WO2021166366A1 - Système de conversion de puissance, dispositif de conversion de puissance, dispositif d'estimation d'état, procédé de conversion de puissance et programme de conversion de puissance - Google Patents

Système de conversion de puissance, dispositif de conversion de puissance, dispositif d'estimation d'état, procédé de conversion de puissance et programme de conversion de puissance Download PDF

Info

Publication number
WO2021166366A1
WO2021166366A1 PCT/JP2020/044906 JP2020044906W WO2021166366A1 WO 2021166366 A1 WO2021166366 A1 WO 2021166366A1 JP 2020044906 W JP2020044906 W JP 2020044906W WO 2021166366 A1 WO2021166366 A1 WO 2021166366A1
Authority
WO
WIPO (PCT)
Prior art keywords
current
power conversion
machine
distribution information
distribution
Prior art date
Application number
PCT/JP2020/044906
Other languages
English (en)
Japanese (ja)
Inventor
善康 高瀬
健太朗 猪又
貞之 佐藤
Original Assignee
株式会社安川電機
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社安川電機 filed Critical 株式会社安川電機
Publication of WO2021166366A1 publication Critical patent/WO2021166366A1/fr

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02P—CONTROL OR REGULATION OF ELECTRIC MOTORS, ELECTRIC GENERATORS OR DYNAMO-ELECTRIC CONVERTERS; CONTROLLING TRANSFORMERS, REACTORS OR CHOKE COILS
    • H02P29/00—Arrangements for regulating or controlling electric motors, appropriate for both AC and DC motors
    • H02P29/02—Providing protection against overload without automatic interruption of supply
    • H02P29/024—Detecting a fault condition, e.g. short circuit, locked rotor, open circuit or loss of load

Definitions

  • One aspect of the disclosure relates to a power conversion system, a power conversion device, a state estimation device, a power conversion method, and a power conversion program.
  • Patent Document 1 the torque command value from the inverter is compared with the torque judgment value, and when the torque command value deviates from the torque judgment value, it is determined that an abnormality has occurred in the drive force transmission mechanism including the drive belt or the drive chain.
  • the control device to be used is described.
  • the power conversion system includes a power conversion device that supplies power to the motor of the machine.
  • the power conversion system produces a current distribution generator that generates current distribution information regarding the current distribution, which is a probability distribution of the frequency components of the current signal, based on the current signal indicating the current operation of the machine, and a past operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal shown.
  • the power converter supplies power to the motor of the machine.
  • the power conversion device generates current distribution information regarding the current distribution, which is a probability distribution of the frequency components of the current signal, based on the current signal indicating the current operation of the machine, and a current distribution generator that generates the past operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal shown.
  • the state estimation device includes a current distribution generator that generates current distribution information regarding a current distribution, which is a probability distribution of frequency components of the current signal, based on a current signal indicating the current operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal indicating the past operation of the machine.
  • the power conversion method is executed by a power conversion system having a power conversion device that supplies power to the motor of the machine.
  • the power conversion method includes a step of generating current distribution information regarding a current distribution, which is a probability distribution of frequency components of the current signal, based on a current signal indicating the current operation of the machine, and a past signal indicating the past operation of the machine. It includes a step of estimating the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of.
  • the power conversion program causes a computer system to function as a power conversion system having a power conversion device that supplies power to a machine motor.
  • the power conversion program has a step of generating current distribution information regarding the current distribution, which is a probability distribution of the frequency component of the current signal, based on the current signal indicating the current operation of the machine, and a past signal indicating the past operation of the machine.
  • the computer system is made to perform a step of estimating the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of.
  • the state of the machine can be estimated appropriately.
  • FIG. 1 is a diagram showing an example of application of the power conversion system 1 according to one aspect of the present disclosure.
  • the electric power conversion system 1 is a control system that supplies electric power to the motor 91 of the machine 9.
  • the power conversion system 1 is connected to one machine 9, but the power conversion system 1 may be connected to a plurality of machines 9.
  • Machine 9 is a device that receives power to perform a predetermined operation according to a purpose and performs useful work.
