Diagnosis apparatus, diagnosis method, and diagnosis program

Abstract

According to the present invention, a diagnosis apparatus, a diagnosis method, and a diagnosis program use a controller that is connected to an input unit to which time-series current data regarding a current flowing through a winding when an impulse voltage is repeatedly applied to the winding is input, and that processes the current data. The controller calculates, on the basis of the current data, a power spectral density that indicates power for each of the frequency components included in the current, and sets a formant bandwidth that includes a frequency component at which the power of the power spectral density reaches a local maximum. The elapsed time until a local maximum disappears in the set formant bandwidth is acquired, and a value obtained by multiplying the elapsed time by a prescribed coefficient is determined as a predicted time until insulation breakdown in the winding.

Description

Diagnostic apparatus, diagnostic method, and diagnostic program

[0001] The present disclosure relates to a diagnostic apparatus, a diagnostic method, and a diagnostic program.

[0002] Patent Literature 1 discloses a technique for estimating the residual dielectric breakdown strength rate of a rotating electrical machine based on the relationship between an applied voltage and the partial discharge charge amount of a stator coil corresponding to the applied voltage, and estimating the remaining life of the rotating electrical machine based on the estimation result.

[0003] Japanese Unexamined Patent Publication No. 2014-235091

[0004] According to the technique described in Patent Literature 1, it is necessary to calculate the differential value of the charge amount. Therefore, when the technique is applied to a winding to which repeated impulse voltages due to a surge voltage are applied, such as a motor driven by a high-frequency PWM inverter, the calculation error becomes large. Therefore, there is a problem that the diagnosis necessary for avoiding insulation breakdown of the winding cannot be performed.

[0005] The present disclosure has been made in view of the above problems. An object of the present disclosure is to provide a diagnostic apparatus, a diagnostic method, and a diagnostic program that can perform necessary diagnosis to avoid insulation breakdown of windings.

[0006] The diagnostic apparatus, diagnostic method, and diagnostic program according to the present disclosure use a controller that is connected to an input unit to which time-series current data related to the current flowing through a winding when repeated impulse voltages are applied to the winding is input, and processes the current data. Based on the current data, the controller calculates a power spectral density indicating the power for each frequency component included in the current, and sets a formant band including the frequency component at which the power of the power spectral density is maximum. In the set formant band, the controller acquires the elapsed time until the maximum disappears, and determines a value obtained by multiplying the elapsed time by a predetermined coefficient as the estimated time until insulation breakdown of the winding.

[0007] The controller may set the formant band to the frequency band containing the nth lowest frequency component (where n is an integer greater than or equal to 1) among the frequency components where the power of the power spectral density is maximum.

[0008] The controller may set the formant bandwidth based on the smoothed power spectral density.

[0009] The controller may count the number of peaks in the power spectral density within the formant band and determine that the timing at which the number of peaks first exceeds a predetermined threshold is the timing at which partial discharge begins.

[0010] The controller may determine that the moment when the frequency of the frequency component related to the maximum becomes higher than the frequency included in the formant band is the moment when the maximum disappears.

[0011] When voltage is repeatedly applied to the test winding until dielectric breakdown occurs, a predetermined coefficient may be set based on the time-series test current data related to the test current flowing through the test winding, and the period from the start of voltage application until dielectric breakdown occurs.

[0012] According to this disclosure, a diagnostic device, a diagnostic method, and a diagnostic program can be provided that can perform the necessary diagnostics to avoid dielectric breakdown of windings.

[0013] This is a block diagram showing the configuration of the diagnostic device relating to this disclosure. This is a flowchart showing the procedure for the determination process. This is a schematic diagram showing an example of the time variation of the voltage and current applied to the winding. This is a schematic diagram showing an example of the power spectral density. This is a schematic diagram showing an example of the power spectral density at the start of partial discharge. This is a schematic diagram showing an example of the power spectral density when the maximum has disappeared. This is a schematic diagram showing an example of the time variation of the formant frequency.

