Motor deterioration diagnosis method and motor deterioration diagnosis system
The method uses pulse voltage and frequency analysis to diagnose inverter-driven motor deterioration in real-time, addressing the limitations of conventional technologies by accurately identifying specific parts' degradation states.
Patent Information
- Application Number
- PCT/JP2025/032324
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-08
- Filing Date
- 2025-09-12
- Publication Date
- 2026-11-12
AI Technical Summary
Conventional methods for diagnosing inverter-driven motor deterioration are limited to evaluating the entire motor and do not support real-time diagnosis of specific parts like windings or bearings, making it difficult to detect abnormalities during operation.
A method that uses pulse voltage from an inverter to generate frequency bands based on the motor's resonant frequency, analyzes the response current for impedance characteristics, and diagnoses the deterioration state of specific parts like windings or bearings in real-time.
Enables real-time detection of motor abnormalities, improving diagnostic accuracy and reducing equipment downtime by identifying specific parts' degradation states, such as windings or bearings, while the motor is operating.
Smart Images

Figure JP2025032324_12112026_PF_FP_ABST
Abstract
Description
Motor Deterioration Diagnosis Method and Motor Deterioration Diagnosis System
[0001] The present disclosure relates to a motor deterioration diagnosis method and a motor deterioration diagnosis system.
[0002] Inverter-driven motors are widely used in industrial equipment or facilities from the perspective of energy saving and higher efficiency, and their stable operation and maintainability have become important management themes. In order to realize these themes and improve the stable operation and maintainability of inverter-driven motors, early detection of the deterioration state of the inverter-driven motor in use is of great importance.
[0003] In general, main deterioration factors of a motor include dielectric breakdown of windings or wear of bearings, and the progression of these factors leads to failure or performance degradation. Among these, for example, regarding the dielectric breakdown of windings, methods for evaluating, inspecting, or diagnosing the insulating performance of motor insulating materials under sinusoidal voltage application have been established.
[0004] Here, regarding methods for evaluating, inspecting, or diagnosing insulating performance against pulse voltages used in inverter-driven motors, for example, there is a known practice in which a pulse test voltage from an inverter is applied to a sample insulating material, and the insulating performance is evaluated based on the insulation lifetime until the dielectric breakdown of the sample insulating material, so as to diagnose the insulation deterioration performance of the motor based on the diagnosis result of a representative component of the motor. However, the above evaluation, inspection, or diagnosis methods are limited to the insulating material of the motor, and do not support the diagnosis of the deterioration state of windings and bearings, or deterioration diagnosis during operation (for example, refer to Patent Document 1).
[0005] On the other hand, in actual inverter-driven motors, main deterioration factors of the motor include dielectric breakdown of windings or wear of bearings, and the progression of these factors leads to failure or performance degradation. In the conventional technology, the diagnosis function of an inverter is limited to the deterioration evaluation of the inverter itself or the entire motor, and it has been considered difficult to evaluate the deterioration state of a specific part in real time while the motor is in operation.
[0006] JP 2007-232517 A JP 2020-20611 A
[0007] B. Mirafzal, et. al. , “Universal Induction Motor Model With Low-to-High Frequency-Response Characteristics”, IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 43, No. 5, SEPTEMBER / OCTOBER, 2007, pp. 1233-1246
[0008] As explained above, in the degradation diagnosis of inverter-driven motors, it is becoming necessary to establish more detailed degradation diagnosis techniques, such as real-time diagnosis during operation and identification of specific deteriorated parts, in addition to diagnosing the degradation of the entire motor.
[0009] This disclosure discloses technology to solve the above-mentioned problems, and aims to establish a motor degradation diagnosis method that uses pulse voltage from an inverter applied to the motor to instantly acquire and analyze the response current during motor operation, immediately issue an alert in the event of an abnormality, and grasp the degradation state of specific parts such as the motor windings or bearings supporting the rotating shaft in real time.
[0010] The method for diagnosing the deterioration of an inverter-driven motor disclosed herein is characterized by: generating a pulse voltage in a frequency band set based on a pre-given resonant frequency of the motor under test, or a resonant frequency identified by impedance measurement; driving the motor with the generated pulse voltage; acquiring the motor's response current obtained by applying the pulse voltage to the motor in real time; determining the impedance characteristics of the motor from the diagnostic pulse voltage output from the inverter, which serves as a diagnostic reference, and the acquired motor response current; and diagnosing the deterioration state of the motor based on the motor's impedance characteristics.
[0011] According to the inverter-driven motor degradation diagnosis method disclosed herein, it is possible to establish a motor degradation diagnosis method that uses pulse voltage from an inverter applied to the motor to immediately acquire and analyze the response current during motor operation, immediately issue an alert in the event of an abnormality, and grasp the degradation state of specific parts such as the motor windings or bearings supporting the rotating shaft in real time.
[0012] This is a block diagram showing an example of a motor degradation diagnosis device for explaining the motor degradation diagnosis method according to Embodiment 1. This is a diagram showing some of the components in detail in Figure 1. This is a block diagram showing an example of a motor degradation diagnosis device for explaining the motor degradation diagnosis method according to Embodiment 3. This is a block diagram showing an example of a motor degradation diagnosis device for explaining the motor degradation diagnosis method according to Embodiment 12. This is a diagram for explaining the outline of the motor degradation diagnosis method using AI according to Embodiments 12-14. This is a flowchart diagram for explaining the motor degradation diagnosis method using AI according to Embodiments 12-14. This is an overview diagram for explaining the motor degradation diagnosis system according to Embodiments 1-14. This is a diagram showing an example of the hardware of the main components constituting the motor degradation diagnosis device according to Embodiments 1-14.
[0013] The following will explain in detail, using diagrams, the method for diagnosing the deterioration of an inverter-driven motor according to this disclosure, in comparison with the prior art. In each diagram, the same reference numerals indicate the same or corresponding parts.
[0014] Embodiment 1. <Basic Configuration> The motor degradation diagnosis method according to Embodiment 1 involves applying a pulse waveform generated from an inverter device to the motor, acquiring the motor's response current in real time, and analyzing it to diagnose the motor's degradation state. The motor degradation diagnosis method according to Embodiment 1 will be described in detail below with reference to the figures.
[0015] The motor degradation diagnostic device 100 used in the motor degradation diagnostic method of this embodiment 1 consists of the following elements shown in the block diagram within the dotted line frame in Figures 1 and 2 (the same applies to the motor degradation diagnostic devices of other embodiments which will be described later). Specifically, the pulse voltage generation command unit 10 consists of an inverter power supply unit 1 that supplies power to the inverter control unit 2, and an inverter control unit 2 that outputs commands for generating pulse voltages for motor driving and diagnosis; a pulse voltage generation unit 3 that generates a set pulse waveform based on commands from the pulse voltage generation command unit 10 and applies it to the motor 4 (hereinafter also referred to as the object under test 4); a motor 4 that generates a response current corresponding to the applied pulse voltage (hereinafter also referred to as pulse wave or pulse waveform); an ammeter side unit 5 that measures the response current; a signal processing unit 6 that analyzes the response current measured by the ammeter side unit 5 by referring to the diagnostic pulse wave from the pulse voltage generation unit 3 and calculates the impedance characteristics; a storage unit 7 that stores the calculated impedance information; an abnormality determination unit 8 that analyzes the impedance information and determines the deterioration state of the motor by comparing it with past data; and a display unit 9 that outputs the determined result as a diagnosis result, presents it visually to the operator, and notifies the operator of abnormalities as necessary.