  • the machine 9 may be a robot or a machine tool.
  • the machine 9 comprises a motor 91, a drive target 92, and a sensor 93.
  • the motor 91 is a device that generates power for driving a drive target 92 that processes a work according to the electric power supplied from the power conversion system 1.
  • the motor 91 may be a rotary motor that rotates the drive target 92, or may be a linear motor that displaces the drive target 92 along a straight line.
  • the motor 91 may be a synchronous motor or an induction motor.
  • the motor 91 may be a permanent magnet type synchronous motor such as an SPM (Surface Permanent Magnet) motor or an IPM (Interior Permanent Magnet) motor.
  • the motor 91 may be a synchronous motor having no permanent magnet, such as a synchronous reluctance motor.
  • the sensor 93 is a device that detects the response of the machine 9 operated by the electric power from the electric power conversion system 1.
  • the response is the output of the machine to a command that is a command to control the machine.
  • the response indicates information about at least one of the operation and state of the machine 9.
  • the response may indicate information about at least one of the movements and states of the driven object 92, for example at least one of the position and speed of the driven object 92.
  • the senor 93 is a rotary encoder that outputs a pulse signal having a frequency proportional to the operating speed of the drive target 92.
  • the rotary encoder can acquire both the position and the speed of the drive target 92.
  • the sensor 93 transmits a response signal indicating a response to the power conversion system 1.
  • the power conversion system 1 further has a function of estimating the current state of the machine 9.
  • the current state of the machine refers to the current state of the machine 9 itself or its components (eg, motor 91, drive target 92, or sensor 93).
  • the current state may be represented by whether the machine 9 is normal or abnormal, or may be represented by the product life of the machine 9.
  • the current state may be represented by the load applied to the machine 9, or may be represented by an abnormal factor of the machine 9.
  • Product life is an index that indicates how long a machine will continue to operate normally.
  • the load is the amount of power consumed by the machine (in other words, the amount of work).
  • Anomalous factors are the causes of abnormalities in the machine.
  • the current state may be represented by at least two combinations selected from normal / abnormal, product life, load, and abnormal factors.
  • the power conversion system 1 includes a power conversion device 10 and a reference device 20 in order to supply power to the motor 91 and estimate the state of the machine 9.
  • the power conversion device 10 is a device that supplies electric power to the motor 91.
  • the power conversion device 10 generates electric power for operating the motor 91 based on an instruction from the host controller (for example, a command signal indicating a command) or an instruction input by the user, and uses the electric power. It is supplied to the motor 91.
  • This supplied electric power corresponds to a driving force command such as a torque command and a current command.
  • the power conversion device 10 may be, for example, a servo amplifier or an inverter.
  • the power conversion device 10 may be incorporated in the machine 9.
  • the power conversion device 10 also has a function of estimating the current state of the machine 9 operated by the electric power. Therefore, the power conversion device 10 also functions as a state estimation device according to the present disclosure.
  • the reference device 20 is a device that transmits information used for estimating the current state of the machine 9 to the power conversion device 10.
  • the above-mentioned host controller may also serve as the reference device 20. That is, in one example, the reference device 20 may be a host controller.
  • FIG. 2 is a diagram showing an example of the hardware configuration of the computer 100 used in the power conversion system 1.
  • the computer 100 includes a main body 110, a monitor 120, and an input device 130.
  • the main body 110 is a device that executes the main functions of the computer.
  • the main body 110 has a circuit 160, which has at least one processor 161, a memory 162, a storage 163, an input / output port 164, and a communication port 165.
  • the storage 163 records a program for configuring each functional module of the main body 110.
  • the storage 163 is a computer-readable recording medium such as a hard disk, a non-volatile semiconductor memory, a magnetic disk, or an optical disk.
  • the memory 162 temporarily stores the program loaded from the storage 163, the calculation result of the processor 161 and the like.
  • the processor 161 constitutes each functional module by executing a program in cooperation with the memory 162.
  • the input / output port 164 inputs / outputs an electric signal to / from the monitor 120 or the input device 130 in response to a command from the processor 161.
  • the input / output port 164 may input / output an electric signal to / from another device.
