JP6481638B2 - Preventive maintenance device for motor drive system - Google Patents

Description

  The present invention relates to a preventive maintenance device for an electric motor drive system.

  There are ex-post maintenance and preventive maintenance. Subsequent maintenance is a method of repairing a product after it has failed. Preventive maintenance is a method of detecting an abnormality before failure and inspecting, exchanging, and maintaining the equipment in advance.

  When a motor drive system (a system including an electric motor and a drive device) breaks down, it is forced to stop operation while it is repaired or replaced. In addition, a great amount of cost is required for restoration, and a great loss occurs. However, in conventional maintenance, a threshold value that is a criterion for failure and a threshold value that is a criterion for issuing an alarm are set for monitored data, and measures are often taken after the data value reaches the threshold value. It was.

  As a precedent example related to maintenance, Patent Document 1 discloses that in a power conversion device, a preset temperature setting value inside the power conversion device, a temperature detected by a temperature detector provided on a cooling fin, and a power conversion device housing It is disclosed that the cause of temperature overheating of the power converter is determined from the detected temperature of the temperature detector provided inside and the detected value of the current detector. According to this document, it is possible to easily determine the cause of abnormal overheating of the power conversion device, and it is possible to determine what to take or replace.

  Further, in Patent Document 2, in an electronic control device, when the device is stopped, the thermal conduction characteristics of the die attach material of the power semiconductor element are measured, so that the degree of progress of fatigue fracture inside the material is reduced by a short pulse power supply. It is disclosed that the remaining life of the device is diagnosed by performing instantaneous heating and then allowing to cool and detecting, observing and measuring in a nondestructive manner.

JP 2014-132929 A International Publication No. 2011/151973

  However, as in the conventional maintenance described above, a threshold value is set in advance in order to determine an abnormality of a device by setting a threshold value that is a criterion for failure and a threshold value that is a criterion for issuing an alarm for monitoring target data. However, it is difficult to set an appropriate threshold value in advance because there are individual differences between devices and ambient conditions.

  Moreover, the technique described in Patent Document 1 is a technique related to post-maintenance that analyzes abnormal overheating in the panel. If the motor drive system is stopped due to a failure, it causes a great loss in the operation of the plant. Therefore, in the maintenance of the electric motor drive system, preventive maintenance that detects a sign of failure or abnormality and performs maintenance is important.

  In addition, since the technique described in Patent Document 2 needs to generate a heat by applying a voltage to the heat-generating thermal device in order to detect the degree of progress of fatigue breakdown of the semiconductor element using the heat-generating thermal device, This is preventive maintenance technology that is performed when the vehicle is stopped. Therefore, it cannot be applied to preventive maintenance during operation of the motor drive system.

  The present invention has been made to solve the above-described problems, and provides a preventive maintenance device for an electric motor drive system capable of performing preventive maintenance during operation without depending on individual differences of devices or ambient conditions. The purpose is to do.

In order to achieve the above object, the present invention is a preventive maintenance device for an electric motor drive system,
An actual measurement data acquisition unit for acquiring actual measurement data in the operating state of the motor drive system;
Among the measured data, the measured data acquisition unit is used during the learning period in the prediction formula that defines the relationship between the monitoring target data that is the target of abnormality monitoring and the monitoring target related data other than the monitoring target data using unknown parameters. A model learning unit that estimates a value of the unknown parameter by applying a plurality of actual measurement data acquired by
After the learning period, by substituting monitoring target related data in the actual measurement data after the learning period acquired by the actual measurement data acquisition unit into the prediction calculation formula, a prediction calculation unit that calculates a predicted value of the monitoring target data;
And an abnormal tendency determination unit that determines an abnormal tendency during operation of the electric motor drive system based on a comparison value between the monitoring target data in the measured data after the learning period and the predicted value.

  According to the present invention, it is possible to estimate the value of an unknown parameter of a prediction calculation formula based on a large number of actually measured data acquired during a learning period. Therefore, it is possible to identify a prediction calculation formula corresponding to individual differences between devices and ambient conditions. Therefore, it is possible to realize a preventive maintenance device that does not depend on individual differences between devices or ambient conditions.