[0016] The motor degradation diagnosis device 100, composed of the multiple elements described above, performs degradation diagnosis of the motor to be diagnosed according to the arrows in the block diagram of the degradation diagnosis device shown in Figures 1 and 2, and diagnoses the motor's degradation state in real time. In particular, a feature of the motor degradation diagnosis method of this disclosure is that the motor degradation state is diagnosed from the impedance calculated by the signal processing unit 6 based on the response current of the motor driven by the drive pulse voltage and the diagnostic pulse voltage, using the motor drive pulse (hereinafter also simply called the drive pulse or drive pulse voltage) and the diagnostic pulse (hereinafter also called the diagnostic pulse voltage) generated by the pulse voltage generation unit 3. The results of the diagnosis are output to the display unit 9 and presented visually to the operator, and abnormality notifications are made as necessary. Note that waveform shaping is used as the diagnostic pulse (details will be explained later). As described above, Figures 1 and 2 clearly show the flow of information for each component and illustrate the diagnosis process based on the signal flow.
[0017] <Diagnostic Pulse> As mentioned earlier, a waveform shaped for use as a diagnostic pulse is used. The details are explained below. When the pulse waveform output from the pulse voltage generation unit is used for diagnosis, it is necessary that at least the harmonic components or rise characteristics meet certain conditions. In particular, since the effective bandwidth of the signal must be sufficiently secured in analysis, it is desirable that the component includes frequencies up to 1.5 times or more the upper limit of the frequency of the measurement target. The reason for this is to ensure a spectrum that is well above the Nyquist frequency (the upper limit of reproducible frequencies based on the sampling theorem) and to prevent waveform distortion or loss of information.
[0018] The pulse waveform output from the pulse voltage generation unit does not necessarily have to be an ideal rectangular waveform, but for diagnostic purposes, it is desirable to use it after applying waveform shaping (for example, controlling the pulse width or rise time). For example, pulse shaping using PWM control can be employed.
[0019] In the motor degradation diagnosis method of Embodiment 1, the pulse voltage generation command unit has information on the motor's resonant frequency and sends a command to the pulse voltage generation unit to generate a pulse waveform in a low-frequency band or a high-frequency band set with the resonant frequency as the boundary. The pulse voltage generation unit then generates a pulse waveform in the commanded frequency band. The response current generated by the pulse wave applied to the motor is acquired in real time by the ammeter unit and transmitted to the signal processing unit. The signal processing unit performs harmonic analysis or impedance analysis to generate data for evaluating the motor's degradation state.
[0020] Next, in the motor degradation diagnosis method of Embodiment 1, specific methods that enable degradation diagnosis for each specific part will be described in detail below, separated by the methods used in Embodiment 1.
[0021] <Utilizing the Frequency Dependence of Impedance> Each part (element) of a motor has different impedance characteristics and exhibits different responses in specific frequency bands. By utilizing these frequency-dependent characteristics, it is possible to diagnose the deterioration state of each part. In other words, the impedance Z of each part of a motor is generally expressed by the following equation (1): Z = R + j(ωL - 1 / (ωC)) ... (1) Here, R is the resistance mainly derived from the windings, ω is the angular frequency of the voltage applied to the motor or each element of the motor (e.g., windings), L is the inductance due to the windings and core, and C is the capacitance due to the insulation between windings and the contact points. As the frequency changes, the frequency band in which these motor elements (windings, core, etc.) respond changes, so if this can be clearly distinguished, it becomes possible to diagnose deterioration specific to each part of the motor.
[0022] <Degradation Diagnosis in the Low-Frequency Band> In the low-frequency band, the impedance of the motor windings and core primarily affects the performance. These will be explained below with reference to equation (1) above. First, regarding winding degradation, as dielectric breakdown or poor conductivity progresses, the resistance component R of the windings increases, which changes the impedance Z. Next, regarding the effect of core degradation, in the low-frequency band, the inductance L has a significant impact, and magnetic loss in the core or damage to the coil manifests as impedance fluctuations. Therefore, by monitoring the impedance response when a pulse wave in the low-frequency band is applied to the motor, the degradation state of the windings and core can be diagnosed.
[0023] <Degradation Diagnosis under High Frequency Bands> In high frequency bands, the skin effect, or the influence of capacitance C, becomes stronger, and the response of bearings or contact points becomes more pronounced. Specifically, regarding bearing wear, contact, or degradation, high-frequency signals are highly sensitive to small fluctuations in contact resistance caused by the degradation of the motor's bearings or terminals. Wear or dirt increases the resistance of the contact points, and this is accompanied by fluctuations in capacitance, resulting in the generation of harmonic components in the impedance. Furthermore, regarding abnormalities in terminals or contacts, it has been confirmed that corrosion or poor contact of the terminals causes specific impedance fluctuations for high-frequency components, leading to an increase in certain harmonics. For example, poor contact of terminals can cause asymmetry in the spectrum in the frequency band, resulting in a symmetrical distribution of sideband peaks appearing before and after the center frequency. From the above, degradation of parts such as bearings or terminals can be detected by analyzing the impedance response obtained from high-frequency pulse waves.
[0024] <Actual Diagnostic Method> Diagnosis is performed by combining frequency band and harmonic analysis. Specifically, winding diagnosis is performed at low frequencies and bearing diagnosis at high frequencies. In this case, low-frequency and high-frequency pulse waves from the inverter are applied sequentially or simultaneously, and the change in impedance and harmonic components in each frequency band are measured. Furthermore, in identifying deterioration, if the impedance fluctuation in the low-frequency band is large, it is judged to be winding deterioration, and if the harmonics in the high-frequency band are strong, it is judged to be a bearing or terminal abnormality. Note that there is a study that compares the performance of the stator coil of an electric motor, with frequency on the horizontal axis, focusing on the change in current waveform before and after deterioration from the perspective of frequency (see, for example, Patent Document 2).
[0025] In other words, diagnostic methods based on impedance changes allow for the use of different frequency bands depending on the object being measured. Specifically, monitoring noise components in the high-frequency range is suitable for diagnosing bearing degradation, while evaluating the phase fluctuations of the first harmonic or the symmetry of the current waveform is suitable for diagnosing rotor eccentricity. Furthermore, detecting iron loss can capture the trend of degradation over time by continuously monitoring the changes in the intensity of third or fifth harmonics.
[0026] <How to set high and low frequency bands> Set the frequency band appropriate for the motor to be diagnosed. To do this, first identify the resonant frequency (details of this will be explained later). Specifically, identify the resonant frequency of the motor to be diagnosed. Next, set the frequency band. Specifically, set the high frequency band and low frequency band using the identified resonant frequency of the motor to be diagnosed as the boundary.