  • the communication port 165 performs data communication with another device via the communication network N in accordance with a command from the processor 161.
  • the monitor 120 is a device for displaying information output from the main body 110.
  • the monitor 120 is a device capable of displaying graphics, such as a liquid crystal panel.
  • the input device 130 is a device for inputting information to the main body 110.
  • Examples of the input device 130 include operation interfaces such as a keypad, a mouse, and operation buttons.
  • the monitor 120 and the input device 130 may be integrated as a touch panel.
  • the main body 110, the monitor 120, and the input device 130 may be integrated.
  • FIG. 3 is a diagram showing an example of the functional configuration of the power conversion system 1.
  • the power conversion device 10 includes a control mode setting unit 11, a power conversion control unit 12, a current distribution generation unit 13, a reference distribution setting unit 14, and an estimation unit 15.
  • the reference device 20 includes a reference distribution generation unit 21.
  • the control mode setting unit 11 is a functional module for setting the control mode of the power conversion device 10.
  • the control mode refers to a motor control method using a power converter.
  • the number and type of control modes depends on the power converter 10 or the motor 91.
  • the control mode of the inverter includes a V / F control mode and a vector control mode.
  • the power converter 10 has a plurality of control modes corresponding to a plurality of types of machines 9.
  • the control mode setting unit 11 sets at least one control mode from the plurality of control modes based on an instruction from the host controller or the user.
  • the power conversion control unit 12 is a functional module that operates the power conversion device 10 in a predetermined pattern.
  • the pattern refers to the type or type of operation of the power converter.
  • the power conversion control unit 12 operates the power conversion device 10 in a pattern corresponding to at least one control mode set by the control mode setting unit 11.
  • the current distribution generation unit 13 is a functional module that generates current distribution information regarding the current distribution, which is a probability distribution of the frequency component of the current signal, based on the current signal indicating the current operation of the machine 9.
  • the current distribution is represented, for example, by the median (mean) and variance.
  • the current signal is represented by the data in the most recent predetermined time width.
  • the current signal may be a command signal obtained from the host controller, a control signal obtained in the power conversion device 10, or a response signal obtained from the sensor 93.
  • the reference distribution setting unit 14 is a functional module that sets reference distribution information regarding a reference distribution, which is a probability distribution based on the frequency component of a past signal indicating the past operation of the machine 9. Like the current distribution, the reference distribution is represented by, for example, the median (mean) and variance.
  • the past signal is a signal obtained before the current signal. Past signals are also represented by data in a predetermined time width. The past signal may be a command signal obtained from the host controller, a control signal obtained in the power conversion device 10, or a response signal obtained from the sensor 93.
  • the reference distribution setting unit 14 includes a reference distribution storage unit 16, a reference distribution selection unit 17, and an adjustment unit 18.
  • the reference distribution storage unit 16 is a functional module that stores reference distribution information, and stores, for example, a plurality of reference distribution information corresponding to a plurality of control modes.
  • the reference distribution selection unit 17 is a functional module that selects reference distribution information corresponding to at least one control mode set by the control mode setting unit 11 from the reference distribution storage unit 16.
  • the adjustment unit 18 is a functional module that adjusts the selected reference distribution information based on the information of the machine 9.
  • the estimation unit 15 is a functional module that estimates the current state of the machine 9 based on the difference between the current distribution information and the reference distribution information.
  • the reference distribution generation unit 21 is a functional module that generates reference distribution information based on past signals and current signals. In one example, the reference distribution generation unit 21 generates a plurality of reference distribution information corresponding to a plurality of control modes.
  • the reference distribution generation unit 21 may generate individual reference distribution information based on the trained model. This trained model is generated based on past signals. For example, the reference distribution generation unit 21 executes machine learning using a set of signals (which corresponds to past signals) obtained by the trial run of the power conversion device 10 under the set control mode as input data. Generate a trained model corresponding to the control mode. When generating the trained model, the reference distribution generation unit 21 uses the signal (past signal) when the power conversion device 10 and the machine 9 normally operate as input data. The reference distribution generation unit 21 generates a trained model for each of the plurality of control modes. Therefore, the trained model is different for each control mode.