  Further, according to the present invention, after the learning period, the monitoring value calculated by substituting the actual measurement value of the monitoring target data acquired from the motor drive system during operation and the actual measurement value of the acquired monitoring target related data into the prediction calculation formula Based on the comparison value with the predicted value of the target data, the abnormal tendency during driving can be determined. Here, the occurrence of a predetermined deviation between the actually measured value and the predicted value of the monitoring target data does not directly indicate an abnormality of the electric motor drive system, but is an index indicating that the system tends to be abnormal. . Therefore, it is possible to carry out preventive maintenance during operation by monitoring the tendency of system abnormalities.

1 is a conceptual configuration diagram of an electric motor drive system according to Embodiment 1 of the present invention. It is a functional block diagram of the preventive maintenance apparatus 5 which concerns on Embodiment 1 of this invention. It is a flowchart of the processing routine which the preventive maintenance apparatus 5 performs.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code | symbol is attached | subjected to the element which is common in each figure, and the overlapping description is abbreviate | omitted.

Embodiment 1 FIG.
[System Configuration of Embodiment 1]
FIG. 1 is a conceptual configuration diagram of an electric motor drive system according to Embodiment 1 of the present invention. The electric motor drive system shown in FIG. 1 includes a drive device 1 and an electric motor 2.

  The drive device 1 is an inverter device that supplies electric energy to the electric motor 2. A sensor 11 for outputting a signal corresponding to the temperature in the panel is attached to the drive device 1. Preferably, the drive device 1 is also provided with a sensor 12 capable of outputting a signal corresponding to the temperature of the semiconductor element used for controlling the device.

  The drive cooling device 3 cools a semiconductor element used for controlling the drive device 1. A sensor 31 that outputs a signal corresponding to the temperature of the coolant and a sensor 32 that outputs a signal corresponding to the flow rate of the coolant per unit time are attached to the drive cooling device 3.

  The electric motor 2 is electrically connected to the drive device 1. The electric motor 2 converts the electric energy supplied from the drive device 1 into mechanical energy and rotates the rotating shaft. A sensor 21 that outputs a signal according to the temperature of the winding of the motor 2 and a sensor 22 that outputs a signal according to the current flowing through the motor 2 are attached to the motor 2.

  The electric motor cooling device 4 cools the heat of the electric motor 2 generated due to the loss. The motor cooling device 4 includes a sensor 41 that outputs a signal according to the temperature of the coolant, a sensor 42 that outputs a signal according to the flow rate of the coolant per unit time, and an inlet side temperature of a passage for cooling the coolant with air. A sensor 43 that outputs a corresponding signal and a sensor 44 that outputs a signal corresponding to the outlet side temperature of the passage are attached.

  The preventive maintenance device 5 is connected to various sensors 11, 12, 21, 22, 31, 32, 41 to 44 shown in FIG. The preventive maintenance device 5 periodically acquires signals output from these sensors. Hereinafter, functions and operations of the preventive maintenance apparatus 5 will be described with reference to FIGS. 2 and 3.

  FIG. 2 is a functional block diagram of the preventive maintenance device 5 according to Embodiment 1 of the present invention. The preventive maintenance device 5 is a general-purpose computer including a processor and a memory. The function of each unit shown in FIG. 2 is realized by a program stored in a memory being executed by a processor.

  The actual measurement data acquisition unit 51 periodically inputs signals output from the various sensors 11, 12, 21, 22, 31, 32, 41 to 44 shown in FIG. The actual measurement data acquisition unit 51 calculates and acquires actual measurement data (operation data) from various signals periodically input. The actual measurement data includes, for example, ambient temperature, current RMS, temperature / flow rate of the coolant, and the like. Specifically, for the drive device 1, the actual measurement data acquisition unit 51 determines the temperature in the panel from the signal of the sensor 11 (ambient temperature of the drive device 1), the temperature of the semiconductor element from the signal of the sensor 12, and the signal of the sensor 31. The coolant flow rate is calculated from the coolant temperature and the sensor 32 signal. In addition, the actual measurement data acquisition unit 51 determines the winding temperature from the signal from the sensor 21, the effective current value (RMS) from the signal from the sensor 22, the coolant temperature from the signal from the sensor 41, and the signal from the sensor 42. The ambient temperature of the electric motor 2 is calculated from the flow rate of the coolant and the deviation between the sensor 43 and the sensor 44.

