, i.e., the interval where the statistical variation calculated by the current variation calculation unit 110 is below a pre-set threshold, is defined as the analysis interval.

[0027] The analysis unit 120 includes a frequency analysis unit 121, a peak detection and calculation unit 122, a rotating frequency band extraction unit 123, a frequency axis conversion unit 124, and an averaging processing unit 125.

[0028] The frequency analysis unit 121 performs frequency analysis on the current data input by the current input unit 7 within the analysis interval extracted by the analysis interval determination unit 111, and calculates the spectrum waveform as the analysis result. Alternatively, the current data can be analyzed using, for example, current FFT (Fast Fourier Transform) analysis or Discrete Fourier Transform (DFT) analysis.

[0029] The peak detection calculation unit 122 detects sideband waves of multiple power supply frequencies from the spectrum waveform obtained by frequency analysis performed by the frequency analysis unit 121, and calculates the peak intensity of the detected sideband waves. Specifically, the peak detection calculation unit 122 extracts the steep slope inversion portion from the detected power supply frequency sideband waves by performing first-order, second-order, and third-order differential calculations, and calculates the peak intensity. When calculating the peak intensity in this way, even small peak intensities can be detected by performing differential calculations up to the third order.

[0030] The rotational frequency extraction unit 123 calculates the rotational frequency from the rated information of the motor 5, and extracts the peak intensity of the power frequency sideband waves generated at positions where the power supply frequency of the motor 5 deviates from the rotational frequency from the peak intensity of multiple sideband waves detected in the peak detection calculation unit 122. Specifically, the rotational frequency is calculated from the rated rotational speed of the rated information stored in the rated information storage unit 9, and the calculated rotational frequency is used to extract sideband waves with equal peak intensities generated at positions where the power supply frequency deviates from the high-frequency and low-frequency sides corresponding to the rotational frequency.

[0031] Regarding the calculation of the rotation frequency, the rotation frequency extraction unit 123 uses the rated information stored in the rated information storage unit 9 to calculate the rotation frequency according to the following formula (1). In addition, the rotation frequency is set to fn, the power supply frequency is set to f, the number of poles is set to P, and the slip of the motor 5 is set to s.

[0032] (Mathematical Formula 1) fn=(2×f) / P×(1-s) Equation (1)

[0033] As described above, the rotating frequency band extraction unit 123 calculates the rotating frequency using formula (1) and extracts the peak intensity of the sideband wave generated in the rotating frequency band from the peak intensity of the sideband wave calculated in the peak detection calculation unit 122. Here, the rotating frequency band refers to the frequency at a position where the power supply frequency of the motor 5 deviates from the rotating frequency, specifically the positions where the power supply frequency deviates from the rotating frequency to the lower frequency side and the positions where the power supply frequency deviates from the rotating frequency to the higher frequency side.

[0034] The frequency axis conversion unit 124 calculates correction values ​​to align the frequency axes of the spectrum waveforms obtained through frequency analysis, thereby correcting the spectrum waveforms calculated in the frequency analysis unit 121. In the spectrum waveforms obtained by frequency analysis of current data, the rotational frequency deviates depending on the load torque of the motor 5. Therefore, in order to accurately overlap and average multiple spectrum waveforms without causing deviations in the averaging processing unit 125 (described later), the frequency axis conversion unit 124 calculates correction values ​​for each spectrum waveform to correct them and align their frequency axes.

[0035] The correction value calculated by the frequency axis conversion unit 124 will be explained. The slip of the motor 5 varies depending on the load torque of the motor 5, and the rotational frequency fn calculated from equation (1) in the rotational frequency band extraction unit 123 also varies depending on the slip of the motor 5.

[0036] For example, with a power supply frequency f = 60 Hz and a pole number P = 4, the rotational frequency fn is calculated by the following formula (2) in the absence of slip (s = 0).

[0037] (Mathematical Formula 2) fn=(2×60) / 4×(1-0)=30Hz Equation (2)

[0038] On the other hand, when the slip is 2%, the rotational frequency fn is calculated according to the following formula (3) based on the load torque condition of the motor 5.

[0039] (Mathematical Formula 3) fn=(2×60) / 4×(1-0.02)=30.6Hz Equation (3)

[0040] Therefore, since the value of the rotational frequency fn calculated using equation (1) varies depending on the load torque of the motor 5, the frequency axis conversion unit 124 corrects for deviations from the reference based on the no-slip condition, i.e., the unloaded state, taking into account the deviation caused by this slip. Specifically, according to equation (3), when the slip is 2%, a deviation of 0.06 Hz occurs compared to the rotational frequency under the unloaded state. In this case, the frequency axis conversion unit 124 sets the correction value to 0.06 Hz to correct the spectrum waveform. As described above, the frequency axis conversion unit 124 uses the rotational frequency of the motor 5 under the no-slip condition, i.e., the unloaded state, as the reference, and corrects the spectrum waveform by using the difference from the reference as the correction value when the rotational frequency calculated in the rotational frequency band extraction unit 123 deviates from the reference.

[0041] The averaging processing unit 125 performs averaging processing on multiple spectral waveforms corrected by the frequency axis conversion unit 124. In the frequency axis conversion unit 124, a correction value for aligning the rotational frequency is calculated for each spectral waveform and correction is performed. Multiple spectral waveforms aligned by the frequency axis are overlapped for averaging processing, thereby enabling accurate overlapping of multiple spectral waveforms for averaging.

