TW201835598A - Diagnostic device and diagnostic method - Google Patents

Abstract

Provided is a diagnostic device for highly accurately diagnosing a rotating machine even when a diagnosis is based on a small amount of current data. This diagnostic device is provided with a current measurement unit for measuring current flowing through at least two locations in a rotating machine and a diagnostic unit for diagnosing, on the basis of current data output by the current measurement unit, the state of the rotating machine and a peripheral device electrically or mechanically connected to the rotating machine. The diagnostic unit creates a Lissajous curve distribution map by layering multiple periods of the two types of current data obtained by the current measurement unit and diagnoses the state of the rotating machine system from the results of evaluating the distribution map.

Description

Diagnostic device and method

The invention relates to a diagnostic device and a diagnostic method.

If a rotating machine such as a motor (electric motor) or a generator incorporated into the production equipment fails suddenly, the rotating machine needs unplanned repair or replacement operations, resulting in a decrease in the operating rate of the production equipment or the need to redo the production plan. Similarly, even if the power conversion device or cable connected to the rotating machine fails, unplanned repair work or replacement work is required, resulting in a decrease in the operating rate of production equipment or the need to redo the production plan. In order to prevent the sudden failure of the rotating machine system (the rotating machine and its attached devices (cables, power conversion devices)), you can stop the rotating machine system properly and perform offline diagnosis to grasp the degree of deterioration and prevent it to a certain extent. Sudden failure. However, in order to perform offline diagnosis, it is necessary to stop the rotating machine system, which will cause the operating rate of the production equipment to decrease. In addition, depending on the type of degradation, there is degradation that occurs only when a voltage is applied. Therefore, there is a need to diagnose the state of the rotating machine based on the information of the current of the rotating machine system. As a prior art related to diagnosis based on current information of a rotating machine system, there is Non-Patent Document 1. In Non-Patent Document 1, a method called Motor Current Signature Analysis (MCSA, Motor Current Signal Analysis) can be used to diagnose the damage of the rotor rod, the eccentricity of the rotor, and the corresponding specific frequency spectrum detection. Damage to the core of the stator, short circuit of the winding wire, deterioration of the bearing, etc. In addition, Patent Document 1 discloses a method for obtaining vibration sensor data at two locations, particularly in bearing diagnosis, and based on the trajectory of the Lissajous figure drawn on the axis according to the instantaneous value of the data of each sensor. Changes in slope or radius to determine anomalies. [Preceding Technical Documents] [Patent Documents] [Patent Documents 1] Japanese Patent Laid-Open No. 2000-258305 [Non-Patent Documents] [Non-Patent Documents 1] TAKADA TECHNICAL REPORT Vol.20 2010 Analysis of Motor Driven by MCSA Using Current Symptoms Machine diagnostic technology