  The actual measurement data acquisition unit 51 outputs monitoring target data, which is the target of abnormality monitoring, of the actual measurement data to the model learning unit 54 and the abnormal tendency determination unit 55. In addition, the actual measurement data acquisition unit 51 outputs monitoring target related data other than the monitoring target data among the actual measurement data to the prediction calculation unit 52 as prediction calculation data. For example, the monitoring target data is the temperature of the winding of the electric motor 2. In this case, the monitoring target related data includes the effective value of the current flowing through the motor 2, the ambient temperature of the motor 2, the temperature of the coolant flowing through the motor cooling device 4, and the flow rate of the coolant.

  The prediction calculation unit 52 has a prediction calculation formula that defines the relationship between the monitoring target data and the monitoring target related data using unknown parameters. The value of the unknown parameter is estimated by the model learning unit 54. The prediction calculation unit 52 substitutes the prediction calculation data (monitoring target related data) into a prediction calculation formula to which the estimated parameter value is applied, and calculates a prediction value of the monitoring target data.

  The switching unit 53 switches the operation state of the preventive maintenance device 5 between the learning mode and the preventive maintenance mode. The switching is performed manually by the operator or by a program executed when a predetermined condition is established. The switching unit 53 outputs the prediction calculation data (monitoring target related data) to the model learning unit 54 in the learning mode, and outputs the predicted value of the monitoring target data to the abnormal tendency determination unit 55 in the preventive maintenance mode.

  The model learning unit 54 is executed in the learning mode. Similar to the prediction calculation unit 52, the model learning unit 54 has a prediction calculation formula that defines the relationship between monitoring target data and monitoring target related data using unknown parameters. A plurality of actual measurement data (monitoring target data and monitoring target related data) acquired by the actual measurement data acquisition unit 51 during the learning period is applied to a prediction arithmetic expression to estimate an unknown parameter value. Using a large number of actual measurement data acquired during the learning period, the monitoring target data and the predicted value of the monitoring target data based on the prediction calculation formula are repeatedly compared to identify the prediction calculation formula. The estimated parameter value is output to the prediction calculation unit 52. The learning period is, for example, 1 to 2 months.

  The abnormal tendency determination unit 55 is executed in the preventive maintenance mode. The abnormal tendency determination unit 55 is based on a comparison value between the actual measurement value of the monitoring target data and the predicted value of the monitoring target data calculated by using the prediction calculation formula of the prediction calculation unit 52. Determine abnormal trends. The occurrence of a predetermined deviation between the actual measurement value and the predicted value of the monitoring target data does not directly indicate a system abnormality, but is an index indicating that the system tends to be abnormal.

  The abnormal trend display unit 56 displays the determination result of the abnormal trend in the abnormal trend determination unit 55 on a screen or the like. As described above, preventive maintenance can be performed by monitoring the abnormal tendency of the system during operation.

  Next, the operation of the preventive maintenance device 5 will be described with reference to FIG. FIG. 3 is a flowchart of a processing routine executed by the preventive maintenance device 5.

  In the routine shown in FIG. 3, first, the actual measurement data acquisition unit 51 acquires actual measurement data in the operating state of the electric motor drive system (step S100). The period for acquiring the actual measurement data is about the same as the detection period of the various sensors shown in FIG.

  Next, the preventive maintenance device 5 determines whether the operation state set by the switching unit 53 is the learning mode or the preventive maintenance mode (step S110).

  In the learning mode, the model learning unit 54 uses the unknown parameter to define the relationship between the monitoring target data that is the target of abnormality monitoring in the measured data and the monitoring target related data other than the monitoring target data. By applying the actual measurement data acquired by the actual measurement data acquisition unit 51 during the learning period to the equation, the value of the unknown parameter is estimated (step S120). Until a predetermined learning period (for example, 1-2 months) elapses, the process is executed again from step S100 (step S130). Thereby, a prediction arithmetic expression is identified. The estimated parameter value is applied to the prediction calculation formula of the prediction calculation unit 52.

  When the predetermined learning period has elapsed, the switching unit 53 switches the set mode from the learning mode to the preventive maintenance mode (step S140). Thereafter, the process is executed again from step S100.

  In step S110, new actual measurement data is acquired. If it is determined in step S110 that the preventive maintenance mode is set, preventive maintenance processing in step S150 and subsequent steps is executed.

  In step S150, after the learning period, the prediction calculation unit 52 substitutes the monitoring target related data in the post-learning period actual measurement data acquired by the actual measurement data acquisition unit 51 into the prediction arithmetic expression, and predicts the monitoring target data. Is calculated.