[0042] The arithmetic unit 112 detects multiple peak intensities of the sideband wave of the power supply frequency from the analysis results of the analysis unit 120. Specifically, the arithmetic unit 112 detects multiple peak intensities of the sideband wave of the power supply frequency from the spectrum waveform that has been averaged in the averaging processing unit 125. The output result of the arithmetic unit 112 is output to the inherent peak extraction unit 131 of the diagnostic unit 130.

[0043] The diagnostic unit 130 includes an inherent peak extraction unit 131 and a judgment unit 132.

[0044] Figure 4 This diagram illustrates the transmission of vibration to the geared section 60 when an abnormality occurs in the motor 5. Here, the geared section 60 refers to the belt and gears that connect the motor 5 and the load device 70.

[0045] The inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the natural frequency (hereinafter referred to as the gear frequency) that occurs when an abnormality occurs in the geared section 60 connecting the motor 5 and the load device 70. That is, the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the gear frequency from the peak intensities of multiple sideband waves detected in the calculation unit 112. Furthermore, the gear frequency refers to the natural frequency that occurs when an abnormality occurs in the geared section 60, but it can also be the natural frequency that occurs when an abnormality occurs in the belt or the natural frequency that occurs when an abnormality occurs in the gear.

[0046] The extraction method is explained. The inherent peak extraction unit 131 calculates the gear frequency based on the rated information stored in the rated information storage unit 9, and extracts the sideband waves with the same peak intensity generated at positions that deviate from the gear frequency from the power supply frequency to the high-frequency and low-frequency sides, respectively.

[0047] The gear frequency is calculated using the rated information of the motor 5 stored in the rated information storage unit 9 according to the following formula (4). Furthermore, the gear frequency is set to F, the power supply frequency is set to f, and the number of poles is set to P.

[0048] (Mathematical Expression 4) F = (2 × f) / P × (reduction ratio) Equation (4)

[0049] For example, given that the rated information of motor 5 is power supply frequency f = 60Hz, number of poles P = 4, and reduction ratio = 1:3, the gear frequency F can be calculated from equation (1) as (2×60) / 4×(1 / 3) = 10Hz.

[0050] As described above, in the example above, the inherent peak extraction unit 131 extracts sideband waves with peak intensities of equal magnitude generated at positions 10 Hz away from the power supply frequency on both the high-frequency and low-frequency sides. That is, sideband waves with peak intensities of equal magnitude are extracted from both the high-frequency and low-frequency sides of the power supply frequency.

[0051] Figure 4 This diagram shows the connection relationship between the electric motor 5, the geared unit 60, and the load device 70. (Example) Figure 4 As shown, when the motor 5 malfunctions, its vibration is transmitted to the geared section 60 connected to it. Furthermore, even when a malfunction occurs in the load device 70, its vibration is also transmitted to the geared section 60. Therefore, by extracting the sideband wave caused by the gear frequency of the geared section 60, the inherent peak extraction unit 131 can detect not only malfunctions in the geared section 60, but also malfunctions in the motor 5 and the load device 70 connected to the geared section 60.

[0052] The judgment unit 132 compares the peak intensity of the sideband wave caused by the gear frequency extracted in the inherent peak extraction unit 131 with a peak intensity threshold to diagnose whether the motor 5 is abnormal. This peak intensity threshold is calculated using the current data of the motor 5 operating normally, which is pre-stored in the reference value storage unit 113. The following will use... Figure 7 Describe the specific treatment steps during diagnosis.

[0053] Furthermore, when an abnormality is diagnosed in the judgment unit 132 of the diagnostic unit 130, the arithmetic processing unit 8 outputs information to the display unit 11, the contactor drive unit 12, the output unit 13, and the communication unit 14 to notify of the abnormality, and the diagnostic result is stored in the diagnostic result storage unit 114. In addition, the diagnostic result of the motor 5 stored in the diagnostic result storage unit 114 can be stored in association with the current data before the frequency analysis unit 121 performs frequency analysis.

[0054] The reference value storage unit 113 stores the peak intensity of the sideband wave caused by the gear frequency, which is obtained in advance from the current data of the motor 5 operating normally, as a reference value, i.e., the peak intensity threshold, used when diagnosing the motor 5. Additionally, the peak intensity threshold can be stored together with the rated information in the rated information storage unit 9. This will be discussed later. Figure 5 The specific processing steps will be described in detail.

[0055] Next, we will use Figures 5 to 7 Explain the procedures for learning and diagnosis.

[0056] First, use Figure 5 Explain the steps involved in the learning process. Figure 5 This is a flowchart illustrating the calculation steps of the peak intensity threshold used when comparing with the peak intensity of a sideband wave.

[0057] In step S101, the rated information of the motor 5 operating normally is input to the rated information setting unit 10 and stored in the rated information storage unit 9. Furthermore, the rated information of the motor 5 can also be input and stored in the rated information storage unit 9. Additionally, when there are multiple motors 5 to be diagnosed, the rated information of each motor 5 is input and stored.

[0058] In step S102, the current input unit 7 inputs the current of the motor 5, which is the object of diagnosis, obtained by the current detector 4.

[0059] In step S103, the current variation calculation unit 110 calculates the variation of the current data input to the current input unit 7, i.e., whether the current data is in a stable state. The analysis interval determination unit 111 extracts the intervals in a stable state and determines them as the analysis intervals for frequency analysis. The current variation calculation unit 110 calculates the current data and compares it with the interval determination threshold. If the interval determination threshold is not met, it is determined that the current data is in an unstable state, and the process returns to step S102. If it is determined that the current data is in a stable state, the process proceeds to step S104.