[0027] It should be noted that precisely identifying the motor's resonant frequency is not essential; if the approximate frequency range is known, it is possible to set high-frequency and low-frequency bands for diagnostic purposes. For example, in the case of typical small induction motors (0.4 kW to 3.7 kW), the resonant frequency tends to be around 5 to 20 kHz. In addition, medium-sized induction motors (7.5 kW to 15 kW) tend to have a resonant frequency of around 10 to 30 kHz, and large motors (22 kW to 55 kW) tend to have a resonant frequency of around 20 to 40 kHz.
[0028] However, since the value of this resonant frequency varies depending on the model, and because it is necessary to set the frequency band while avoiding the vicinity of the resonance point, it is desirable to identify the resonant frequency in some way when using the diagnostic method of Embodiment 1 on an actual device.
[0029] <Method for determining the resonant frequency> The resonant frequency can be determined by performing impedance measurements at multiple frequencies in advance and extracting the frequency at which the impedance changes abruptly from a Bode plot or similar.
[0030] The measurement range for this resonant frequency can typically be assumed to be 5 kHz to 40 kHz, corresponding to motor capacities of 0.4 kW to 55 kW (see, for example, Non-Patent Document 1). For example, Non-Patent Document 1 shows resonant characteristics extending up to approximately 60 kHz for a 3.3 kV class high-voltage motor, covering a frequency band where the impedance peak characteristics are particularly pronounced.
[0031] One specific method for identifying the resonance frequency is to use a sweep signal (frequency sweep signal) that continuously changes multiple frequency components as the measurement pulse. This allows for measuring the impedance response over a wide bandwidth and identifying the resonance frequency. Alternatively, even without using a sweep signal, it is possible to apply a pulse wave containing a wide range of frequency components, such as a square wave, and then perform FFT (Fast Fourier Transform) processing on the response current to identify the peak corresponding to the resonance frequency on the frequency spectrum.
[0032] In addition, as explained above, the signal processing unit and the abnormality detection unit analyze using impedance characteristics or impedance information. However, diagnosis is not limited to these methods; it is possible to diagnose using multiple methods, such as response current waveforms or frequency characteristics.
[0033] Specifically, methods include evaluating rise time, transient response decay time, and peak current value from the time waveform of the response current, which is the response signal when a pulse voltage is applied, and methods that apply FFT to the response current to extract current components for each frequency and analyze abnormal trends. It is also possible to identify specific abnormal patterns by mapping the time axis and applied voltage level in two dimensions and visualizing the nonlinear response current. Furthermore, it is possible to manage the history of the response current in a time series and analyze trends, or to extract abnormal harmonics that occur in a short time using wavelet transform.
[0034] A key feature of the motor degradation diagnosis method of Embodiment 1 is that, unlike conventional techniques, it can perform degradation diagnosis in real time while maintaining the motor's operating state. Conventional methods often only perform motor diagnosis when the motor is stopped, making it difficult to detect abnormalities early in actual operating environments. The motor degradation diagnosis method of Embodiment 1 allows diagnosis even when the motor is running, reducing equipment downtime and enabling immediate detection of abnormalities.
[0035] Furthermore, in the motor degradation diagnosis method of Embodiment 1, more precise diagnosis is possible by setting a frequency band based on the resonant frequency. In the low frequency band, the effects of winding resistance changes or bearing wear can be detected, while in the high frequency band, the effects of iron loss or insulation degradation can be accurately grasped. This makes it possible to analyze motor abnormalities from multiple angles and improve diagnostic accuracy.
[0036] As described above, by applying the motor deterioration diagnosis method of Embodiment 1, it is possible to optimize equipment maintenance management and contribute to extending the lifespan of motors or improving preventive maintenance. This embodiment achieves higher diagnostic accuracy compared to conventional technology and enables real-time detection of motor abnormalities.
[0037] Embodiment 2. Next, the motor degradation diagnosis method of Embodiment 2 will be described. The motor degradation diagnosis method according to Embodiment 2, in addition to the diagnosis method described in Embodiment 1, applies pulse waveforms of different frequency bands sequentially or simultaneously and analyzes the impedance characteristics and signal response in each frequency band.
[0038] In other words, in this second embodiment, similar to the diagnostic method in the first embodiment, the pulse voltage generation unit generates pulse waves in the low-frequency band and the high-frequency band based on a command from the pulse voltage generation command unit, and applies the generated pulse waves to the motor sequentially or simultaneously. Subsequently, the response current measured on the ammeter side is transmitted to the signal processing unit, and the impedance characteristics for each frequency are analyzed. The abnormality determination unit identifies the location of the abnormality and notifies the display unit of the diagnostic result. Here, when applying pulse waves to the motor simultaneously, a pulse wave (composite wave) that combines (combines) both high-frequency and low-frequency bands can be generated and applied to the motor.
[0039] In this second embodiment, the impedance characteristics of the motor exhibit different responses depending on the frequency band. This is achieved by utilizing the fact that in the low-frequency band, the resistance component of the windings or bearing wear has an effect, while in the high-frequency band, iron loss or insulation degradation is mainly reflected. In other words, by applying pulse waves in the low-frequency and high-frequency bands and analyzing the respective responses, specific abnormal factors can be identified.
[0040] Specifically, in diagnosing deterioration of winding insulation, applying a high-frequency pulse waveform allows damage to the winding's insulation layer to be detected as impedance fluctuations. In diagnosing bearing wear, analyzing response changes in the low-frequency range allows for the detection of vibrations caused by wear and the understanding of the progression of the abnormality.
[0041] Here, the wear of bearings is strongly correlated with the increase in the rotational speed of the motor and the accompanying vibration frequency. When operated at a high rotational speed for a long time, deterioration of the lubricating grease or wear due to metal contact tends to progress in the bearing. In addition, in an environment where the load torque fluctuates, asymmetric forces are repeatedly applied to the shaft, which affects the relative positional relationship between the rotor and the stator, thereby inducing rotor eccentricity, amplifying vibration, and disrupting magnetic balance, increasing mechanical stress on the bearing or shaft, and ultimately becoming a factor that accelerates the progression of wear or damage. For example, the disruption of magnetic balance causes time-axis asymmetry in the current waveform, that is, in one cycle waveform of the current, the shapes of the positive half-cycle and the negative half-cycle do not match, resulting in a difference in peak current.
[0042] As described above, by implementing the motor deterioration diagnosis method according to the second embodiment, diagnosis with higher accuracy can be realized. That is, abnormalities that were difficult to detect in the prior art can be identified with high accuracy by comparing analysis results of different frequency bands. In addition, by recording the characteristics for each frequency over a long period of time and monitoring secular changes, it becomes possible to detect signs of abnormality occurrence in advance.
[0043] Third Embodiment. Next, a motor deterioration diagnosis method according to the third embodiment will be described with reference to FIG. 3. In addition to the diagnosis method described in the first embodiment, the motor deterioration diagnosis method according to the third embodiment acquires the rotational speed of the motor, the load torque, and the ambient temperature of the motor, selects an appropriate low frequency band or high frequency band based on these operating conditions, and applies a pulse waveform in the selected frequency band.