[Problems to be Solved by the Invention] However, the above-mentioned non-patent literature 1 and patent literature 1 have the following problems. In the technique disclosed in the above-mentioned Non-Patent Document 1, detection of a specific frequency spectrum is required. In order to detect a specific frequency spectrum with high accuracy, an expensive data logger is required for diagnosis corresponding to long-time measurement at a high sampling speed, and there is diagnosis. The problem of increased costs. In addition, when the driving conditions of the motor are changed during a long period of data measurement, the above-mentioned specific frequency spectrum appears unintentionally, and there is a problem of false alarm. Also, similarly, when the driving condition of the motor changes during a long period of data measurement, the fundamental wave frequency changes and there is a problem of underreporting when the frequency spectrum differs from the assumed frequency. In addition, in the technology disclosed in the above Patent Document 1, diagnosis is performed based on vibration sensor information, and it is necessary to install a vibration sensor at a position sensitive to a motor failure, which has a problem of limited installation places. In addition, a diagnostic sensor for diagnosis and an expensive data logger with a sufficient sampling speed to obtain the trajectory of the Lissajous figure are needed, and there is a problem that the diagnosis cost increases. An object of the present invention is to provide a diagnostic device capable of performing high-precision diagnosis even in a case where data is measured in a short time, low sampling speed, and a general-purpose machine. [Technical means to solve the problem] In order to solve the above-mentioned problems and achieve the purpose of the present invention, they are constituted as follows. That is, the diagnostic device of the present invention includes: a current measurement section that measures current flowing in at least two places of the rotating machine; and a classification section that sets the current data measured by the current measurement section according to each command value of the power conversion device Classification of information; and has a diagnosis section that divides the distribution of Lissajous figures obtained by superimposing multiple periods of current data of at least two phases in the classification section (here defined as connecting points not in time series with lines but rather The distribution drawn as a collection of points. It is not that the points connected in time series are different from the trajectory of Li Shayu's figure), compared with the distribution of Li Shayu's figure set as a normal state in advance, according to the change of the distribution of Li Shayu's figure, Machine or power conversion device. In addition, the diagnostic method of the present invention measures the current flowing in at least two places of the rotating machine, classifies the measured current data according to each command value information of the power conversion device, and classifies the classified current of at least 2 phases and a plurality of cycles. The distribution of the Lissajous figure created by overlapping data is compared with the distribution of the Lissajous figure set as a normal state in advance, and the state of the peripheral machine such as the rotary machine or the power conversion device connected to the rotary machine is diagnosed according to the change. The details of the structure of the other diagnostic method machine and the diagnostic method will be described in the embodiment. [Effects of the Invention] According to the present invention, no expensive data logger is required, and the state of the rotating machine system can be diagnosed even when the driving conditions of the motor are changed.