[0014] Several exemplary embodiments will be described below with reference to the drawings. Common parts in each drawing are denoted by the same reference numerals, and redundant explanations will be omitted.

[0015] [Configuration of the diagnostic device] Figure 1 is a block diagram showing the configuration of the diagnostic device according to the present disclosure. As shown in Figure 1, the diagnostic device 20 comprises an input unit 21 and a controller 25. The diagnostic device 20 may also comprise an output unit 23 and an operation unit 27. The controller 25 is connected to the input unit 21, the output unit 23, and the operation unit 27 so as to be able to communicate with them.

[0016] In addition, the input unit 21, output unit 23, and operation unit 27 may be provided within the diagnostic device 20 itself, or they may be installed outside the diagnostic device 20 and connected to the diagnostic device 20.

[0017] The input unit 21 receives time-series current data relating to the current flowing through the winding when a repetitive impulse voltage is applied to the winding. For example, the input unit 21 may be connected to a measuring device 10 and acquire current data from the measuring device 10. Alternatively, the input unit 21 may read current data from a recording medium. The current data acquired by the input unit 21 is transmitted to the controller 25 and processed by the controller 25.

[0018] Here, the measuring device 10 applies an impulse voltage to the winding to be measured and acquires current data relating to the current flowing through the winding. For example, the measuring device 10 may be equipped with a high-voltage probe and a high-frequency current sensor. The measuring device 10 may measure the voltage applied to the winding with the high-voltage probe and measure the current flowing through the winding with the high-frequency current sensor.

[0019] The measuring device 10 may include an oscilloscope. The oscilloscope may be connected to a high-voltage probe and a high-frequency current sensor and generate voltage data relating to the voltage applied to the winding and current data relating to the current flowing through the winding. The generated voltage data and current data are transmitted to the controller 25. The measuring device 10 only needs to be capable of acquiring current data and is not limited to the examples given herein.

[0020] Herein, we will note the term "winding" as it is used in this disclosure. A "winding" is a type of electric wire, mainly consisting of a conductor such as copper or aluminum covered with an insulating coating. For example, a "winding" is an element that constitutes an electric motor or generator, and it converts electrical energy and magnetic energy to each other. Specifically, it is used in equipment such as electric motors or generators for the purpose of converting electrical energy into mechanical energy, or conversely, mechanical energy into electrical energy.

[0021] For example, in electric motors driven by high-frequency PWM inverters, surge voltages repeatedly apply impulsive voltages to the windings. When impulsive voltages are applied to the windings, partial discharge may occur between the windings. Repeated partial discharges can gradually destroy the insulation coating of the windings, leading to dielectric breakdown. Dielectric breakdown in the windings can cause failure of the equipment, including the windings.

[0022] For example, to suppress failures in equipment including windings, winding structures or materials that prevent partial discharge have been employed. In addition, the operating range of equipment including windings has been selected to suppress failures in such equipment.

[0023] However, the requirements for the structures or materials that can be used in windings, or for the operating range of equipment including windings, have become increasingly stringent in recent years. For example, in the case of electric motors mounted on aircraft, windings are used in high-altitude regions with low atmospheric pressure, making partial discharge unavoidable. Furthermore, high operating temperatures of electric motors, or structural constraints that prevent the securing of sufficient insulation thickness and insulation distance for windings, also make partial discharge unavoidable.

[0024] Therefore, it is necessary to determine whether or not dielectric breakdown will occur in the winding before it occurs. According to this disclosure, it is possible to determine whether or not dielectric breakdown will occur in the winding before it occurs, and to utilize the determination result for winding maintenance, etc. As a result, it is possible to suppress equipment failures including windings caused by dielectric breakdown.

[0025] Figure 3 is a schematic diagram showing an example of the time variation of the voltage and current applied to the winding. As shown in Figure 3, the voltage applied to the winding varies impulsively over time during the interval P1. Specifically, the voltage applied to the winding increases sharply, then a predetermined voltage (Vmax) is maintained for a predetermined time, and then it decreases sharply. For example, by repeating the interval P1 over a predetermined period, an impulsive voltage is repeatedly applied to the winding.