  • Machine learning is a method of autonomously finding a law or rule by iteratively learning based on given information. Machine learning may be supervised learning or unsupervised learning such as reinforcement learning.
  • the trained model is a calculation model that outputs reference distribution information when a signal indicating the operation of the machine 9 is input.
  • the trained model is a computational model that is presumed to be optimal for outputting reference distribution information, and is not necessarily a “realistic optimal computational model”.
  • the trained model may be constructed using a neural network. Considering that the past signal is time series data, the trained model may include a recurrent neural network (RNN). Alternatively, the trained model may include a convolutional neural network (CNN) that processes the frequency components of the signal indicating the operation of the machine.
  • the reference distribution generation unit 21 executes a frequency analysis method such as a fast Fourier transform (FFT) on the signal to convert the signal into a frequency domain.
  • FFT fast Fourier transform
  • the reference distribution generation unit 21 inputs the individual frequency components obtained by the conversion to the CNN.
  • the trained model may have a configuration that does not include a neural network.
  • the reference distribution generation unit 21 includes a model database 22 and a model selection unit 23.
  • the model database 22 is a functional module that stores at least one trained model that generates reference distribution information.
  • the model database 22 stores a plurality of trained models corresponding to a plurality of control modes.
  • the reference distribution generation unit 21 stores the individual trained models generated by the above method in the model database 22.
  • the model selection unit 23 is a functional module that selects a trained model corresponding to at least one control mode set by the control mode setting unit 11 from the model database 22.
  • the reference distribution generation unit 21 may set a threshold value to be used for estimating the current state based on the generation of the trained model.
  • the reference distribution generation unit 21 may set a threshold value based on the standard deviation indicated by the reference distribution obtained by the trained model.
  • the reference distribution generation unit 21 may set a multiple of the standard deviation ⁇ as a threshold value, for example, the threshold value may be set to ⁇ , 2 ⁇ , 3 ⁇ , or 4 ⁇ .
  • the reference distribution generation unit 21 updates the reference distribution information in parallel with the estimation of the current state of the machine 9.
  • the model selection unit 23 selects a trained model corresponding to at least one control mode set by the control mode setting unit 11 from the model database 22, and the reference distribution generation unit 21 is based on the trained model. Generate new reference distribution information.
  • the reference distribution generation unit 21 inputs the current signal obtained by operating the power conversion device 10 in the control mode into the trained model to generate reference distribution information. That is, the reference distribution generation unit 21 may generate the reference distribution information based on the trained model that is generated based on the past signal and outputs the reference distribution information when the current signal is input.
  • the reference distribution generation unit 21 calculates the frequency component of the current signal corresponding to the set control mode by a frequency analysis method such as FFT, inputs the frequency component to the selected trained model, and inputs the reference distribution information. To generate. That is, the reference distribution generation unit 21 generates reference distribution information based on a trained model that is generated based on the frequency component of the past signal and outputs the reference distribution information when the frequency component of the current signal is input. May be good.
  • the reference distribution generation unit 21 may reset (update) the threshold value to be used for estimating the current state based on the update of the reference distribution information.
  • the reference distribution generation unit 21 transmits the generated or updated reference distribution information to the power conversion device 10.
  • the reference distribution setting unit 14 stores the reference distribution information in the reference distribution storage unit 16.
  • the reference distribution information corresponding to the currently set control mode is registered or updated.
  • the reference distribution generation unit 21 also transmits the threshold value to the power conversion device 10, and the reference distribution setting unit 14 stores the threshold value in the reference distribution storage unit 16. do. This process automatically registers or updates the threshold.
  • FIG. 4 is a flowchart showing an example of the operation of the power conversion system 1 as a processing flow S1. That is, the power conversion system 1 executes the processing flow S1.
  • step S11 the control mode setting unit 11 sets the control mode of the power conversion device 10.
  • the control mode setting unit 11 sets at least one control mode from the plurality of control modes of the power conversion device 10 based on an instruction from the host controller or the user.
  • step S12 the power conversion control unit 12 controls the power conversion device 10 based on the set control mode.