Hereinafter, a mode for implementing the present invention (hereinafter referred to as an “embodiment”) will be described with reference to the drawings as appropriate. In addition, the following is only an example of an embodiment, and it is not intended to limit the scope of the present invention to the following examples. Rotary machines equipped with electric motors (motors) or generators, cables attached to the rotary machines, and power converters have a lot of fault generating locations and factors. For example, consider insulation degradation, bearing degradation, short circuit, disconnection, water ingress, and so on. In addition, there are cases where the motor is installed in a harsh environment for a long time, and a diagnosis technique corresponding to the installation condition is required. An example of the previous diagnostic device 14 is shown in FIG. 2. In the conventional diagnostic device, the current measurement information of one phase was obtained by the current measurement unit 11, and the diagnosis was performed based on the value of a specific frequency spectrum of the frequency spectrum obtained by the Fourier transform in the diagnosis unit 12. Because the Fourier transform is used to measure changes in a specific frequency spectrum, continuous measurements need to be performed at a certain sampling rate. Therefore, it is necessary to increase the capacity of the memory for temporarily storing the measurement data, and to improve the communication speed of the device for storing and storing the data, and an expensive device is required. In addition, since no change in the control signal is assumed, there is a problem that the frequency of false positives and false negatives is high. The inventors have studied the sensor value of the two-phase discontinuity in the load current value of the rotating machine on a plane, and the state of the rotating machine is visualized as a data distribution diagram similar to Li Shayu's figure. The so-called Li Shayu pattern refers to a planar pattern obtained by combining two waves. The current sensor data of the three-phase motor is offset by 120 degrees from each other, so if the two phases are combined, it is an inclined elliptical shape. Usually, it is created by using continuous data, but the inventors and others overlap data obtained at a certain frequency obtained intermittently, and use the obtained data distribution for evaluation. As a result, the state of the rotating machine system can be diagnosed with high accuracy based on the discontinuous data, and equipment such as an expensive data logger can be omitted. The diagnostic device of this embodiment capable of solving the above problems includes a current measuring unit that measures current flowing in at least two places of the rotating machine; and a diagnosis unit that analyzes the rotating machine and the rotating machine based on the current data output from the current measuring unit. Diagnosis of the state of peripheral equipment that is electrically or mechanically connected to the rotating machine. In the diagnosis section, the two types of current data obtained by the current measurement section are overlapped for a plurality of periods to form a distribution chart of the Lissajous figure, and the state of the rotating machine system is diagnosed based on the evaluation results of the distribution chart. Furthermore, even when the device described above is not provided, by measuring the current flowing in at least two places of the rotating machine, the measured current data is classified according to each command value information of the power conversion device, and Each command value information overlaps the classified current data with a plurality of cycles to create a distribution chart of the Lissajous figure. The distribution map of the Lissajous figure created can be evaluated, and the status of the rotating machine system can be diagnosed based on the results. Moreover, the above-mentioned diagnostic device can also be incorporated into a rotating machine system. In particular, it is desirable to share a current sensor, a current measuring unit, and the like provided in the power conversion device for controlling the rotating machine, so that the number of parts can be reduced. Furthermore, a plurality of rotating machines may be connected to the power conversion device. Furthermore, by classifying and evaluating the current data according to each command value information of the power conversion device, the state of the rotating machine system can be appropriately diagnosed even when the driving conditions of the rotating machine are changed. Specifically, a rotating machine system is provided, which includes: one or more rotating machines; and a power conversion device electrically connected to the rotating machine to control the current flowing through the rotating machine; the power conversion device has at least There are two current measuring sections for measuring the current flowing, and a control section for outputting a command value for controlling the rotating machine. The above rotating machine system further includes a diagnosis section that diagnoses the electrical or mechanical connection of the rotating machine. State; and a classification unit that classifies the current data output from the current measurement unit and input to the diagnosis unit according to each of the above-mentioned command values; and the diagnosis unit creates the current data of the rotating machine with at least two phases and a plurality of cycles overlapping and creates The distribution of Li Shayu's graphics. Hereinafter, in this embodiment, the following example will be described: each of the instructions includes a current measurement unit that measures current flowing in at least two places of the rotating machine, and current data measured by the current measurement unit according to each instruction of the power conversion device. A classification section of value information classification, and a diagnosis section, which diagnoses the distribution of the Lissajous figure obtained by superimposing a plurality of cycles of the current data of at least 2 phases of the classification section (here, it is defined as that each point is not a line in time series) The connection is a distribution drawn as a collection of points. It is not that the connection of points in time series is different from the trajectory of Li Shayu graphics), and the distribution of Li Shayu graphics set as a normal state in advance is compared, and according to the distribution of Li Shayu graphics Changes to diagnose the status of the rotating machine or power conversion device. Embodiment 1 FIG. 1 shows a configuration diagram of a diagnostic apparatus according to Embodiment 1, and describes a diagnostic apparatus and a diagnostic method. The common parts with the description of FIG. 2 are omitted. In the first embodiment, the power source 1, the cable 2, and the power conversion device 7 are electrically connected, and a three-phase AC voltage is