[0044] Here, the relationship between the rotational speed, the load torque, the ambient temperature of the motor, and the deterioration of each part of the motor will be briefly described. The deterioration of the winding is mainly caused by thermal deterioration of the insulating material due to temperature rise. The combined effect of Joule heat accompanying energization or an increase in ambient temperature accelerates the deterioration of the insulating coating, and the risk of partial discharge or dielectric breakdown increases with long-term use. Particularly in an environment with a high ambient temperature, the progression of deterioration becomes significant.
[0045] Therefore, in order to clarify these effects on motor degradation, as shown in Figure 3, a sensor 11 is connected to the motor to acquire operating parameters such as load torque, rotational speed, and ambient temperature. The acquired data is transmitted to the signal processing unit, where the current operating state of the motor is analyzed. Subsequently, a diagnostic pulse wave of an appropriate frequency band is selected, and a command is sent from the pulse voltage generation command unit to the pulse voltage generation unit. The pulse voltage generation unit generates a pulse wave of the selected frequency band and applies it to the motor. The ammeter unit acquires the response current, and the signal processing unit performs impedance analysis, enabling the abnormality detection unit to detect abnormalities adapted to the load conditions.
[0046] In this motor degradation diagnosis method of Embodiment 3, the impedance characteristics of the motor, which change depending on the load or operating conditions, are utilized. Since the impedance characteristics of the motor change depending on the load or operating conditions, for example, at high loads, the winding resistance of the motor increases, affecting the low-frequency components of the impedance. On the other hand, at low loads, the influence of harmonic components in the current waveform becomes more pronounced, and fluctuations in high-frequency components become apparent. Therefore, by appropriately selecting the diagnostic frequency band according to the load torque or rotational speed, more accurate degradation diagnosis becomes possible. It is known that confirming these load conditions is particularly important for motors with gearheads.
[0047] Specific examples of applications for the motor degradation diagnosis method of this third embodiment include degradation diagnosis of wind power generation equipment and degradation diagnosis of industrial pumps. In the former case of degradation diagnosis of wind power generation equipment, attention is paid to the fact that the load torque of wind power generation motors fluctuates greatly depending on the wind speed. At low wind speeds, diagnosis is performed in the high-frequency band, and if an abnormality is found, signs of iron loss or insulation degradation can be detected. On the other hand, at high wind speeds, diagnosis is performed in the low-frequency band, and changes in winding resistance or bearing wear can be identified.
[0048] In the deterioration diagnosis of the latter industrial pump, when the rotational speed of the pump is within a certain range, diagnosis in a low frequency band that emphasizes changes in winding resistance under a high load state is performed. On the other hand, when the operating load of the pump is reduced, the effect of the cavitation phenomenon appears, so diagnosis in a high frequency band is applied to perform abnormality detection.
[0049] As described above, by applying the motor deterioration diagnosis method according to the third embodiment, optimal diagnosis in consideration of load conditions becomes possible. This reduces the risk of misdiagnosis, and enables more accurate detection of motor abnormalities compared to conventional diagnosis using a fixed frequency.
[0050] Embodiment 4. Next, the motor deterioration diagnosis method according to the fourth embodiment will be described in detail below. In addition to the diagnosis method described in Embodiment 1, the motor deterioration diagnosis method according to the fourth embodiment monitors load fluctuations of the motor in real time, and applies a diagnostic pulse waveform only during time periods when the load is stable. Here, the time period when the load is stable refers to a time period during which fluctuations in the current value or the torque command value remain below a certain threshold (e.g., ±5%). In contrast, the time period during which the load fluctuates significantly refers to a time period in which a large change (e.g., ±10% or more) occurs in a short period of time.
[0051] In the fourth embodiment, similarly to FIG. 3, a load monitoring sensor is connected to the motor to monitor load fluctuations of the motor in real time. When the fluctuation range of the load becomes equal to or less than the threshold, the pulse voltage generation command unit transmits an application command for a diagnostic pulse waveform to the pulse voltage generation unit. The pulse voltage generation unit generates the selected pulse waveform and applies it to the motor. The diagnosis result is stored in a storage unit, and analyzed by an abnormality determination unit while being compared with past data.
[0052] When performing motor diagnostics, if measurements are taken when the load fluctuates significantly, disturbances can be introduced into the diagnostic results, leading to misdiagnosis. In particular, when rapid torque fluctuations occur, the motor's impedance response becomes unstable, reducing diagnostic accuracy. Therefore, it is necessary to avoid measurements in such situations. To this end, it is important to detect the period when the load rotation speed is stable within a certain range and apply the diagnostic pulse at that time. This enables highly accurate diagnostics with reduced noise interference. As mentioned above, it is necessary to perform the diagnosis when the load fluctuation range falls below a threshold.
[0053] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 4 can be applied include, for example, the diagnosis of elevator drive motors or conveyor drive motors. In the former case, since the load on an elevator motor fluctuates greatly when passengers get on and off, or when the elevator starts and stops, performing the diagnosis only while the elevator is moving at a constant speed where the load is stable eliminates external disturbances and enables accurate diagnosis. On the other hand, in the latter case, since the load on a conveyor motor in a production line fluctuates depending on the product transport conditions, performing the diagnosis in a section where the weight fluctuation of the products is small enables the acquisition of more stable data.
[0054] As described above, by applying the motor degradation diagnosis method of Embodiment 4, it is possible to eliminate noise caused by load fluctuations and improve diagnostic accuracy. In particular, even in environments with large load fluctuations, the reliability of anomaly detection can be increased by performing the diagnosis at the appropriate timing.
[0055] Embodiment 5. Next, the motor degradation diagnosis method of Embodiment 5 will be described. The motor degradation diagnosis method according to Embodiment 5 is characterized in that, in addition to the diagnosis method described in Embodiment 1, a diagnosis pulse waveform is applied only when the motor is stopped, the response current is acquired, and the degradation state is diagnosed by analyzing the fluctuations in harmonic components and impedance.
[0056] In the motor degradation diagnosis method of this embodiment 5, the inverter control unit monitors the operating state of the motor and, if it determines that diagnosis in the stopped state is necessary, transmits a command signal to the pulse voltage generation unit to instruct the generation of a pulse waveform. The pulse voltage generation unit generates a pulse waveform based on the command and applies it to the motor, and the ammeter unit measures the resulting response current. The signal processing unit analyzes the harmonic components and impedance characteristics of this response current, and while storing past diagnostic data in the memory unit, the abnormality determination unit evaluates the progression of the abnormality.
[0057] In the motor degradation diagnosis method of Embodiment 5, motor diagnosis is usually performed by real-time analysis during operation, but since it is not possible to completely eliminate the influence of the load, it is important to perform the diagnosis in a stopped state. This is because the characteristics of the motor itself can be measured more accurately in a stopped state. In particular, insulation degradation of the windings or wear of the bearings clearly manifest as impedance changes when stationary, so with the motor completely stopped, a pulse waveform of a specific frequency band is applied and the response current is obtained. The data obtained by the signal processing unit is then compared with past reference data to determine whether or not there is an abnormality.
[0058] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 5 can be applied include, for example, the diagnosis of a pump drive motor or a machine tool motor. In the former case, the diagnosis is performed only when the pump is stopped, eliminating the influence of the load during operation and enabling a more accurate diagnosis. In the latter case, by measuring the insulation degradation in a stopped state after a long period of operation and predicting the progression of the abnormality, it becomes possible to prevent sudden failures.