[0026] For example, the voltage applied to the windings increases sharply in section P2. At this time, charging current flows between the windings and between the windings and the core due to parasitic capacitance present between the windings and between the windings and the core. Also, the voltage applied to the windings decreases sharply in section P3. At this time, charging current flows between the windings and between the windings and the core due to parasitic capacitance present between the windings and between the windings and the core. The direction of the current flowing through the windings in section P2 and the direction of the current flowing through the windings in section P3 are opposite.

[0027] The output unit 23 outputs various types of information obtained by the controller 25, which will be described later. In particular, the output unit 23 outputs the predicted time until dielectric breakdown in the winding. The output unit 23 may also output a determination result regarding whether or not dielectric breakdown is likely to occur in the winding. The output unit 23 may output control information for the equipment including the winding, which is determined based on the determination result, along with the determination result, or in lieu of the determination result.

[0028] For example, the output unit 23 may be connected to a monitor (not shown). The monitor may then display to the user the predicted time until dielectric breakdown of the winding, or the result of the determination. The output unit 23 may also be connected to a control device that controls the operation of equipment including the winding.

[0029] The operation unit 27 is an input device that allows the user of the diagnostic device 20 to perform operations. For example, the operation unit 27 may be a keyboard, mouse, trackball, touch panel, etc. The operation unit 27 is not limited to the examples given herein. User operations entered via the operation unit 27 are transmitted to the controller 25.

[0030] For example, the operation unit 27 may acquire information that identifies a predetermined period of time to be processed from the current data processed by the controller 25, based on the user's operation.

[0031] Here, "predetermined period" refers to the time range within the time-series current data that is processed by the controller 25. The "predetermined period" includes multiple intervals in which the current changes rapidly in accordance with the rising and falling of multiple impulse-like voltages applied to the winding. For example, the operation unit 27 may acquire the start timing of the period in which impulse-like voltages are repeatedly applied to the winding, and the end timing of the said period, based on user operation.

[0032] The controller 25 is a general-purpose computer equipped with a CPU (Central Processing Unit), memory, and an input / output unit. The controller 25 has a computer program (diagnostic program) installed for functioning as the diagnostic device 20. By executing the computer program, the controller 25 functions as one of the multiple information processing circuits (251, 253, 255, 257, 259) of the diagnostic device 20. The computer program may be stored on a storage medium readable and writable by the computer, or it may be distributed via a telecommunications line.

[0033] The controller 25 may be incorporated into a control device that controls the operation of equipment including windings. Alternatively, the diagnostic device 20 itself may be incorporated into a control device that controls the operation of equipment including windings.

[0034] This disclosure provides an example of implementing multiple information processing circuits (251, 253, 255, 257, 259) using software. However, it is also possible to configure the information processing circuits (251, 253, 255, 257, 259) by preparing dedicated hardware for each of the information processing operations described below. Alternatively, the multiple information processing circuits (251, 253, 255, 257, 259) may be configured using separate hardware.

[0035] As shown in Figure 1, the controller 25 includes a plurality of information processing circuits (251, 253, 255, 257, 259), namely a power spectral density calculation unit 251, a formant bandwidth setting unit 253, a determination unit 255, an elapsed time acquisition unit 257, and a predicted time determination unit 259.

[0036] The power spectral density calculation unit 251 calculates the power spectral density, which indicates the power of each frequency component contained in the current, based on the current data.

[0037] More specifically, the power spectral density calculation unit 251 divides the time-series current data over a predetermined period into short intervals of a certain length, and performs a Fast Fourier Transform (FFT) on each interval to convert it to the frequency domain. The power spectral density calculation unit 251 calculates the power spectrum by calculating the square of the absolute value of the FFT result. Then, it calculates the power spectral density by dividing the calculated power spectrum by the frequency resolution of the FFT.