output to the power conversion device 7. The output of the three-phase AC voltage is controlled by adjusting the operation timing of the switching elements of the inverter in such a way that the number of revolutions and torque of the motor becomes the desired value. This control is determined based on preset control information and information on the current output from the inverter. The current information is obtained by the current sensors 4a and 4b and the current measurement unit 9, and is fed back to the control unit 8. The current information of a plurality of cycles of the current sensors 4a and 4b of the current measurement section is input to the classification section, and the diagnosis section obtains the distribution of the Lissajous figure drawn by taking the current sensors 4a and 4b as orthogonal axes. Diagnostic motor system. The current information of the current sensors 4a and 4b obtained by the current measurement section does not have to be a fixed sampling interval, and it does not have to be a continuous measurement. The interval between the data acquisition of the current sensor 4a and the data acquisition of the current sensor 4b is preferably fixed to allow a certain deviation. It is possible to set a fixed data acquisition interval that allows a certain deviation by applying a measurement device that is excellent in real-time processing. An example is a measurement device using a personal computer. Furthermore, by setting the interval between the data acquisition of the current sensor 4a and the data acquisition of the current sensor 4b to be fixed, continuous measurement is not required. For example, the current measurement unit 9 such as storing a certain amount of data in the memory of a personal computer and inserting the data communication processing to the memory device, clearing the memory, and restarting the data storage can be implemented, so that the universal use Current sensor, memory system. In the following, as an example of a diagnostic method using the diagnostic device of the first embodiment, a method of diagnosing a specific frequency to generate a frequency spectrum when the motor is operated at a fixed fundamental wave frequency without changing the control pattern is described to explain the diagnosis Ministry of Functions. Figure 3 shows the U-phase current waveform. Fig. 3a is a diagram showing the appearance of a frequency spectrum with a difference of 1 Hz from the fundamental wave frequency of the U-phase current at 50 Hz. For example, when a bearing is degraded, a sideband wave of a fundamental wave frequency is generated due to the deterioration. In this case, as shown in FIG. 3b, the U-phase current appears as a waveform having a difference frequency with a period of 1 Hz. In the conventional diagnostic device, in order to separate the components of 50 Hz and 51 Hz from the waveform with good accuracy, when measuring at a frequency of 200 Hz, it is necessary to continuously measure 20,000 points of data for at least 100 seconds. Therefore, a measurement device equipped with a large-capacity memory corresponding to a measurement of 20,000 points or more, or a measurement device capable of writing to a memory device at 200 Hz is required. In addition, if measurement errors are considered, it is effective to increase one or both of the sampling speed and the measurement data length. Therefore, a special and expensive measuring device is required for accurate separation. On the other hand, in this embodiment, by using any two kinds of current values for diagnosis, the diagnosis of the rotating machine can be performed without increasing the sampling speed and the length of the measurement data. FIG. 4 shows current waveforms of the U-phase and W-phase in a normal state, that is, a 50 Hz sine wave without a difference frequency. The phase difference between the U-phase and W-phase currents is 120 degrees. On the other hand, FIG. 5 is a conceptual diagram showing the U-phase and W-phase current waveforms when a sideband wave is generated in the same manner as in FIG. 3. The sampling speed is 100 Hz, and the data measurement time is 1 second. As shown in FIGS. 4 and 5, based on the current waveforms obtained by high-speed sampling, the U-phase current is placed on the horizontal axis and the W-phase current is placed on the vertical axis. FIG. 6a is an example based on the normal state of FIG. 4, and FIG. 6b is a conceptual diagram based on the case where a sideband wave is generated due to deterioration of a rotating machine or the like in FIG. By comparing FIG. 6A and FIG. 6B, it can be seen that in the degraded state, the distribution of the Lissajous pattern becomes thicker than in the normal state, and as the tendency of the deterioration progresses, the thickness of the Lissajous pattern distribution changes. Fig. 7 shows an example of the distribution of a Lisa Yu pattern for each of the normal and degraded states at U-phase and W-phase currents obtained at a sampling speed of 4.975 Hz (Fig. 7a: Normal, Fig. 7b: Degraded). The measured data length is 1000 points, and the data length is consistent with Figure 6. Figure 6 and Figure 7 show the distribution of Lissajous patterns that are approximately the same. Therefore, by evaluating the distribution of the two-phase data, deterioration can be detected regardless of the sampling speed. The difference between the degraded state and the normal state can be confirmed by human eyes, and the difference in the distribution of Li Shayu's figures can also be compared using mechanical learning and other methods. That is, according to this embodiment, the generation of a difference frequency can be easily detected in the case of high-speed sampling or the case of slower sampling frequency. Furthermore, when the sampling speed is slow, it is desirable to perform the measurement asynchronously with the fundamental frequency. Specifically, at a sampling speed with a period that is an integer multiple of the fundamental wave, only a certain phase value is always obtained, so there is a danger of underreporting in the case of a degradation that changes only at a specific phase. Therefore, the sampling speed is preferably a frequency different from an integer multiple of the period of the fundamental wave. As a result, it is possible to obtain a diagnostic profile similar to the case where data is acquired at a high sampling rate. In addition, in order to obtain high-frequency information, it is desirable to perform measurement asynchronously with the timing of the switches of the power conversion device. Furthermore, if the points of the Li Shayu figure are filled with lines in time series (the trajectory of the Li Shayu figure), the interior of the distribution of the Li Shayu figure is filled with the line, and there may be no difference between the normal state and the degraded state. Ideally, they should be displayed in trajectories instead of being connected in time series. The diagnosis section 10 outputs the results. The method of notifying the user of the diagnosis result