  • the power conversion control unit 12 operates the power conversion device 10 in a pattern corresponding to the control mode.
  • the power conversion device 10 supplies electric power to the motor 91 based on the pattern, and the motor 91 drives the drive target 92 according to the electric power.
  • the power conversion system 1 estimates the current state of the machine 9 that operates by the processing.
  • the reference distribution setting unit 14 sets the reference distribution information corresponding to the set control mode.
  • the reference distribution selection unit 17 reads the reference distribution information corresponding to the control mode from the reference distribution storage unit 16.
  • the adjusting unit 18 adjusts the reference distribution information based on the information of the machine 9.
  • the information on the machine 9 may include, for example, at least one of a model number, serial number, operating status, and maintenance history.
  • the operating status refers to information on the operation of the machine 9 or the surrounding environment of the machine 9. Examples of the operating status include the number of years of operation, the average load factor which is the average of the load factors applied to the machine 9 (for example, the motor 91) during operation, and the environmental temperature which is the temperature around the operating machine 9.
  • the maintenance history is a record related to inspection or repair of the machine 9.
  • the adjusting unit 18 may acquire the information of the machine 9 from a predetermined memory in the power conversion device 10 or the machine 9. As the adjustment of the reference distribution information, the adjusting unit 18 may change the median value (mean) of the reference distribution or the width (variance) of the reference distribution.
  • the current distribution generation unit 13 acquires the current signal corresponding to the operation of the power conversion device 10 under the pattern corresponding to the set control mode.
  • the power conversion control unit 12 acquires a signal in the latest predetermined time width from a device related to the power conversion system 1 such as a host controller, a power conversion device 10, and a sensor 93, and obtains a signal in the latest predetermined time width.
  • the signal is output to the current distribution generation unit 13.
  • the current distribution generation unit 13 acquires the signal as a current signal.
  • the acquired current signal differs depending on the control mode.
  • the current signal in V / F control mode can be at least one of a current signal and a voltage signal
  • the current signal in vector control mode is a torque signal and a speed signal. Can be at least one of them. Since the current signal is information used to estimate the current state of the machine, it can be said to be a variable (state variable) for the estimation. Since the current signal corresponds to the control mode, it can be said that the control mode setting is a process corresponding to the switching of state variables.
  • the current distribution generation unit 13 generates current distribution information based on the current signal.
  • the current distribution generation unit 13 executes a frequency analysis method such as FFT on the current signal to convert the current signal into a frequency domain.
  • the current distribution generation unit 13 executes a calculation based on the individual frequency components of the current signal obtained by the conversion, obtains the probability distribution of the frequency components of the current signal as the current distribution, and generates current distribution information regarding the current distribution. do.
  • the current distribution generation unit 13 may execute Bayesian estimation based on the frequency component of the current signal to generate current distribution information.
  • the current distribution generation unit 13 generates current distribution information by a calculation method having a lower calculation cost than the reference distribution generation unit 21. For example, when the reference distribution generation unit 21 generates a reference distribution by a trained model, Bayesian inference is a method having a lower calculation cost than the trained model.
  • the estimation unit 15 estimates the current state of the machine 9 based on the current distribution information and the reference distribution information.
  • the estimation unit 15 uses the reference distribution information when the power conversion device 10 is operated in a predetermined pattern. In one example, the estimation unit 15 uses the reference distribution information when the power conversion device 10 is operated in a pattern corresponding to the set control mode. Alternatively, the estimation unit 15 may use the reference distribution information adjusted by the adjustment unit 18. In any case, the estimation unit 15 calculates the difference between the current distribution information and the reference distribution information. For example, the estimation unit 15 calculates the Mahalanobis distance between the current distribution and the reference distribution as the difference. The Mahalanobis distance may be the distance between the median (mean) of the current distribution and the median (mean) of the reference distribution.
  • the estimation unit 15 estimates the current state of the machine 9 based on the difference. For example, the estimation unit 15 expresses the current state of the machine 9 by using at least one of normal / abnormal, product life, load, and abnormal factor. The estimation unit 15 can output a detailed estimation result such as "the machine is normal but overloaded".