[0059] As described above, by applying the motor degradation diagnosis method of Embodiment 5, more accurate abnormality diagnosis becomes possible without being affected by the load during operation. In addition, by periodically acquiring data while stationary, the trend of deterioration over time can be grasped more clearly.
[0060] Embodiment 6. Next, the motor degradation diagnosis method of Embodiment 6 will be described. The motor degradation diagnosis method according to Embodiment 6, in addition to the diagnosis method described in Embodiment 5, monitors the motor's operating status, switches the operating mode at a timing when diagnosis is possible, and applies a diagnosis pulse waveform when the motor is stopped.
[0061] In the motor degradation diagnosis method of Embodiment 6, the pulse voltage generation command unit monitors the motor's operating schedule and determines the timing when it is scheduled to stop. Based on this information, the signal processing unit generates an appropriate diagnostic pulse waveform and transmits it to the pulse voltage generation unit. Next, the pulse voltage generation unit performs the diagnosis in a stopped state according to the operating schedule, and the signal processing unit analyzes the data measured by the ammeter unit. Finally, the abnormality determination unit compares the results with past measurement results to evaluate whether there are any signs of abnormality and outputs the result to the display unit.
[0062] To improve the accuracy of deterioration diagnosis, it is important to perform the diagnosis at the appropriate time. In the motor deterioration diagnosis method of Embodiment 6, the operating state of the motor is monitored in real time, and the diagnosis is performed at the appropriate stopping time. For example, by utilizing the timing when a factory's production line periodically stops, abnormalities can be identified more accurately.
[0063] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 6 can be applied include, for example, diagnostics utilizing the periodic shutdown of a manufacturing line and diagnostics of an elevator system. In the former, diagnostics can be automatically performed after the end of daily operations, allowing for long-term monitoring of abnormal trends. In the latter, by performing diagnostics during periods of low usage, it is possible to accurately evaluate whether abnormalities are progressing and to formulate maintenance plans as needed.
[0064] As described above, by applying the motor deterioration diagnosis method of Embodiment 6, it is possible to perform the diagnosis at an appropriate timing considering the operating conditions, thereby improving the accuracy of abnormality detection and preventing misdiagnosis.
[0065] Embodiment 7. Next, the motor degradation diagnosis method of Embodiment 7 will be described. The motor degradation diagnosis method according to Embodiment 7, in addition to the diagnosis method of Embodiment 1, stores the response current during steady-state operation of the motor as reference data in the storage unit and diagnoses the degradation state by comparing it with the response current acquired periodically. Steady-state operation refers to a situation where the change in current, torque, and rotational speed over the last few seconds is within ±5%.
[0066] In the motor degradation diagnosis method of Embodiment 7, the memory unit stores response current data during steady-state operation of the motor over a long period of time and manages it as reference data. The signal processing unit periodically compares newly acquired response current data with past data, and the abnormality determination unit analyzes the data trend. If an abnormality is detected, the information is output to the display unit and used to formulate an appropriate maintenance plan.
[0067] Since motor characteristics change over time, it is important to accumulate diagnostic data at regular intervals and monitor abnormal trends over the long term when diagnosing motor degradation. Therefore, in the motor degradation diagnosis of this embodiment 7, response current data during normal operation is saved and compared with newly acquired data to identify subtle degradation.
[0068] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 7 can be applied include, for example, life prediction of cooling fan motors and maintenance management of motors for railway vehicles. In the former, it is possible to identify abnormal trends by comparing long-term data and analyzing minute changes in rotational speed. In the latter, it is possible to utilize long-term operational data to perform diagnoses to prevent sudden failures.
[0069] As described above, by applying the motor deterioration diagnosis method of Embodiment 7, it becomes possible to grasp the progression of aging deterioration over the long term, thereby enabling early detection of abnormalities. This allows for the optimization of the motor maintenance plan and the prevention of sudden failures.
[0070] Embodiment 8. Next, the motor degradation diagnosis method of Embodiment 8 will be described. The motor degradation diagnosis method of Embodiment 8, in addition to the diagnosis method of Embodiment 1, identifies specific degradation factors such as changes in winding resistance, bearing wear, rotor eccentricity, and increased iron loss based on the analysis results of the acquired response current.
[0071] In the motor degradation diagnosis method of Embodiment 8, the pulse voltage generation unit applies a diagnostic pulse voltage to the motor based on a command signal from the pulse voltage generation command unit, and the ammeter unit measures the response current. The signal processing unit performs frequency analysis of the response data or extracts harmonic components, and the abnormality determination unit identifies the degradation factors of each part by comparing them with specific abnormality patterns. The diagnostic results are stored in the memory unit, and the progress of degradation can be checked by comparing them with past diagnostic data.
[0072] Each part of a motor has a different degradation mechanism, requiring appropriate analysis depending on the motor degradation diagnosis method. For example, winding insulation degradation manifests as an increase in impedance, and bearing wear affects the response current as vibrations appearing in specific frequency components (several kHz to tens of kHz) due to vibrations generated with shaft rotation (ripple generation). Also, rotor eccentricity causes rotor misalignment or unbalance, resulting in fluctuations in the first harmonic component (corresponding to the power supply frequency). Furthermore, physical displacement of the rotor causes asymmetry in the frequency band corresponding to the rotational speed, resulting in left-right asymmetry or phase distortion of the current waveform on the time axis. In addition, increased iron loss leads to deterioration or saturation of magnetic properties, increasing the harmonic components (third, fifth, etc.) associated with iron loss. As a result, the energy of the entire harmonic spectrum increases, and power consumption tends to rise over time. Therefore, the increase in iron loss can be identified through trend analysis or spectral historical comparison (a method that visualizes spectral data accumulated over time in a heat map format, etc., and evaluates spectral changes such as peak shifts that accompany the progression of degradation over time). In this embodiment 8, the change pattern of the response current is analyzed to identify each degradation factor.
[0073] As described above, by applying the motor deterioration diagnosis method of Embodiment 8, it becomes possible to identify not only whether or not there is a simple abnormality, but also the location of the abnormality and its cause in detail. This improves the efficiency of maintenance work and enables planned maintenance.
[0074] Embodiment 9. Next, the motor deterioration diagnosis method of Embodiment 9 will be described. The motor deterioration diagnosis method according to Embodiment 9, in addition to the diagnosis method of Embodiment 1, outputs the diagnosis results to a display unit and visually displays the degree of deterioration in the form of numerical values, graphs, alerts, etc.
[0075] In the motor degradation diagnosis method of Embodiment 9, the signal processing unit analyzes the diagnostic data, and the abnormality determination unit quantifies the severity of the abnormality. It also compares this with past data stored in the memory unit to determine the progression of the abnormality. The display unit then displays a graph showing the degradation status or an alert when a threshold is exceeded. For example, spectral data is displayed as visualized data such as a heatmap. This makes it easier to understand the temporal changes in the spectrum (peak shift, energy increase, etc.) that accompany the progression of degradation.
[0076] The diagnostic results are presented visually, not just as indicators of whether or not there is an abnormality, but also as indicators of the progression or degree of deterioration of the abnormality, thereby facilitating the development of maintenance plans. Furthermore, the results of the response current analysis are graphed, allowing operators to intuitively understand the status of the abnormality.