may be appropriately selected, and as a method of communicating to the user, in addition to the display on the display, the lighting of the indicator light, the notification of the mail, etc. may be mentioned. The content also considers (1) the method of displaying the distribution of Li Shayu graphics on the screen for the user to judge whether there is a corresponding method, (2) the method of notifying the user of the difference in the distribution of Li Shayu graphics by some method, and (3) A method of notifying the user when a predetermined threshold is exceeded. As a method of digitizing the difference in the distribution of Li Shayu's graphics (2) above, consider the application of mechanical learning. As the algorithm of mechanical learning, it is only necessary to select one that makes the difference between Li Shayu's graphics clear, such as the local subspace method. The local subspace method is as follows: For all points in the distribution of the Lissajous pattern of the diagnosis object, the nearest 2 points are selected from the distribution of the Lissajous pattern defined as the normal state, and according to the straight line connecting the two points and the diagnosis object The distance between points, and defines the degree of degradation. In addition to the method of calculating the distance of all points of the diagnosis object and digitizing the changes in the distribution of the Lissajous figure, the average value of the distances of all the points of the diagnosis object, the digitization of the distance of only the points with a specific phase, etc. can also be used. Select arbitrary evaluation method for the deviation error of the waveform. When computing speed is a priority, a clustering method such as vector quantization clustering or K-means clustering can be used. In addition, a method called deep neural network that can automatically find a feature based on a large amount of data can be applied. Next, the case where the control pattern changes will be described. When the control pattern changes, the fundamental wave frequency and torque of the motor also change, so the previous method sometimes diagnoses it as abnormal. In addition, if the state of the control pattern change is included as the normal state learning, there may be cases where changes due to deterioration are missed and missed. Therefore, it is desirable to diagnose normal and abnormalities for each control pattern. In this embodiment, the control instruction of the control unit is combined with the current information to improve the diagnostic accuracy. The diagnosis unit 10 diagnoses the state of the motor system based on the distribution of the Lissajous patterns drawn by the U-phase and W-phase currents that classify the same control commands. Therefore, a classification unit 6 is provided as described in FIG. 1. The function of combining the classification unit 6 and the diagnosis unit 10 will be described below. As the control command, a voltage command value, a current command value, an excitation current command value, a torque current command value, a speed command value, a frequency command value, and the like can be arbitrarily selected from the command values that the power conversion device 7 can output. The classification unit 10 does not necessarily use all the command values that can be output by the power conversion device 7, and may use only the command values with high sensitivity to the degradation of the detection target. The so-called high-sensitivity command value can be selected by comparing the distribution of the Lissajous pattern in the degraded state with the distribution of the Lissajous pattern in the normal state, and it is easy to see the difference between the degraded state and the normal state. Hereinafter, the operation of the classification unit 6 and the diagnosis unit 10 will be described based on a pattern diagram of the current waveforms when the motor is driven under the two conditions of the control command A and the control command B having different voltage command values and frequency command values as control commands. The classification unit 6 assigns the current information obtained by the current measurement unit 9 based on the control instruction values (control instruction A and control instruction B) obtained by the control unit 8. Based on the allocation by the classification unit 6, the diagnosis unit 10 saves the current information and the command value information at intervals of about one minute, and creates a distribution chart of the Lissajous pattern for each control command value. FIG. 8 shows the distribution of the Lissajous patterns in normal and degraded states classified by each control instruction. Figures 8a and 8c are the results of the distribution of the Lissajous graphics depicted under the control instruction A, and Figures 8b and 8d are the results of the distribution of the Lissajous graphics depicted under the control instruction B. As is clear from the comparison between FIG. 8a and FIG. 8b, in the case of the control instruction A and the control instruction B, even under normal conditions, the obtained Lissajous figures are different. Therefore, if only the normal state of the control instruction A (Fig. 8a) is taken as the normal learning, and the distribution of the Lissajous pattern of the control instruction B with different control instructions (Fig. 8b) is made, the change will be misreported as a change due to degradation. Therefore, by classifying the current waveform based on the control information, false alarms can be suppressed. By classifying the current information obtained by the current measurement unit into the current information related to the control instruction A and the current information related to the control instruction B, each of the control instruction A and the control instruction B is delineated into two types: normal state and degraded state. The distribution of the graph. As a result, if the normal (Figure 8a) and the degradation (Figure 8c) of the control instruction A are compared, it can be seen that the difference in the distribution of the Li Shayu pattern can be seen and the degradation can be detected. In addition, if the control instruction B is focused on, the normal (Fig. 8b) and the deterioration (Fig. 8d) of the control instruction B are compared, and it can be seen that the difference in the distribution of the Li Shayu pattern can be used to detect the deterioration. Next, a case where the diagnosis target data X obtained when the control instruction A is evaluated based on the result that the control instruction A of FIG. 8a is normally learned as a normal state is taken as an example to describe a method of digitizing a change in the Lissajous figure. FIG. 9 shows the learning data Y (control command A is normal) near the diagnosis object data X (control command A is normal or degraded) in the Li Shayu graph. Find the data that is closest to the diagnosis object's data X (control instruction A is normal or degraded), the learning data Y1 (control