[0038] Furthermore, when dividing time-series current data into short intervals of a certain length, each interval may have a length that is an integer multiple of the interval P1 shown in Figure 3. The power spectral density calculation unit 251 may reduce the effects of sharp changes in data at the ends of the intervals by applying a window function (such as a Hanning window) to the data in each interval. Alternatively, the power spectral density calculation unit 251 may average the power spectra of multiple intervals and calculate the power spectral density based on the averaged power spectra.

[0039] The method for calculating the power spectral density by the power spectral density calculation unit 251 is not limited to the examples given herein.

[0040] The formant band setting unit 253 sets a formant band that includes the frequency component where the power of the power spectral density is maximized.

[0041] The "formant band" is a frequency band in current data where energy is concentrated near a specific frequency band. In the formant band, the power spectral density is stronger than in other frequency bands.

[0042] The setting of formant bands will be described with reference to FIG. 4. FIG. 4 is a schematic diagram showing an example of power spectral density.

[0043] In FIG. 4, graph GF shows the relationship between frequency and power of power spectral density. For example, graph GF has maxima at several frequencies, reflecting characteristics included in time-series current data. For example, graph GF has maxima at formant frequencies M2 and M3. Furthermore, on both sides of formant frequency M2 and both sides of formant frequency M3, graph GF has locations where the power is minimal.

[0044] A formant band setting unit 253 identifies, as a formant frequency, a frequency at which the power of the power spectral density reaches a maximum. The formant band setting unit 253 also sets a frequency band including the formant frequency as a formant band.

[0045] For example, the formant band setting unit 253 may identify frequencies at which the power of the power spectral density becomes minimal before and after the formant frequency, and set a frequency band having the minimum-power frequencies as the lower limit and upper limit as the formant band.

[0046] According to FIG. 4, a formant band FW2 is set corresponding to formant frequency M2, and a formant band FW3 is set corresponding to formant frequency M3. As described above, the formant band setting unit 253 may set, as a formant band, a frequency band including the n-th lowest frequency component (where n is an integer of 1 or more) among frequency components at which the power of the power spectral density is maximum.

[0047] In addition, the formant band setting unit 253 may set the formant band based on the smoothed power spectral density. For example, the formant band setting unit 253 may calculate the smoothed power spectral density by averaging a plurality of temporally continuous power spectral densities. Alternatively, the formant band setting unit 253 may calculate the smoothed power spectral density by performing a moving average in frequency space.

[0048] FIG. 5 is a schematic diagram showing an example of the power spectral density when partial discharge starts. Unlike the power spectral density shown in FIG. 4, according to the power spectral density shown in FIG. 5, fine vibrations occur in the range TP, and it can be seen that a plurality of peaks are generated around the formant frequencies M2 and M3. When the power spectral density with such fine vibrations remaining is used as the processing target, there is a risk that the formant frequencies and the formant bands cannot be accurately specified.

[0049] Therefore, the formant band setting unit 253 may smooth the power spectral density to remove fine vibrations. The formant band setting unit 253 may specify the formant frequencies and the formant band based on the smoothed power spectral density.

[0050] The determination unit 255 determines whether or not partial discharge has occurred within the set formant band. The determination unit 255 also determines whether or not the maximum indicated by the formant frequency has disappeared within the set formant band.

[0051] Here, the "formant band" used in the determination by the determination unit 255 is set based on the power spectral density calculated at a predetermined timing. If the power spectral density fluctuates after the formant band is set, the formant band may also fluctuate, but the formant band after fluctuation is not used for determination. For example, the determination unit 255 may use a formant band set based on the initially calculated power spectral density.

[0052] The determination unit 255 may count the number of peaks in the power spectral density in the formant band and determine that the timing at which the number of peaks first exceeds a predetermined threshold is the timing at which partial discharge has begun.