[0077] Specific diagnostic examples applying the motor degradation diagnostic method of this embodiment 9 include, for example, numerical display of diagnostic results and graphical display of abnormal trends. By applying this method, the former allows the degradation index to be expressed on a scale of 0 to 100, quantifying the progression of the abnormality. The latter allows the impedance fluctuations over the past six months to be displayed in time series, confirming whether the abnormality is progressing.
[0078] As described above, the motor deterioration diagnosis method of Embodiment 9 allows the operator to intuitively grasp the progression of the abnormality and formulate an appropriate maintenance plan.
[0079] Embodiment 10. Next, the motor degradation diagnosis method of Embodiment 10 will be described. The motor degradation diagnosis method of Embodiment 10, in addition to the diagnosis method of Embodiment 1, outputs a control signal that changes the motor's operating mode to reduce the load or stops operation if the degradation diagnosis result exceeds a set threshold. Here, the threshold can be set, for example, based on the data distribution under normal conditions (mean ± 3σ, etc.), taking into account the number of years of use or the motor model by adding a correction coefficient. Note that when using AI, which will be described in Embodiment 12 and later, it can be automatically adapted by the learning model.
[0080] In the motor degradation diagnosis method of Embodiment 10, the signal processing unit analyzes the diagnosis result, and the abnormality determination unit determines whether the threshold has been exceeded. If the abnormality exceeds a certain level, the pulse voltage generation command unit outputs a command to change the operating mode, and the pulse voltage generation unit either reduces the load or performs an emergency stop.
[0081] In other words, in the motor degradation diagnosis method of this embodiment 10, when an abnormality is detected, instead of simply issuing a notification, the motor's operating mode is automatically changed to prevent further degradation or serious failure. Furthermore, the abnormality determination unit analyzes the diagnosis result and outputs an appropriate control command according to the level of abnormality.
[0082] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 10 can be applied include, for example, load reduction in the event of an abnormality and emergency stop control. By applying this method, in the former case, the load torque can be reduced when the motor temperature rise exceeds a certain threshold. In the latter case, if the impedance change is sudden, the motor can be stopped immediately to prevent further degradation.
[0083] As described above, the motor deterioration diagnosis method of Embodiment 10 not only detects abnormalities but also appropriately controls the motor's operation, thereby preventing sudden failures and improving the overall health of the equipment.
[0084] Embodiment 11. Next, the motor degradation diagnosis method of Embodiment 11 will be described. The motor degradation diagnosis method of Embodiment 11, in addition to the diagnosis method of Embodiment 1, applies pulse waveforms at different voltage levels during diagnosis and analyzes the change in response current in order to identify the insulation degradation of the motor.
[0085] In the motor degradation diagnosis method of Embodiment 11, the pulse voltage generation command unit commands the generation of pulse waveforms at different voltage levels and applies them to the motor via the pulse voltage generation unit. The ammeter unit measures the response current at each voltage level, and the signal processing unit performs data analysis. The abnormality determination unit identifies changes in the insulation state and stores the diagnosis results in the storage unit. If there are signs of an abnormality, an alert is displayed on the display unit.
[0086] In other words, the insulation of motor windings gradually deteriorates due to aging or overvoltage stress, and as dielectric breakdown progresses, it causes short circuits or arc discharges. Therefore, in the motor deterioration diagnosis method of this embodiment 11, changes in insulation characteristics are captured by applying pulse waveforms of different voltage levels and comparing their responses. In particular, if an impedance change or an increase in leakage current is detected when a high voltage is applied, it can be determined that there is a high possibility of insulation deterioration. Examples of applying different voltage levels as described above include a pulse of about 200V in normal diagnosis (within the rated range of the inverter output) and a high-voltage pulse of 400V to 600V in insulation diagnosis (conditions close to the dielectric breakdown voltage). The reason for using different voltage levels is that dielectric breakdown or leakage current becomes apparent at high voltage, and the purpose is to intentionally make visible insulation deterioration due to minute cracks etc. that do not surface at normal operating voltages.
[0087] Specific diagnostic examples applying the motor degradation diagnostic method of this embodiment 11 include, for example, leakage current detection by applying high voltage and impedance change analysis for each voltage. In the former, if there is no problem at low voltage but the leakage current increases sharply at high voltage, it is possible to determine signs of dielectric breakdown. In the latter, the change in dielectric constant due to the degradation of the insulating material is evaluated in combination with frequency characteristics.
[0088] As described above, the motor degradation diagnosis method of Embodiment 11 makes it possible to detect motor insulation degradation early and take appropriate measures before the abnormality progresses. Furthermore, this reduces the risk of failure due to sudden dielectric breakdown and improves the reliability of the equipment.
[0089] Embodiment 12. Next, the motor degradation diagnosis method of Embodiment 12 will be described with reference to Figure 4. The motor degradation diagnosis method according to Embodiment 12, in addition to the diagnosis method of Embodiment 1, collects the response current obtained by applying a pulse waveform generated from an inverter device to the motor as a data set, inputs the data set into an AI model to estimate the fluctuations in the motor's impedance characteristics and harmonic components, and diagnoses the degradation state based on these fluctuations.
[0090] In the motor degradation diagnosis method of Embodiment 12, motor degradation diagnosis is performed using the motor degradation diagnosis device 102 shown in Figure 4. This motor degradation diagnosis device 102 is the motor degradation diagnosis device 100 described in Embodiment 1, with the addition of the AI analysis unit 12 described below. The inverter control unit commands the generation of a pulse waveform, and the pulse voltage generation unit applies the generated pulse waveform. The ammeter side unit collects response data, and the signal processing unit extracts specific feature quantities. The AI analysis unit 12 then learns the collected data and predicts the possibility of an abnormality. The abnormality determination unit performs a diagnosis based on the AI estimation result and outputs the result to the display unit. The number of motors to be measured may be one or multiple, and desired diagnostic data is collected and analyzed for each motor (see Figure 4). The motor degradation diagnosis device 102 may also be the motor degradation diagnosis device 101 described in Embodiment 3, with the addition of the AI analysis unit 12.
[0091] Generally, motor degradation progresses due to a variety of factors, making accurate anomaly detection difficult with conventional methods. Therefore, the motor degradation diagnosis method of this embodiment 12 utilizes AI (artificial intelligence) to detect signs of abnormality by comparing them with past diagnostic data. Specifically, by using a neural network or decision tree model to learn subtle fluctuations in impedance characteristics or anomalies in harmonic components, highly accurate diagnosis becomes possible.
[0092] Specific diagnostic examples applying the motor degradation diagnosis method of this embodiment 12 include, for example, the application of an anomaly prediction algorithm and its application to predictive maintenance. In the former, data from normal motor operation and data from abnormal situations are learned, and anomaly patterns are automatically detected by AI. In the latter, if signs of degradation exceed a certain threshold, the AI issues an early warning and recommends planned maintenance.
[0093] As described above, the motor degradation diagnosis method of Embodiment 12 utilizes AI to detect minute abnormalities that were difficult to find with conventional methods at an early stage. This enables more efficient equipment maintenance and prevents sudden motor failures.