instruction A is normal), and the second closest learning data Y2 (control instruction A is normal), according to The degree of abnormality is defined by the distance between the straight line of the data of Y1 and Y2 and the data of the diagnosis target's data X (the control command A is normal or degraded). By implementing the measurement of the data length of the above content, the degree of abnormality of the data group X1 to n (control command A is normal or degraded) relative to the learning data group Y can be quantified. FIG. 10 shows the judgment and evaluation results of the diagnosis results of the normal state and the abnormal state. It indicates the average value of the abnormality of the normal current data under the control instruction A and the average value of the abnormality of the current data when the machine is degraded under the control instruction A. The average value of abnormality increases due to deterioration. If the threshold value is specified in advance through research, an increase in abnormality, that is, progress of deterioration can be displayed to the user. Furthermore, in the first embodiment, the case where a frequency spectrum is generated at a specific frequency is described, but even if the frequency spectrum is not degraded for a specific frequency, the current will appear due to the load or impedance change of the motor due to the degradation. Some changes can be detected by the diagnostic device and diagnostic method of Embodiment 1. As the deterioration that does not occur with a change in a specific spectrum, it is specifically assumed that deterioration other than the deterioration that occurs as a peak of a specific frequency that can be detected by the MCSA, that is, such as deterioration of grease, thermal deterioration of an insulating material, and moisture absorption. In addition, according to the diagnostic device of Embodiment 1, in addition to the motor, the motor system including a cable, a power conversion device, a load, and the like that are electrically or mechanically connected to the motor can be diagnosed. In a motor system including peripheral equipment connected to the motor, even if the peripheral equipment other than the motor fails or deteriorates, the current flowing in the motor will change due to changes in the impedance or load of the other equipment, so the degradation can be detected by this method. . Embodiment 2 FIG. 11 is an example of a diagnosis system including a command unit 13. The instruction unit 13 sets a control instruction to the control unit 8 and changes the control instruction. In the first embodiment, the information of the control unit 8 is input to the classification unit for diagnosis, but the information of the command unit 13 may be input to the classification together with the control command value obtained by the control unit 8 or in place of the control command value. Department for diagnosis. As shown in FIG. 11, when a plurality of types of information such as a rotating load are stored, and a change of a control command value using time is stored in the command section 13, the information of the command section 13 may be input to the classification section for diagnosis. . Thereby, diagnostic accuracy can be improved. Embodiment 3 In Embodiments 1 and 2, an example in which one motor 3 is connected to one power conversion device 7 has been described. In the third embodiment, as shown in FIG. 12, an example in which a plurality of motors 3 are connected to one power conversion device 7 will be described. In FIG. 12, a current sensor in the power conversion device is used to perform diagnosis by using a result of measuring all load currents. The method of diagnosis is the same as that in the case where one rotary machine is connected, and the diagnostic data measured and the classification of the use control command value and the reflection of the command value information are used as the distribution of the Lissajous graph to diagnose. Furthermore, since the current signals of multiple rotating machines are diagnosed as one load current, the change in the current waveform of a degraded motor is thinner than the current waveforms of other multiple motors, so it is more desirable to use machinery Learn, especially the local subspace method, to evaluate the changes in the distribution of Li Shayu's figures. Example 4 In Examples 1 to 3, the current sensor and the current measurement unit in the power conversion device were used, but as shown in FIG. 13, the current sensors 4a, 4b, and The current measuring section 9 can also obtain the effect of the present invention. There is no need to perform long-term data measurement corresponding to high sampling speed, so it is possible to use inexpensive general-purpose measurement equipment. Furthermore, in the third embodiment, when a plurality of motors are connected to one power conversion device in FIG. 12, the current sensor and the current measurement unit in the power conversion device are used, but it may also be used in a plurality of motors. Each of them is provided with a current sensor. In addition, a current sensor independent of the power conversion device may be provided at a position where the motor is connected to the power conversion device to measure all load currents. When measuring the load current of a plurality of motors by using a plurality of current sensors, a classification section and a diagnosis section may be provided in each rotating machine, and a single classification section and a diagnosis section may be used to diagnose and process the information of the plurality of sensors. . Embodiment 5 Embodiment 5 describes an example in which three current sensors are respectively provided for a zero-phase current and a two-phase load current. In Examples 1 to 4, the state of the motor system was diagnosed based on the information of the two current sensors, but as shown in FIG. 14, the distribution of the Li Shayu pattern can also be obtained based on the information of three or more current sensors. In this case, when performing a distribution comparison, two or more combinations can be selected from three or more current sensors to obtain the distribution of the Lissajous figure for evaluation. In addition, three or more current sensors may be taken as orthogonal axes, and the change in the distribution of the Lissajous pattern in a three-dimensional or more-dimensional space may be evaluated by mechanical learning, such as a local subspace method. As a current value measured by a current sensor, in the case of a three-phase AC rotating machine, in addition to the load current of each of the U-phase, V-phase, and W-phase, the three-phase clamp is arbitrarily selected to measure zero. Diagnostic sensitivity changes such as phase current, two-phase current measured by arbitrarily selected two phases, leakage current measured by clamping both the start winding and final winding of the motor winding, and the current flowing from the motor to ground. High current, and set a current sensor at this position.