[0060] In step S104, the frequency analysis unit 121 performs frequency analysis on the current data within the analysis range determined to be in a stable state, and calculates the spectral waveform as the analysis result. The spectral waveform, as the analysis result obtained through frequency analysis, is output to the arithmetic unit 112. Furthermore, it will be used later... Figure 1 Describe the details of the frequency analysis in step S104.

[0061] In step S105, the inherent peak extraction unit 131 calculates the gear frequency caused by the geared unit 60 corresponding to the motor from the information of the motor 5 stored in the rated information storage unit 9.

[0062] In step S106, the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the geared frequency from the peak intensity of the sideband wave input by the calculation unit 112. That is, the inherent peak extraction unit 131 uses the geared frequency calculated in step S105 to extract sideband waves with peak intensities of the same degree generated at positions that deviate from the geared frequency from the power supply frequency towards both high and low frequencies.

[0063] In step S107, the reference value storage unit 113 stores the peak intensity of the sideband wave extracted by the inherent peak extraction unit 131 in step S106 as the threshold used for diagnosis.

[0064] Next, use Figure 6 The diagnostic procedures are explained. Figure 6 This is a flowchart illustrating the diagnostic process.

[0065] In step S201, the current input unit 7 inputs the current of the motor 5, which is the object of diagnosis, obtained by the current detector 4.

[0066] In step S202, the current variation calculation unit 110 calculates the variation of the current data input to the current input unit 7, i.e., whether the current data is in a stable state. The analysis interval determination unit 111 extracts the intervals in a stable state and determines them as the analysis intervals for frequency analysis. The current variation calculation unit 110 calculates the current data and compares it with the interval determination threshold. If the interval determination threshold is not met, it is determined that the current data is in an unstable state, and the process returns to step S201. If it is determined that the current data is in a stable state, the process proceeds to step S203.

[0067] In step S203, the frequency analysis unit 121 performs frequency analysis on the current data within the analysis range determined to be in a stable state in step S202, and calculates the spectral waveform as the analysis result. The spectral waveform, obtained as the analysis result from the frequency analysis, is input to the arithmetic unit 112. Furthermore, it will be used later... Figure 7 Describe the details of the frequency analysis.

[0068] In step S204, the inherent peak extraction unit 131 calculates the gear frequency caused by the geared section 60 of the motor 5 being diagnosed from the rated information of the motor 5 stored in the rated information storage unit 9.

[0069] In step S205, the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the geared frequency from the peak intensity of the sideband wave input by the calculation unit 112. That is, the inherent peak extraction unit 131 uses the geared frequency calculated in step S105 to extract sideband waves with peak intensities of the same degree generated at positions that deviate from the geared frequency from the power supply frequency towards both high and low frequencies.

[0070] In step S206, the determination unit 132 compares the peak intensity of the sideband wave caused by the gear frequency extracted in step S205 with the peak intensity threshold stored in the reference value storage unit 113.

[0071] In step S207, if the result of comparing the peak intensity of the sideband wave caused by the gear frequency with the peak intensity threshold stored in the reference value storage unit 113 is that the peak intensity of the sideband wave meets the peak intensity threshold, the determination unit 132 determines that there is an abnormality in the motor 5 and proceeds to step S208. On the other hand, if the peak intensity threshold is not met, the determination unit 132 determines that the motor 5 is normal and returns to step S201.

[0072] In step S208, the judgment unit 132 outputs the diagnostic results to the display unit 11, the contactor drive unit 12, the output unit 13, and the communication unit 14. Furthermore, the diagnostic results are stored in the diagnostic result storage unit 114.

[0073] Next, use Figure 7 Provide details of the frequency analysis for S104 and S203. Figure 7 This is a diagram illustrating the frequency analysis process. Figure 7 The processes shown in S301 to S308 correspond to Figure 5 S101 to S103, Figure 6 The processes S201 to S203. Additionally... Figure 7 S301 is the same as S102 and S201, and S302 is the same as S103 and S202, so the description is omitted.

[0074] In step S303, the frequency analysis unit 121 performs frequency analysis on the current data within the analysis range that is determined to be in a stable state, and calculates the spectral waveform as the analysis result. The frequency analysis unit 121 inputs the calculated spectral waveform to the peak detection calculation unit 122.

[0075] In step S304, the peak detection calculation unit 122 detects multiple peak intensities of the sideband wave of the power supply frequency from the spectrum waveform, which is the result of frequency analysis, input by the frequency analysis unit 121.

[0076] In step S305, the rotating frequency band extraction unit 123 extracts the peak intensity of the sideband waves in the rotating frequency band from the peak intensities of the multiple sideband waves detected in the peak detection calculation unit 122. The rotating frequency band extraction unit 123 extracts the peak intensity of the sideband waves in the rotating frequency band on each of the high-frequency and low-frequency sides of the power supply frequency. The rotation frequency is calculated from the rated rotation speed of the rated information stored in the rated information storage unit 9, and the peak intensity of the sideband waves generated at positions that deviate from the power supply frequency to both sides corresponding to the rotation frequency is extracted.

[0077] In step S306, in order to reduce the influence of the rotational frequency deviation caused by the load torque, the frequency axis conversion unit 124 calculates a correction value for the spectrum waveform and uses the calculated correction value to perform correction to align the frequency axis of the spectrum waveform. This eliminates the rotational frequency deviation caused by the load torque condition.