[0094] Embodiment 13. Next, the motor degradation diagnosis method of Embodiment 13 will be described. The motor degradation diagnosis method according to Embodiment 13 is a combination of the diagnosis method of Embodiment 12 and involves training an AI model using a neural network with measured data of normal and abnormal states as training data.
[0095] In the motor degradation diagnosis method of Embodiment 13, the signal processing unit transmits the collected data to the AI analysis unit, which inputs it into the AI model as training data. The abnormality determination unit performs a diagnosis based on the output results of the learned AI model and outputs the results to the display unit.
[0096] Generally, to improve the accuracy of anomaly diagnosis using machine learning, it is important to train the model with a large amount of actual measurement data. In the motor degradation diagnosis method of this embodiment 13, motor data in both normal and abnormal states are input into the AI model to enhance the learning of anomaly patterns. By utilizing the trained model, the accuracy of real-time diagnosis is improved, and higher reliability can be provided than with conventional rule-based diagnosis.
[0097] Specific diagnostic examples to which the motor degradation diagnostic method of this embodiment 13 can be applied include, for example, analysis of long-term data using AI and feedback learning when an anomaly occurs. In the former, the accuracy of anomaly detection can be improved by learning the progression pattern of anomalies based on several years of operating data. In the latter, the accuracy of the diagnosis can be improved by comparing the diagnostic results with the actual anomaly occurrence situation and updating the model as needed.
[0098] As described above, the motor degradation diagnosis method of Embodiment 13 enables more accurate abnormality diagnosis by utilizing a pre-trained AI model. This improves the efficiency of predictive motor maintenance and minimizes equipment downtime.
[0099] Embodiment 14. Next, the motor degradation diagnosis method of Embodiment 14 will be described. The motor degradation diagnosis method according to Embodiment 14, in addition to the diagnosis method of Embodiment 12, involves an AI model operating in a cloud environment and comprehensively analyzing the diagnosis data of multiple motors to analyze the correlation of abnormal trends among motors of the same model.
[0100] In the motor degradation diagnosis method of Embodiment 14, the motor degradation diagnosis device 100 or the motor degradation diagnosis device 101 transmits the diagnosis data of each motor to a cloud-based server which is a motor degradation diagnosis support server, and the AI analysis unit in the cloud environment performs integrated data analysis. Alternatively, the motor degradation diagnosis device 102 may be used (in this case, the AI-analyzed data can also be used as is). By training the AI in the cloud environment with data acquired from multiple motors, the accuracy of abnormality trend analysis is improved. The diagnosis results from the abnormality judgment unit are transmitted to the operators of each motor, and if an abnormality occurs, a warning is output to the display unit. In addition, the abnormality data is stored in the storage unit and used to optimize the long-term diagnosis model.
[0101] Generally, to detect motor abnormalities with higher accuracy, it is effective to analyze not only data from individual motors but also information obtained from motors of the same model or in similar environments. In the motor degradation diagnosis method of this embodiment 14, an AI model is operated in a cloud environment to comprehensively analyze diagnostic data collected from multiple motors installed in a factory or facility. This analysis makes it possible to identify correlations in abnormality trends or common factors of degradation, enabling failure prediction with higher accuracy than individual motor diagnosis.
[0102] Specific diagnostic examples applying the motor degradation diagnostic method of this embodiment 14 include, for example, applications to equipment management in large-scale factories and remote monitoring systems. In the former, diagnostic data from multiple motors on the same line are integrated and analyzed on the cloud to detect abnormal trends common to specific models or operating conditions. In the latter, diagnostic data from motors installed at multiple distributed locations is collected on the cloud to grasp the overall degradation trend and optimize preventive maintenance.
[0103] As described above, the motor degradation diagnosis method of Embodiment 14 enables a more advanced predictive diagnosis because it can comprehensively analyze the abnormality trends of not only a single motor but also multiple motors. Furthermore, by statistically identifying the causes of abnormalities and discovering common problems in advance, it contributes to optimizing the maintenance management of the entire facility. In addition, by utilizing a cloud environment, remote monitoring becomes possible, enabling a rapid response in the event of an abnormality.
[0104] The motor degradation diagnosis method of Embodiments 1-12 described above may also be performed using AI. Details of the motor degradation diagnosis method using AI will be specifically explained below with reference to Figures 5 and 6. <Motor Degradation Diagnosis Method Using AI> When performing degradation diagnosis using AI, the degradation diagnosis is performed in the following order (see Figures 5 and 6). First, estimation is performed using an AI model. Specifically, the response current obtained by applying a pulse wave generated from an inverter device to the motor is collected as a dataset (Step S1 in Figures 5 and 6), and this dataset is input into the AI model (Step S2 in Figure 6) to estimate the motor's impedance characteristics and harmonic component fluctuations. At this time, the impedance: Z(f) (=V(f) / I(f)) obtained from the frequency spectrum of the pulse voltage waveform: V(f) and the frequency spectrum of the response current: I(f) is used (Step S3 in Figures 5 and 6). Then, based on the estimated motor's impedance characteristics and harmonic component fluctuations, degradation diagnosis is performed (Step S4 in Figure 6). In this process, a pre-trained AI model is used. Specifically, the pre-trained AI model is a model that takes the motor's response current as input and learns the fluctuations in impedance characteristics and harmonic components as ground truth data. The learning data consists of past measured values or results from simulations that simulate abnormal or normal conditions.
[0105] Specifically, the anomaly detection unit employs a configuration that accurately identifies various degradation factors by utilizing machine learning methods such as a probability distribution model based on training data, anomaly scoring, clustering, and support vector machines (SVM), in addition to analysis results obtained by trend analysis that manages the history of the response current in a time series or analysis using wavelet transform. This enables flexible and highly accurate anomaly detection for a variety of degradation factors. Furthermore, in addition to the conventional threshold method, the anomaly detection unit can also use statistical determination methods based on a probability distribution model or anomaly scoring based on training data (data with normal and abnormal labels).
[0106] As a concrete example of the method using the above-mentioned training data, response waveforms of normal and various abnormal conditions (winding short circuit, bearing wear, rotor eccentricity, etc.) collected through experiments can be labeled and stored in a database. A classification model can then be constructed using Random Forest or SVM. This allows for high identification accuracy for diagnostic data of unknown conditions and reduces the false positive rate. Furthermore, as an example of visual output, it is possible to plot the data in two dimensions using PCA (Principal Component Analysis) to visualize the appearance of abnormal clusters.
[0107] The pulse voltage generation command unit 10, signal processing unit 6, abnormality determination unit 8, display unit 9, and the AI analysis unit 12, which are common main components of the motor degradation diagnostic devices 100, 101, and 102, and the AI analysis unit 12, which is a main component of the motor degradation diagnostic device 102, are composed of a processor 300 and a storage device 301, as shown in Figure 8 as an example of the hardware. Here, although the storage device 301 is not shown, it comprises a volatile storage device such as random access memory and a non-volatile auxiliary storage device such as flash memory. Alternatively, an auxiliary storage device such as a hard disk may be provided instead of flash memory. The processor 300 executes the program input from the storage device 301. In this case, the program is input from the auxiliary storage device to the processor 300 via the volatile storage device. The processor 300 may also output data such as calculation results to the volatile storage device of the storage device 301, or it may store the data in the auxiliary storage device via the volatile storage device.