[0053] For example, according to the power spectral density shown in Figure 5, fine oscillations occur in the range TP, and multiple peaks are generated around the formant frequencies M2 and M3. The determination unit 255 counts the number of peaks in the formant band FW2 and determines whether the number of peaks is above a predetermined threshold. The determination unit 255 may then determine that partial discharge is occurring if the number of peaks is above the predetermined threshold.

[0054] The determination unit 255 may determine the occurrence of partial discharge based on the number of peaks in the power spectral density, or it may determine the occurrence of partial discharge in the winding by a method other than those listed herein.

[0055] Furthermore, the determination unit 255 may determine that the timing at which the frequency of the frequency component related to the maximum that was included in the formant band becomes higher than the frequency of the other frequency components included in the formant band is the timing at which the maximum has disappeared.

[0056] Figure 6 is a schematic diagram showing an example of the power spectral density when the maximum disappears. When partial discharge is repeated in the winding, the formant frequency corresponding to the maximum included in the formant band of interest tends to increase over time.

[0057] For example, when partial discharge is repeated, the power spectral density shown in Figure 5 changes to the power spectral density shown in Figure 6. The formant frequency M2, which was included in the formant band FW2 shown in Figure 5, increases over time and becomes higher than the frequency included in the formant band FW2. As a result of the formant frequency M2 falling outside the formant band FW2, it is not possible to identify the formant frequency M2 in Figure 6.

[0058] In addition, the determination unit 255 may output the determination result of partial discharge occurrence in the winding, or the determination result of maximum annihilation that was included in the formant band, via the output unit 23.

[0059] The elapsed time acquisition unit 257 acquires the elapsed time in a set formant band from the time a partial discharge occurs until the maximum disappears. More specifically, the elapsed time acquisition unit 257 acquires the time from the start timing corresponding to the power spectral density in which the first partial discharge was determined to have occurred, to the end timing corresponding to the power spectral density in which the maximum disappeared, as the elapsed time.

[0060] Figure 7 is a schematic diagram showing an example of the time evolution of formant frequency. In Figure 7, the horizontal axis represents time and the vertical axis represents formant frequency, plotting the formant frequencies M2 and M3 specified at predetermined timing intervals.

[0061] As shown in Figure 7, formant frequencies M2 and M3 are shown in time period T1, while in time period T2 only formant frequency M3 is shown, indicating that formant frequency M2 could not be identified. It is assumed that partial discharge occurred at the start of time period T1, and insulation breakdown of the winding occurred at the end of time period T2.

[0062] When focusing on the formant band FW2, the elapsed time acquisition unit 257 acquires the length of the time period T1 as the elapsed time from when a partial discharge occurs until the maximum disappears.

[0063] The prediction time determination unit 259 determines a value obtained by multiplying the elapsed time by a predetermined coefficient as the prediction time until dielectric breakdown of the winding. The predetermined coefficient may vary depending on the type of winding, the environment in which the winding is placed, etc.

[0064] For example, according to Figure 7, the total time for time zones T1 and T2 is approximately twice the elapsed time determined based on time zone T1. Therefore, typically, the predetermined coefficient may be set to approximately 2.

[0065] For example, in order to set a predetermined coefficient, time-series test current data related to the test current flowing through the test winding may be acquired when the voltage is repeatedly applied to the test winding until dielectric breakdown occurs. The period from when the application of voltage to the test winding is started until dielectric breakdown occurs in the test winding may also be acquired. A predetermined coefficient may be set in advance based on the power spectral density, formant bandwidth, and formant frequency calculated based on the test current data.

[0066] Alternatively, the prediction time determination unit 259 may output the prediction time via the output unit 23.

[0067] [Judgment Processing by Diagnostic Device] Figure 2 is a flowchart showing the procedure for the judgment processing. The judgment processing shown in Figure 2 is performed on the winding to be measured.

[0068] In step S101, the measuring device 10 repeatedly applies an impulse-like voltage to the winding to be measured.