[0078] In step S307, the averaging processing unit 125 repeats steps S301 to S306 multiple times and collects multiple spectral waveforms obtained from the processes in steps S301 to S306. Furthermore, there is no particular limitation on the number of repetitions, but for example, the number of repetitions can be preset in the rated information storage unit 9, and the preset number of repetitions can be repeated. When the number of repetitions is satisfied, the process proceeds to step S318.

[0079] In step S308, multiple frequency-axis-aligned spectral waveforms collected in step S307 are overlapped for averaging. This eliminates the effects of noise, etc.

[0080] As described above, the condition diagnosis method for the motor 5 includes: a step of inputting current data detected by the current detector 4 from the motor 5; a step of performing frequency analysis on the input current data and calculating the analysis result; a step of detecting multiple peak intensities of the sideband wave of the power supply frequency of the motor 5 from the analysis result; a step of calculating the natural frequency of the geared section 60 from the rated information of the motor, and extracting the peak intensity of the sideband wave generated at a position corresponding to the natural frequency based on the power supply frequency of the motor 5 from the detected multiple peak intensities; and a step of diagnosing the motor 5 by comparing the peak intensity threshold calculated from the motor 5 operating normally with the peak intensity. Therefore, the condition diagnosis of the motor 5 can be performed using a single peak intensity of the sideband wave caused by the geared frequency as a parameter, without the need to calculate multiple parameters.

[0081] Figure 8 This graph compares the frequency analysis results of motor 5 under normal and abnormal conditions. Solid lines represent the analysis results when motor 5 is operating normally, while dashed lines represent the analysis results when motor 5 malfunctions. Figure 8 As shown, when the motor 5 malfunctions, the peak intensity of the sideband wave caused by the gear frequency increases compared to the normal state. Therefore, the motor 5 state diagnosis device 100 according to this embodiment can use a threshold value of the peak intensity of the sideband wave caused by the gear frequency, which is obtained in advance from the motor 5 operating normally, as a threshold value for determining whether the motor 5 is malfunctioning. If the peak intensity threshold value is met, it is determined that the motor 5 is malfunctioning, thereby diagnosing the malfunction of the motor 5.

[0082] As described above, the condition diagnosis apparatus 100 for the electric motor 5 according to this embodiment performs frequency analysis on the current data acquired from the electric motor 5, and uses the peak intensity of the sideband wave caused by the geared frequency extracted from the spectrum waveform as an analysis result by the inherent peak extraction unit 131 to diagnose the condition of the electric motor 5. Therefore, one of the peak intensities of the sideband wave caused by the geared frequency can be used as a parameter for diagnosis to perform condition diagnosis of the electric motor 5. Thus, the number of parameters required for diagnosis can be reduced, and the time spent on diagnosis can be shortened.

[0083] Furthermore, in this embodiment, the analysis unit 120 is described as including a peak detection calculation unit 122, a rotating frequency band extraction unit 123, a frequency axis conversion unit 124, and an averaging processing unit 125, but this embodiment is not limited to this. The frequency analysis unit 121 can perform frequency analysis on the current data input from the current input unit 7, calculate the spectrum waveform as the analysis result, and output the calculated spectrum waveform to the calculation unit 112. In this case, the analysis unit 120 may include the frequency analysis unit 121.

[0084] Furthermore, it has been explained that the state diagnostic device 100 of this embodiment diagnoses an anomaly when the peak intensity of the sideband wave extracted by the inherent peak extraction unit 131 meets the peak intensity threshold. However, the method for determining whether the peak intensity threshold is met is not limited. For example, an anomaly can be determined when the time for meeting the peak intensity threshold exceeds a certain period of time. Alternatively, a set number of times can be set, and an anomaly can be determined when the number of times the peak intensity threshold is met exceeds the set number.

[0085] Implementation method 2. This implementation method uses Figures 9 to 11 Please provide an explanation. Figure 9 This is a structural diagram of the arithmetic processing unit 80 of the electric motor condition diagnosis device according to this embodiment. Figure 10 This is a flowchart illustrating the processing steps during the learning process of the motor condition diagnosis device according to this embodiment. Figure 11 This is a flowchart illustrating the processing steps during diagnosis by the condition diagnosis device for the electric motor according to this embodiment.

[0086] In Embodiment 1, a condition diagnosis device 100 is shown, wherein the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave generated at positions deviating from the power supply frequency towards both low and high frequencies from the spectral waveform calculated by the analysis unit 120. The diagnosis unit 130 compares the extracted peak intensity of the sideband wave with a peak intensity threshold to perform condition diagnosis of the motor 5. In this embodiment, a condition diagnosis device 101 for the motor 5 is shown, wherein the calculation unit 140 includes an OA value calculation unit 141, which calculates the overall value (hereinafter referred to as "OA value") used when performing condition diagnosis of the motor 5. The structure is otherwise the same as in Embodiment 1, and the same reference numerals are used for structures identical to those in Embodiment 1, and descriptions are omitted.

[0087] like Figure 9As shown, the processing unit 80 of the condition diagnostic device 101 includes a current variation calculation unit 110, an analysis interval determination unit 111, a reference value storage unit 113, a diagnostic result storage unit 114, an analysis unit 120, a diagnostic unit 130, and a calculation unit 140. With this structure, the diagnostic unit 130 can perform condition diagnosis of the motor 5 by comparing the peak intensity of the sideband wave with a peak intensity threshold, thereby reducing the number of parameters required for condition diagnosis and shortening the diagnosis time.