1‧‧‧ Power

2‧‧‧ cable

3‧‧‧ Motor

4a‧‧‧Current sensor

4b‧‧‧Current sensor

4c‧‧‧Current sensor

5‧‧‧Extraction Department

6‧‧‧Classification Department

7‧‧‧Power Conversion Device

8‧‧‧Control Department

9‧‧‧Current measurement department

10‧‧‧Diagnosis Department

11‧‧‧Current Measurement Department

12‧‧‧Diagnosis Department

13‧‧‧Command Department

14‧‧‧ previous diagnostic device

FIG. 1 is a configuration diagram of a diagnostic device of Embodiment 1 FIG. 2 is a configuration diagram of a diagnostic device of a previous method. 3a and b are diagrams showing examples of U-phase current waveforms. 4a and b are diagrams showing current waveforms of the U-phase and the W-phase in a normal state. 5a and 5b are diagrams showing current waveforms of U-phase and W-phase in a deteriorated state. Figures 6a and 6b are diagrams showing the distribution (high-speed sampling) of the Lissajous pattern in normal and degraded states. Figures 7a and b are diagrams showing the distribution (low-speed sampling) of the Lissajous pattern in normal and degraded states. 8a to 8d are diagrams showing the distribution of the Lissajous patterns in the normal state and the degraded state classified by each control instruction. FIG. 9 is a diagram showing a diagnosis method using mechanical learning. FIG. 10 is a graph showing the diagnosis results of Example 1. FIG. FIG. 11 is a configuration diagram of the second embodiment. Fig. 12 is a configuration diagram of the third embodiment. FIG. 13 is a configuration diagram of the fourth embodiment. Fig. 14 is a configuration diagram of the fifth embodiment.