[0069] In step S103, the measuring device 10 acquires time-series current data relating to the current flowing through the winding when a repeatedly impulse-like voltage is applied to the winding to be measured.

[0070] In step S105, the power spectral density calculation unit 251 calculates the power spectral density, which indicates the power of each frequency component contained in the current, based on the current data.

[0071] In step S107, the formant band setting unit 253 sets the formant band based on the power spectral density.

[0072] In step S109, the determination unit 255 determines the timing at which partial discharge began to occur and the timing at which the maximum that was included in the formant band disappeared.

[0073] In step S111, the elapsed time acquisition unit 257 acquires the elapsed time in the set formant band from the occurrence of partial discharge until the maximum disappears.

[0074] In step S113, the prediction time determination unit 259 determines the predicted time until dielectric breakdown of the winding.

[0075] [Effects of the Embodiment] As described in detail above, the diagnostic device, diagnostic method, and diagnostic program according to this disclosure use a controller that is connected to an input unit that receives time-series current data relating to the current flowing through the winding when a repeatedly impulsive voltage is applied to the winding, and processes the current data. Based on the current data, the controller calculates the power spectral density, which shows the power of each frequency component contained in the current, and sets a formant band that includes the frequency component in which the power of the power spectral density is maximum. In the set formant band, the controller obtains the elapsed time until the maximum disappears, and determines the value obtained by multiplying the elapsed time by a predetermined coefficient as the predicted time until dielectric breakdown of the winding.

[0076] This allows for the necessary diagnostics to avoid dielectric breakdown in windings. Furthermore, since it can calculate the predicted time until dielectric breakdown in the winding, it is possible to determine if the winding being measured is in a state where dielectric breakdown is highly likely to occur before it actually happens.

[0077] Furthermore, in windings to which repeated impulse voltages due to surge voltages are applied, it becomes possible to take necessary measures to prevent dielectric breakdown of the windings. As a result, the quality of equipment including windings is improved. Moreover, the burden (cost, workload) of operating equipment including windings can be reduced.

[0078] The controller may set the formant band to include the frequency band containing the nth lowest frequency component (where n is an integer greater than or equal to 1) among the frequency components where the power of the power spectral density is maximum. This improves the accuracy of calculating the predicted time until dielectric breakdown in the winding.

[0079] The controller may set the formant bandwidth based on the smoothed power spectral density. This improves the accuracy of identifying the formant frequency and formant bandwidth. In particular, even when fine oscillations occur in the power spectral density due to partial discharge or the like, the formant frequency and formant bandwidth can be reliably identified.

[0080] The controller may count the number of peaks in the power spectral density within the formant band and determine that the timing when the number of peaks first exceeds a predetermined threshold is the timing when partial discharge begins. This allows for the determination of partial discharge in the winding. Furthermore, the elapsed time can be determined with high accuracy. As a result, the accuracy of calculating the predicted time until dielectric breakdown in the winding can be improved.

[0081] The controller may determine that the moment when the frequency of the frequency component related to the maximum becomes higher than the frequency included in the formant band is the moment when the maximum disappears. This makes it possible to capture the tendency for the formant frequency corresponding to the maximum included in the formant band of interest to increase over time as partial discharge is repeated in the winding. Furthermore, the elapsed time can be determined with high precision. As a result, the accuracy of calculating the predicted time until dielectric breakdown in the winding can be improved.

[0082] When voltage is repeatedly applied to the test winding until dielectric breakdown occurs, a predetermined coefficient may be set based on the time-series test current data related to the test current flowing through the test winding, and the period from the start of voltage application until dielectric breakdown occurs. This allows the predetermined coefficient to be set while capturing the characteristics of dielectric breakdown in the test winding. As a result, it is possible to determine whether there is a high probability of dielectric breakdown occurring in the winding being measured before dielectric breakdown actually occurs in the winding being measured.

[0083] Each of the functions described in the embodiments above may be implemented by one or more processing circuits. These processing circuits may include programmed processors, electrical circuits, and further may include devices such as application-specific integrated circuits (ASICs), or circuit components arranged to perform the described functions.