[0088] Furthermore, the arithmetic unit 140 of the arithmetic processing unit 80 involved in this embodiment includes an OA value calculation unit 141.

[0089] The OA value calculation unit 141 calculates the average amplitude, i.e., the OA value, from the spectrum waveform, which is the analysis result calculated by the analysis unit 120, as information for determining the abnormality of the motor 5. The OA value calculated during learning is stored in the reference value storage unit 113, and the OA value calculated during diagnosis is output to the judgment unit 133.

[0090] Furthermore, the OA value refers to the average amplitude of the entire frequency spectrum waveform obtained through frequency analysis of current data. However, the signal strength in the power supply frequency and its harmonic components accounts for a large proportion of the calculated OA value. Therefore, the peak values ​​of higher-order frequency components in the power supply frequency and integer multiples of the power supply frequency, i.e., the harmonic components of the power supply frequency, can be excluded from the OA value calculated by the OA value calculation unit 141. In this way, by excluding values ​​that account for a large proportion of the OA value, misdiagnosis can be prevented.

[0091] Next, we will use Figure 10 and Figure 11 Explain the procedures for learning and diagnosis. Additionally, Figure 10 In the middle, to and Figure 5 The same processed content is labeled with the same number. Similarly, Figure 11 In the middle, to and Figure 6 The same processing content is labeled with the same number. Therefore, the following will explain... Figure 10 Steps S107 and S108 Figure 11 Steps S207, S209, and S210 are omitted, and other descriptions are omitted.

[0092] First, the processing steps during learning will be explained. In step S108, the OA value calculation unit 141 calculates the OA value based on the spectrum waveform, which is the analysis result output from the analysis unit 120. The OA value calculated in the OA value calculation unit 141 is output to the reference value storage unit 113.

[0093] In step S107, the reference value storage unit 113 stores the OA value output from the OA value calculation unit 141 as a reference value used during diagnosis, i.e., the OA threshold. Furthermore, the reference value storage unit 113 stores the peak intensity of the sideband wave caused by the geared frequency extracted by the inherent peak extraction unit 131 in S106 as a reference value used during diagnosis, i.e., the peak intensity threshold. As described above, in step S107, the OA threshold and the peak intensity threshold are saved as learning data to the reference value storage unit 113. Alternatively, the OA threshold and the peak intensity threshold can be stored in the nominal information storage unit 9.

[0094] Next, the processing steps during diagnosis will be explained. In step S209, the OA value calculation unit 141 calculates the OA value based on the spectrum waveform of the analysis result output from the analysis unit 120. The calculated OA value is output from the OA value calculation unit 141 to the judgment unit 133 of the diagnosis unit 130.

[0095] In step S210, the determination unit 133 compares the OA value calculated by the OA value calculation unit 141 in step S209 with the OA threshold stored in the reference value storage unit 113.

[0096] In step S207, if it is determined in step S210 that the OA value meets the OA threshold and in step S206 that the peak intensity meets the peak intensity threshold, the determination unit 133 determines that there is an abnormality in the motor 5 and proceeds to step S208. Alternatively, if either threshold is not met, the motor 5 is determined to be normal, and the process returns to step S201.

[0097] As described above, the condition diagnosis method for the motor 5 includes: a step of inputting current data detected by the current detector 4 from the motor 5; a step of performing frequency analysis on the input current data and calculating the analysis result; a step of detecting multiple peak intensities of the sideband wave of the power supply frequency of the motor 5 from the analysis result; a step of calculating the natural frequency of the geared section 60 from the rated information of the motor 5, and extracting the peak intensities of the sideband wave generated at a position that deviates from the natural frequency based on the power supply frequency of the motor 5 from the detected multiple peak intensities; a step of calculating the OA value from the analysis result; and a step of comparing the peak intensity threshold calculated from the motor 5 operating normally with the peak intensity, and comparing the OA threshold calculated from the motor 5 operating normally with the OA value, to diagnose the motor 5. Therefore, compared with the condition diagnosis method for the motor 5 according to Embodiment 1, the accuracy of the diagnosis result can be improved.

[0098] As described above, in the condition diagnosis device 101 for the electric motor 5 according to this embodiment, the calculation unit 140 includes an OA value calculation unit 141, and the judgment unit 133 uses the OA value calculated in the OA value calculation unit 141 and the peak intensity of the sideband wave caused by the gear frequency extracted in the inherent peak extraction unit 131 to perform condition diagnosis of the electric motor 5. This reduces the number of parameters required for diagnosis and shortens the diagnosis time. Furthermore, compared to the condition diagnosis method for the electric motor 5 according to Embodiment 1, the accuracy of the diagnosis results can be improved.

[0099] Implementation method 3. will use Figure 12 This implementation method is described below. Figure 12 This is a system overall structure diagram of the motor condition diagnosis device involved in this embodiment.

[0100] In this embodiment, a state diagnosis device 102 for the electric motor 5 is shown. Based on the state diagnosis device 100 for the electric motor 5 described in Embodiment 1, it further includes a learning unit 15, a model storage unit 16, and a prediction unit 17. The structure is otherwise the same as in Embodiment 1, and the same reference numerals are used for the structures identical to those in Embodiment 1; descriptions are omitted.