Claims (12)

A diagnostic device for a rotating machine system, comprising: a current measuring unit that measures current flowing in at least two places of the rotating machine; and a diagnosis unit that diagnoses the above based on current data output from the current measuring unit The state of the rotating machine and peripheral devices electrically or mechanically connected to the rotating machine; and the diagnosis section overlaps the two types of current data obtained by the current measuring section with a plurality of cycles to make a distribution chart of the Lissajous figure, and according to the above distribution The evaluation results of the chart diagnose the state of the rotating machine system. For example, the diagnostic device of the rotating machine system of claim 1 includes a classification unit that classifies the current data according to each command value information of the power conversion device. The power conversion device is electrically connected to the rotation and is controlled by The current flowing through the rotating machine, and the diagnosis section uses current data of at least two phases classified by the classification section. For example, the diagnostic device of the rotating machine system of claim 1 or 2, wherein the above-mentioned diagnosis section compares the distribution of the Lissajous pattern created above with the distribution of the Lissajous pattern set as a normal state in advance to diagnose the presence or absence of the abnormality. For example, the diagnostic device for a rotating machine system according to claim 3, wherein the diagnosis unit numerically diagnoses a change in the distribution of the Lissajous pattern created above with respect to the distribution of the Lissajous pattern previously set as a normal state. The diagnostic device for a rotating machine system according to any one of claims 1 to 4, wherein the above-mentioned current measurement unit is provided in a power conversion device, and the power conversion device is electrically connected to the rotating machine and is controlled by the rotating machine. Current. The diagnostic device for a rotating machine system according to any one of claims 1 to 5, wherein the current measuring section measures current data at a frequency different from an integer multiple of a period of a fundamental wave of the rotating machine. A diagnostic method of a rotating machine system, characterized in that the rotating machine system includes one or more rotating machines and a power conversion device, the power conversion device is electrically connected to the rotating machine, and controls the current flowing through the rotating machine. The above diagnosis method measures the current flowing in at least two places of the rotating machine, classifies the measured current data according to each command value information of the power conversion device, and classifies each of the command value information according to the above command value information. The current data is superimposed on a plurality of cycles to make a distribution chart of the Lissajous figure, and based on the distribution chart of the Lissajous figure created above, the state of the above-mentioned rotating machine system is diagnosed. The diagnostic method of the rotating machine system as claimed in claim 7, wherein at least two-phase current data is used. For example, the diagnosis method of the rotating machine system according to claim 7, wherein the above-mentioned diagnosis compares the distribution of the Lissajous pattern prepared above with the distribution of the Lissajous pattern set as a normal state in advance. For example, the diagnosis method of the rotating machine system according to claim 9, wherein the diagnosis is a digitization of a change in the distribution of the Lissajous pattern created above with respect to the distribution of the Lissajous pattern previously set as a normal state. The diagnostic method of the rotating machine system as claimed in claim 7, wherein the measurement of the current data is performed at a frequency different from an integer multiple of the period of the fundamental wave of the rotating machine. A rotating machine system, comprising: one or a plurality of rotating machines; and a power conversion device electrically connected to the rotating machine to control a current flowing through the rotating machine; the power converting device includes a rotating machine; A current measurement unit for measuring current flowing in at least two places, and a control unit that outputs a command value for controlling the rotating machine. The rotating machine system further includes: a diagnosis unit that is electrically or mechanically connected to the rotating unit. And diagnose the state of the machine; and the classification section, which classifies the current data output from the current measurement section and input to the diagnosis section according to each of the above instruction values; and the diagnosis section makes a distribution of the Li Shayu graph, the Li Shayu The distribution of the graph is obtained by superimposing the current data of the rotating machine in at least two phases and a plurality of periods.