  In step S160, the abnormal tendency determination unit 55 determines the abnormal tendency during the operation of the motor drive system based on the comparison value between the measured value of the monitoring target data in the measured data after the learning period and the predicted value calculated in step S150. Determine.

  In step S170, the abnormal tendency display unit 56 displays the determination result of the abnormal tendency in the abnormal tendency determination unit 55 on a screen or the like.

  As described above, according to the routine shown in FIG. 3, in the learning mode, it is possible to identify the prediction calculation formula based on a large number of actually measured data acquired during the learning period. Therefore, it is possible to realize a preventive maintenance device that does not depend on individual differences between devices or ambient conditions. Further, in the subsequent preventive maintenance mode, among the actual measurement data acquired from the motor drive system during operation, the actual measurement value of the monitoring target data and the actual measurement value of the monitoring target related data were substituted into the identified prediction calculation formula and calculated. Based on the comparison value with the predicted value of the monitoring target data, the abnormal tendency during driving can be determined. Here, the occurrence of a predetermined deviation between the actually measured value and the predicted value of the monitoring target data does not directly indicate an abnormality of the electric motor drive system, but is an index indicating that the system tends to be abnormal. . Therefore, it is possible to implement preventive maintenance during operation by monitoring the tendency of system abnormalities (for example, abnormalities in the cooling system, etc.).

(Modification)
By the way, in the preventive maintenance device 5 of the first embodiment described above, the temperature of the winding of the electric motor 2 has been described as an example of the monitoring target data calculated by the prediction calculation formula, but the prediction calculation formula is limited to this. Is not to be done. For example, the temperature of the semiconductor element that controls the inverter of the drive device 1 may be adopted as the monitoring target data calculated by the prediction calculation formula. In this case, as shown in FIG. 1, a sensor 12 that outputs a signal corresponding to the temperature of the semiconductor element is attached to the drive device 1. The monitoring target related data used in Step S120 and Step S150 are the effective value of the current flowing through the electric motor 2, the ambient temperature of the drive device 1, the temperature of the coolant flowing through the drive cooling device 3, and the flow rate of the coolant. Note that both the temperature of the winding of the electric motor 2 and the temperature of the semiconductor element that controls the inverter of the drive device 1 may be adopted as the monitoring target data.

DESCRIPTION OF SYMBOLS 1 Drive apparatus 2 Electric motor 3 Drive cooling apparatus 4 Electric motor cooling apparatus 5 Preventive maintenance apparatus 11, 12, 21, 22, 31, 32, 41, 42, 43, 44 Sensor 51 Actual measurement data acquisition part 52 Prediction calculation part 53 Switching part 54 Model learning unit 55 Abnormal trend determination unit 56 Abnormal trend display unit

Claims (3)

An actual measurement data acquisition unit for acquiring actual measurement data in the operating state of the electric motor drive system;
Among the measured data, the measured data acquisition unit is used during the learning period in the prediction formula that defines the relationship between the monitoring target data that is the target of abnormality monitoring and the monitoring target related data other than the monitoring target data using unknown parameters. A model learning unit that estimates a value of the unknown parameter by applying a plurality of actual measurement data acquired by
After the learning period, by substituting monitoring target related data in the actual measurement data after the learning period acquired by the actual measurement data acquisition unit into the prediction calculation formula, a prediction calculation unit that calculates a predicted value of the monitoring target data;
An abnormal tendency determination unit that determines an abnormal tendency during operation of the electric motor drive system based on a comparison value between the monitoring target data in the actual measurement data after the learning period and the predicted value;
A preventive maintenance device for an electric motor drive system. The electric motor drive system includes an electric motor and an electric motor cooling device that cools the electric motor,
The monitoring target data includes the temperature of the winding of the electric motor,
The monitoring target related data includes an effective value of a current flowing through the motor, an ambient temperature of the motor, a temperature of a coolant flowing through the motor cooling device, and a flow rate of the coolant.
The preventive maintenance device for an electric motor drive system according to claim 1. The electric motor drive system includes a drive device that supplies electric energy to the electric motor and a drive cooling device that cools the drive device,
The monitoring target data includes a temperature of a semiconductor element that controls the drive device,
The monitoring target related data includes an effective value of a current flowing through the electric motor, an ambient temperature of the drive device, a temperature of a coolant flowing through the drive cooling device, and a flow rate of the coolant,
The preventive maintenance device for an electric motor drive system according to claim 1 or 2.

Images