[0101] like Figure 11 As shown, the status diagnostic device 102 includes, for example, a current input unit 7, an arithmetic processing unit 81, a rated information storage unit 9, a rated information setting unit 10, a display unit 11, a contactor drive unit 12, an output unit 13, a communication unit 14, a learning unit 15, a model storage unit 16, and a prediction unit 17.

[0102] In the condition diagnostic device 102, the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the gear frequency from the peak intensity of the sideband wave detected by the calculation unit 112 from the spectrum waveform, and compares the extracted peak intensity with a peak intensity threshold to perform condition diagnosis of the motor 5. Therefore, since the peak intensity of the sideband wave caused by the gear frequency can be used as a parameter for diagnosing abnormalities in the motor 5, the number of parameters required for diagnosis can be reduced, and the diagnosis time can be shortened.

[0103] The learning unit 15 acquires the current data obtained from the current input unit 7 and the diagnostic results of the motor 5 stored in the diagnostic result storage unit 114, and performs learning by associating the diagnostic results with the current data to generate a complete learning model. That is, the diagnostic results stored in the diagnostic result storage unit 114 are stored in association with the current data before frequency analysis, and the learning unit 15 learns the time series patterns representing abnormal symptoms from the time series data of the current data obtained from the current input unit 7 based on the diagnostic results to generate a complete learning model.

[0104] Furthermore, the learning unit 15 uses the following method when generating the completed learning model: it uses deep learning to infer the time series pattern of current data, which is a common time series pattern during abnormal periods and is not included in the time series pattern during normal periods, from the current data of a certain period before the determination of the motor 5 as abnormal. In addition, known machine learning methods such as genetic programming, inductive logic programming, and support vector machines can be used.

[0105] The model storage unit 16 stores the completed learning model generated by the learning unit 15. Furthermore, the stored completed learning model is not limited to the one generated by the learning unit 15; for example, pre-learned data can be read from an external source and stored as a completed learning model. Moreover, the model storage unit 16 is not limited to being located in the state diagnostic device 102; it can also be located in an external server or the like.

[0106] The inference unit 17 uses the completed learning model stored in the model storage unit 16 to infer abnormal signs of the motor 5 from the newly acquired current data from the current input unit 7, and outputs the inference result to the monitoring device 200. That is, it acquires current data from the current input unit 7, uses the completed learning model stored in the model storage unit 16 to infer whether the current data is a time series pattern that is detected as a symptom of an abnormality in the motor 5, and when an abnormal symptom is detected, it determines that there is an abnormality in the motor 5 and outputs the inference result to the monitoring device 200.

[0107] Furthermore, the prediction results are not limited to being output to the monitoring device 200, but can be output to the display unit 11, the contactor drive unit 12, the output unit 13, and the communication unit 14 via the arithmetic processing unit 8. Thus, by detecting abnormal signs of the motor 5 based on the current data, maintenance of the motor 5 can be performed in a planned manner, and the stopping period of the mechanical equipment 6 connected to the motor 5 can be adjusted to minimize the stopping period.

[0108] As described above, the condition diagnosis device 102 for the electric motor 5 according to this embodiment further includes a learning unit 15, a model storage unit 16, and an inference unit 17. The learning unit 15 generates a completed learning model from the current data obtained based on the analysis results, and the inference unit 17 uses the completed learning model to infer abnormalities of the electric motor 5 from the current data. According to this structure, since abnormalities of the electric motor 5 can be predicted based on the current data, the electric motor 5 can be used reliably, and maintenance plans can be executed without unnecessary equipment downtime.

[0109] In addition, in this embodiment, a state diagnostic device 102 is described that further includes a learning unit 15, a model storage unit 16, and a prediction unit 17 on the basis of the state diagnostic device 100 of embodiment 1. However, it may also be configured to further include a learning unit 15, a model storage unit 16, and a prediction unit 17 on the basis of the state diagnostic device 101 of embodiment 2.

[0110] Implementation method 4. This implementation method uses Figures 13 to 15 Please provide an explanation. Figure 13 This is a system overall structure diagram of the abnormal motor prediction device involved in this embodiment. Figure 14 This is a flowchart illustrating the processing steps of the motor abnormality prediction device according to this embodiment during learning. Figure 15 This is a flowchart illustrating the processing steps when the abnormality prediction device for an electric motor according to this embodiment is applied.

[0111] In this embodiment, an abnormality prediction device 300 for motor 5, externally connected to the condition diagnosis device 100 for motor 5, is shown. The structure is otherwise the same as in Embodiment 1; structures identical to those in Embodiment 1 are labeled with the same reference numerals, and descriptions are omitted.

[0112] like Figure 13 As shown, the condition diagnostic device 100 includes, for example, a current input unit 7, an arithmetic processing unit 8, a rated information storage unit 9, a rated information setting unit 10, a display unit 11, a contactor drive unit 12, an output unit 13, and a communication unit 14. In the condition diagnostic device 100, the inherent peak extraction unit 131 extracts the peak intensity of the sideband wave caused by the gear frequency from the peak intensity of the sideband wave detected from the spectrum waveform by the arithmetic unit 112, and compares the extracted peak intensity with a peak intensity threshold to perform condition diagnosis of the motor 5. Therefore, since the peak intensity of the sideband wave caused by the gear frequency can be used as a parameter for diagnosing abnormalities in the motor 5, the number of parameters required for diagnosis can be reduced, and the diagnosis time can be shortened.

[0113] In addition, the abnormality prediction device 300 for the electric motor 5 includes a data acquisition unit 310, a completed learning model generation unit 320, a completed learning model storage unit 330, and a current data prediction unit 340.