[0084] According to this disclosure, it becomes possible to take the necessary measures to avoid dielectric breakdown of windings when repeated impulse voltages due to surge voltages are applied to the windings. As a result, the quality of equipment including windings is improved. Therefore, for example, it can contribute to United Nations Sustainable Development Goal (SDG) 9, "Build resilient infrastructure, promote inclusive and sustainable industrialization and foster innovation."

[0085] Although several embodiments have been described, it is possible to modify or transform the embodiments based on the above disclosure. All components of the above embodiments, and all features described in the claims, may be taken individually and combined, provided that they do not conflict with each other.

[0086] The entire contents of Japanese Patent Application No. 2025-053219 (Filing Date: March 27, 2025) are incorporated herein by reference.

[0087] 10 Measuring device 20 Diagnostic device 21 Input unit 23 Output unit 25 Controller 27 Operation unit 251 Power spectral density calculation unit 253 Formant band setting unit 255 Judgment unit 257 Elapsed time acquisition unit 259 Prediction time determination unit

Claims

1. A diagnostic device comprising: an input unit that receives time-series current data relating to the current flowing through a winding when a repeatedly impulsive voltage is applied to the winding; and a controller that processes the current data, wherein the controller calculates a power spectral density indicating the power of each frequency component included in the current based on the current data; sets a formant band that includes the frequency component in which the power of the power spectral density is maximum; obtains the elapsed time from the occurrence of partial discharge to the disappearance of the maximum in the set formant band; and determines a value obtained by multiplying the elapsed time by a predetermined coefficient as the predicted time until dielectric breakdown of the winding.

2. The diagnostic device according to claim 1, wherein the controller sets the formant band to include the frequency band that has the nth lowest frequency (where n is an integer of 1 or more) among the frequency components in which the power of the power spectral density is maximized.

3. The diagnostic apparatus according to claim 1, wherein the controller sets the formant bandwidth based on the smoothed power spectral density.

4. The diagnostic device according to claim 1, wherein the controller counts the number of peaks of the power spectral density in the formant band, and determines that the timing at which the number of peaks first exceeds a predetermined threshold is the timing at which the partial discharge begins to occur.

5. The diagnostic device according to claim 1, wherein the controller determines that the timing at which the frequency of the frequency component relating to the maximum becomes higher than the frequency included in the formant band is the timing at which the maximum disappears.

6. The diagnostic device according to any one of claims 1 to 5, wherein the predetermined coefficient is set based on time-series test current data relating to the test current flowing through the test winding and the period from the start of voltage application until dielectric breakdown occurs, when the voltage is repeatedly applied to the test winding until dielectric breakdown occurs.

7. A diagnostic method for controlling a controller that processes current data, which is connected to an input unit that receives time-series current data relating to the current flowing through a winding when a repeatedly impulsive voltage is applied to the winding, wherein the controller calculates a power spectral density indicating the power of each frequency component included in the current based on the current data, sets a formant band that includes the frequency component in which the power of the power spectral density is maximum, obtains the elapsed time from the occurrence of partial discharge to the disappearance of the maximum in the set formant band, and determines a value obtained by multiplying the elapsed time by a predetermined coefficient as the predicted time until dielectric breakdown of the winding.

8. A diagnostic program executed in a controller that processes current data and is connected to an input unit that receives time-series current data relating to the current flowing through a winding when a repeatedly impulsive voltage is applied to the winding, the diagnostic program comprising: a step of calculating a power spectral density indicating the power of each frequency component included in the current based on the current data; a step of setting a formant band that includes the frequency component in which the power of the power spectral density is maximum; a step of obtaining the elapsed time from the occurrence of partial discharge to the disappearance of the maximum in the set formant band; and a step of determining a value obtained by multiplying the elapsed time by a predetermined coefficient as the predicted time until dielectric breakdown of the winding.

Images