[0108] While this disclosure describes exemplary embodiments, the various features, aspects, and functions described in the embodiments are not limited to the application of any particular embodiment, but can be applied individually or in various combinations to the embodiments.
[0109] For example, motor degradation diagnostic data obtained by the motor degradation diagnostic device 100 or the motor degradation diagnostic device 101 can be transmitted in real time to a cloud-based motor degradation diagnostic support server 120 via a wired or wireless communication device 110, and a motor degradation diagnostic system 200 can be constructed by the motor degradation diagnostic support server (see Figure 7). This constructed motor degradation diagnostic system 200 can store time-series data and perform trend analysis, predict the progression of motor degradation using an AI model, and, in the case of multiple motors, can grasp abnormal trends across multiple motors and send alert notifications to administrators regarding data on abnormal motors when abnormal motors occur. In this case, the communication device 110 is connected (including wirelessly) to the motor degradation diagnostic device 100, or the motor degradation diagnostic device 101, and the motor degradation diagnostic support server 120. As a result, the motor degradation diagnostic data obtained by the motor degradation diagnostic device is sent to the motor degradation diagnostic support server via the communication device, and the results obtained from the analysis tools of the motor degradation diagnostic support server are sent to the motor degradation diagnostic device via the communication device. This further enhances the diagnostic performance.
[0110] Accordingly, countless variations not illustrated are conceivable within the scope of the art disclosed in this specification. These include, for example, modifications, additions, or omissions of at least one component.
[0111] 1 Inverter power supply unit, 2 Inverter control unit, 3 Pulse voltage generation unit, 4 Motor (object under test), 5 Ammeter side unit, 6 Signal processing unit, 7 Memory unit, 8 Anomaly detection unit, 9 Display unit, 10 Pulse voltage generation command unit, 11 Sensor, 12 AI analysis unit, 100, 101, 102 Motor degradation diagnosis device, 110 Communication device, 120 Motor degradation diagnosis support server, 200 Motor degradation diagnosis system
Claims
1. A method for diagnosing motor deterioration, characterized by: generating a pulse voltage in a frequency band set based on a pre-given resonant frequency of the motor under test, or a resonant frequency identified by impedance measurement, using an inverter; driving the motor with the generated pulse voltage; acquiring the motor's response current obtained by applying the pulse voltage to the motor in real time; determining the motor's impedance characteristics from a diagnostic pulse voltage output from the inverter, which serves as a diagnostic reference, and the acquired motor response current; and diagnosing the motor's deterioration state based on the motor's impedance characteristics.
2. A motor degradation diagnosis method according to claim 1, characterized in that the preset frequency band includes two or more different frequency bands, pulse voltages of each frequency band are applied to the motor sequentially or simultaneously, and the impedance characteristics and response signals of the motor in each frequency band are analyzed.
3. A method for diagnosing motor deterioration according to claim 1 or 2, characterized by acquiring the rotational speed, load torque, or ambient temperature of the motor, selecting an appropriate frequency band based on the acquired data, and applying a pulse voltage in the selected frequency band.
4. A method for diagnosing motor deterioration according to any one of claims 1 to 3, characterized in that the load fluctuations of the motor are monitored in real time, and the diagnostic pulse voltage is applied to the motor only during periods when the load is stable.
5. A method for diagnosing motor deterioration according to any one of claims 1 to 3, characterized in that the diagnostic pulse voltage is applied only when the motor is stopped, the response current of the motor is obtained, and the harmonic components and impedance characteristics of the response current are analyzed.
6. A motor deterioration diagnosis method according to claim 5, characterized in that the operating status of the motor is monitored, the operating mode of the motor is switched at a timing when the motor can be diagnosed, and the diagnostic pulse voltage is applied when the motor is stopped.
7. A method for diagnosing motor deterioration according to any one of claims 1 to 4, characterized in that the response current of the motor during steady-state operation is stored as reference data and compared with the response current of the motor acquired periodically.
8. A method for diagnosing motor deterioration according to any one of claims 1 to 7, characterized in that, based on the analysis results of the response current of the motor obtained, specific deterioration factors including changes in winding resistance, bearing wear, rotor eccentricity, and increased iron loss are identified.
9. A motor deterioration diagnosis method according to any one of items 1 to 8, characterized in that it outputs the motor diagnosis results and visually displays the degree of deterioration in the form of numerical values, graphs, and alerts.
10. A motor degradation diagnosis method according to any one of claims 1 to 9, characterized in that, if the result of the degradation diagnosis exceeds a set threshold, the motor's operating mode is changed and a control signal is output to reduce the load or stop the operation.
11. A method for diagnosing motor deterioration according to any one of claims 1 to 10, characterized in that, in order to identify the insulation deterioration of the motor, pulse waveforms are applied at different voltage levels during diagnosis and the change in response current is analyzed.
12. A method for diagnosing motor deterioration according to any one of claims 1 to 11, characterized in that the generated pulse voltage is applied to the motor and the response current obtained is collected as a dataset, the dataset is input into an AI model to estimate the impedance characteristics and harmonic component fluctuations of the motor, and the deterioration state is diagnosed based on the fluctuations.
13. The motor deterioration diagnosis method according to claim 12, characterized in that an AI model using a neural network is trained using measured data of normal and abnormal states as training data.
14. A motor degradation diagnosis method according to claim 12 or 13, characterized in that an AI model is operated in a cloud environment and diagnostic data from multiple motors is comprehensively analyzed to analyze the correlation of abnormal trends among motors of the same model.
15. A motor degradation diagnostic device for performing a motor degradation diagnosis, comprising: a pulse voltage generation command unit that outputs commands from an inverter that generates pulse voltages for driving and diagnosing a motor; a pulse voltage generation unit that generates a diagnostic pulse voltage and a driving pulse voltage applied to the motor in accordance with the commands from the pulse voltage generation command unit; an ammeter side unit that measures the response current generated by the motor; a signal processing unit that calculates the impedance characteristics of the motor by analysis based on the diagnostic pulse voltage input from the pulse voltage generation unit and the response current measured by the ammeter side unit; a storage unit that stores the impedance characteristics of the motor calculated by the signal processing unit; an abnormality determination unit that analyzes the impedance characteristics of the motor and determines the degradation state of the motor by comparing it with stored past impedance characteristic data; and a display unit that outputs the determined result as a diagnostic result, visually presents it to the operator, and issues a warning if an abnormality occurs, the motor degradation diagnostic device for performing a motor degradation diagnosis, a cloud-based motor degradation diagnosis support server, and a communication device that connects the motor degradation diagnostic device and the motor degradation diagnosis support server by wire or wireless. A motor degradation diagnosis system characterized by transmitting motor degradation diagnosis data obtained by the motor degradation diagnosis device to the motor degradation diagnosis support server in real time via the communication device, and by having the motor degradation diagnosis support server perform time-series data storage and trend analysis, predict the progression of motor degradation using an AI model, and, in the case of multiple motors, grasp the abnormal trends across multiple motors, and notify the administrator of data regarding abnormal motors when abnormal motors occur.