[0114] The data acquisition unit 310 acquires current data from the current input unit 7 of the status diagnostic device 100, and acquires the diagnostic results of the motor 5 associated with the current data before frequency analysis from the diagnostic result storage unit 114 of the arithmetic processing unit 8.

[0115] The learning model generation unit 320 learns from the diagnostic results using the current data acquired by the data acquisition unit 310 to generate a learning model. Specifically, based on the diagnostic results, it learns time-series patterns representing abnormal symptoms from the time-series data of the current data acquired by the current input unit 7 to generate a learning model. Furthermore, there is no particular limitation on the amount of data used to learn the learning model; however, for example, the amount of data to be learned can be preset in the learning model generation unit 320, and learning can end when the preset amount of data has been learned.

[0116] The learning process used to generate the completed learning model employs the following method: Deep learning is used to infer a common time-series pattern of current data during abnormal periods—a pattern not included in the normal time-series pattern—from current data over a certain period prior to the determination of an anomaly in motor 5. Alternatively, machine learning methods such as genetic programming, inductive logic programming, and support vector machines can be used, not limited to deep learning.

[0117] The completed learning model storage unit 330 stores the completed learning model generated in the completed learning model generation unit 320. Furthermore, the completed learning model to be stored is not limited to the completed learning model generated by the completed learning model generation unit 320; for example, completed learning data can be read from an external source and stored as a completed learning model. Moreover, the completed learning model storage unit 330 is not limited to being configured in the anomaly prediction device 300; it can also be installed, for example, in an external server.

[0118] The current data estimation unit 340 uses a completed learning model to estimate abnormal signs of the motor 5 based on the newly acquired current data from the data acquisition unit 310, and outputs the estimation result to the monitoring device 200. Specifically, the data acquisition unit 310 acquires current data from the current input unit 7, estimates whether the current data acquired using the completed learning model stored in the completed learning model storage unit 330 is a time series pattern indicating an abnormality in the motor 5, and when an abnormal sign is detected, determines that the motor 5 is abnormal, and outputs the estimation result to the monitoring device 200.

[0119] Furthermore, the prediction results are not limited to being output to the monitoring device 200, but can also be output to the display unit 11, the contactor drive unit 12, the output unit 13, and the communication unit 14 via the arithmetic processing unit 8. Thus, abnormal signs of the motor 5 can be detected based on the current data, allowing for planned maintenance of the motor 5 and adjustment of the stopping period of the mechanical equipment 6 connected to the motor 5 to minimize the stopping time.

[0120] Next, we will use Figure 14 Explain the learning steps. Figure 14This is a flowchart illustrating the learning phases.

[0121] In step S301, the data acquisition unit 310 acquires current data from the current input unit 7 and acquires diagnostic results associated with the current data prior to the frequency analysis from the diagnostic result storage unit 114 of the arithmetic processing unit 8.

[0122] In step S302, the learning model generation unit 320 learns the abnormal signs of the motor 5 based on the current data output from the data acquisition unit 310 and the diagnostic results. That is, it learns the time series pattern of the common current data when the abnormality is diagnosed from the current data of the motor 5 before the current input unit 7 diagnoses the abnormality, based on the diagnostic results, and generates a learning model.

[0123] In step S303, the completed learning model storage unit 330 stores the completed learning model generated in step S302.

[0124] Next, we will use Figure 15 Explain the application steps. Figure 15 This is a flowchart illustrating the application phase.

[0125] In step S401, the data acquisition unit 310 acquires current data from the current input unit 7.

[0126] In step S402, the current data acquired by the data acquisition unit 310 in step S401 is input into the completed learning model stored in the completed learning model storage unit 330.

[0127] In step S403, the current data estimation unit 340 outputs the estimation result of the current data input to the completed learning model, that is, the result obtained by estimating whether there are any abnormal signs in the motor 5 based on the current data obtained from the data acquisition unit 310.

[0128] In step S404, the prediction result output from the current data prediction unit 340 is output to the monitoring device 200. In addition, the prediction result is not limited to being output to the monitoring device 200, but can also be output to the display unit 11, the contactor drive unit 12, the output unit 13 and the communication unit 14 via the arithmetic processing unit 8.

[0129] As described above, the abnormality prediction device 300 for the motor 5 according to this embodiment learns from the current data acquired from the condition diagnosis device 100 based on the diagnostic results, generates a complete learning model, and inputs newly acquired current data into the complete learning model to predict abnormalities in the motor 5. Thus, before the condition diagnosis device 100 of the motor 5 diagnoses an abnormality, abnormalities are predicted from the current data, and the presence or absence of abnormalities is output to the monitoring device 200, thereby enabling planned maintenance of the motor 5. Furthermore, by pre-planning maintenance, unnecessary downtime of the mechanical equipment 6 connected to the motor 5 can be shortened.

[0130] In addition, in this embodiment, the system is configured to generate a completion learning model based on the current data and diagnostic results obtained by the state diagnostic device 100 of embodiment 1, and input the newly acquired current data into the completion learning model to predict abnormal signs of the motor 5. However, it can also be configured to generate a completion learning model based on the current data and diagnostic results obtained by the state diagnostic device 101 of embodiment 2, and input the newly acquired current data into the completion learning model to predict abnormal signs of the motor 5. Label Explanation

[0131] 1 Main circuit, 2 Circuit breaker for wiring, 3 Electromagnetic contactor, 4 Current detector, 5 Motor, 6 Mechanical equipment, 7 Current input unit, 8, 80 Calculation and processing unit, 9 Rated information storage unit, 10 Rated information setting unit, 11 Display unit, 12 Contactor drive unit, 13 Output unit, 14 Communication unit, 15 Learning unit, 16 Model storage unit, 17 Prediction unit, 20 Processor, 30 Memory, 40 Display, 50 Input interface, 60 Geared unit, 70 Load device, 100, 101 Status diagnostic device, 110 Current variation calculation unit, 111 Analysis interval determination unit 112, 140 Calculation Unit, 113 Reference Value Storage Unit, 114 Diagnostic Result Storage Unit, 120 Analysis Unit, 121 Frequency Analysis Unit, 122 Peak Detection Calculation Unit, 123 Rotating Frequency Band Extraction Unit, 124 Frequency Axis Conversion Unit, 125 Averaging Processing Unit, 130 Diagnostic Unit, 131 Inherent Peak Extraction Unit, 132, 133 Judgment Unit, 141 OA Value Calculation Unit, 200 Monitoring Device, 300 Abnormal Prediction Device, 310 Data Acquisition Unit, 320 Completed Learning Model Generation Unit, 330 Completed Learning Model Storage Unit, 340 Current Data Inference Unit.

Claims

1. A motor status diagnosis device, characterized in that: include: a current input section that inputs current data detected by the current detector from the motor; an analysis unit that performs frequency analysis on the current data input from the current input unit and calculates an analysis result; a calculation unit that detects a plurality of peak intensities of a sideband wave of a power supply frequency of the motor from the analysis result; a natural peak extraction unit that calculates the natural frequency of the geared portion from the rated information of the motor and extracts the peak intensity of the sideband wave generated at a position where a deviation corresponding to the natural frequency occurs with respect to the power supply frequency of the motor from the plurality of peak intensities detected in the calculation unit; as well as A determination unit compares a peak intensity threshold value calculated from the motor operating normally with the peak intensity extracted by the unique peak extraction unit to diagnose the motor.

2. The motor status diagnosis device according to claim 1, characterized in that: The analysis unit comprises: a frequency analysis unit that performs the frequency analysis of the current data; a peak detection operation unit that detects a plurality of peak intensities of sideband waves of the power supply frequency of the motor from a spectrum waveform that is a result of the frequency analysis by the frequency analysis unit; a rotation frequency band extraction unit that calculates a rotation frequency from rated information of the motor and extracts, from the plurality of peak intensities detected in the peak detection operation unit, the peak intensity of the sideband wave generated at a position where a deviation corresponding to the rotation frequency occurs with respect to the power supply frequency of the motor; a frequency axis conversion unit that calculates a correction value for aligning the frequency axes of the spectrum waveforms and corrects the spectrum waveforms respectively; and an averaging processing unit that performs averaging processing on the plurality of spectrum waveforms corrected by the frequency axis conversion unit, The analysis unit calculates the analysis result.

3. The motor status diagnosis device according to claim 1 or 2, characterized in that: The analyzing unit performs the frequency analysis in an analysis period in which it is determined that the current data is in a stable state.

4. The motor status diagnosis device according to claim 3, characterized in that: The analyzing unit performs the frequency analysis of the current data in the analysis section extracted by the current variation calculating unit and the analysis section determining unit, The current variation calculation unit calculates whether the current data is in the stable state. The analysis section determination unit makes a determination by comparing a section determination threshold value calculated in advance from the motor operating normally with the calculation result of the current fluctuation calculation unit.

5. The motor status diagnosis device according to any one of claims 1 to 4, characterized in that: The calculation unit includes an OA value calculation unit that calculates an OA value that is an average value of amplitudes from the analysis result calculated by the analysis unit.

6. The motor status diagnosis device according to claim 5, characterized in that: The determination unit compares the OA threshold value calculated from the motor that is operating normally with the OA value calculated by the OA value calculation unit to diagnose the motor.

7. The motor status diagnosis device according to any one of claims 1 to 6, characterized in that: include: a learning unit that acquires the current data acquired by the current input unit and a diagnosis result of the motor, associates the diagnosis result with the current data to perform learning, and generates a completed learning model; as well as An estimation unit estimates a sign of abnormality of the motor based on the current data newly acquired by the current input unit using the completed learning model, and outputs an estimation result.

8. A device for predicting abnormality signs of a motor, characterized in that: include: a data acquisition unit that acquires current data and a diagnosis result of the motor associated with the current data before frequency analysis from the motor state diagnosis device according to any one of claims 1 to 6; a completed learning model generating unit that learns the current data acquired by the data acquiring unit based on the diagnosis result to generate a completed learning model; as well as A current data estimation unit estimates a sign of abnormality of the motor based on the current data newly acquired by the data acquisition unit using the completed learning model, and outputs an estimation result.

9. A method for diagnosing the state of an electric motor, characterized in that: include: The step of inputting current data detected by the current detector from the motor; A step of performing frequency analysis on the input current data and calculating the analysis result; A step of detecting a plurality of peak intensities of a sideband wave of a power frequency of the motor from the analysis result; The steps of calculating the natural frequency of the gear portion from the rated information of the motor, and extracting the peak intensity of the sideband wave generated at a position where a deviation corresponding to the natural frequency occurs with respect to the power supply frequency of the motor from the detected plurality of peak intensities; and A step of comparing a peak intensity threshold value calculated from the motor operating normally with the peak intensity to diagnose the motor.

Images