  • the estimation unit 15 may compare the difference with a given threshold value to determine whether the machine 9 is normal or abnormal, and estimate the result of the determination as the current state of the machine 9. .. The estimation unit 15 determines that the machine 9 is normal when the difference is equal to or less than the threshold value, and determines that the machine 9 is abnormal when the difference exceeds the threshold value.
  • the estimation unit 15 may determine the product life or load of the machine 9 based on the degree of deviation of the difference from the threshold value, and estimate the result of the determination as the current state of the machine 9.
  • the degree of divergence is an index indicating how far the difference is from the threshold value, in other words, the distance between the difference and the threshold value.
  • the estimation unit 15 determines that the smaller the degree of deviation, the longer the product life (or the lower the load).
  • the estimation unit 15 may determine that the machine 9 is overloaded. If the difference exceeds the threshold value, it means that the machine 9 is abnormal. Therefore, the estimation unit 15 may determine that the product life is zero in that case.
  • the estimation unit 15 may estimate the abnormal factor corresponding to the difference as the current state of the machine 9 based on the given relationship between the difference and the abnormal factor of the machine 9.
  • the relationship between the difference and the anomalous factor may be expressed by a database.
  • the estimation unit 15 refers to the database and extracts the anomalous factor corresponding to the difference.
  • the database shows the correspondence between the order corresponding to the difference (an index indicating an integral multiple of the reference frequency based on the rotation speed of the motor 91) and the abnormal factor.
  • the relationship between the difference and the anomalous factor may be represented by a trained model.
  • This trained model is a calculation model that outputs anomalous factors when a difference is input, and is generated by machine learning using teacher data showing a large number of combinations of differences and anomalous factors, for example.
  • the estimation unit 15 inputs a difference into the trained model and acquires an abnormal factor.
  • the abnormality factor is set in consideration of various attributes such as the type of the machine 9 and the components of the machine 9. For example, an abnormality of the motor, an abnormality of the driving target, and the like can be mentioned as an example of the abnormality factor.
  • Examples of abnormal factors related to the motor 91 which generally have a high occurrence rate, include deterioration of bearings, deterioration of insulation, burning, magnet peeling, and demagnetization.
  • the threshold value used for estimating normal / abnormal, product life, or load may be set based on the generation of the trained model.
  • the estimation unit 15 reads the threshold value from, for example, the reference distribution storage unit 16 and uses it.
  • the estimation unit 15 outputs the estimation result.
  • the estimation unit 15 may store the estimation result in a recording medium such as storage 163.
  • the estimation unit 15 may display the estimation result on the monitor 120 in a format such as text, a moving image by computer graphics (CG), or a still image.
  • CG computer graphics
  • the power conversion system 1 may repeatedly execute the processes after step S14.
  • the power conversion system 1 may repeatedly execute the processes after step S14 at a given interval while the machine 9 continues to operate.
  • the current distribution generation unit 13 acquires a new current signal in step S14, and generates new current distribution information based on the current signal in step S15.
  • the estimation unit 15 estimates the current state of the machine 9 based on the current distribution information and the reference distribution information, and outputs the estimation result in step S17.
  • the reference distribution generation unit 21 may update the reference distribution information in parallel with the series of processes in steps S14 to S17. For example, the reference distribution generation unit 21 updates the reference distribution information using the current signal acquired in step S14, and also updates the threshold value if necessary. The updated reference distribution information (and updated thresholds, if any) will be used in subsequent estimates based on the new current signal. The reference distribution generation unit 21 updates the reference distribution information (and the threshold value) using the current signal only when the machine 9 is determined to be normal in the estimation based on the current signal acquired in step S14. You may.
  • Each functional module of the power conversion system 1 is realized by loading a power conversion program on the processor 161 or the memory 162 and causing the processor 161 to execute the program.
  • the power conversion program includes a code for realizing each functional module of the power conversion system 1.
  • the processor 161 operates the input / output port 164 or the communication port 165 according to the power conversion program, and executes reading and writing of data in the memory 162 or the storage 163. By such processing, each functional module of the power conversion system 1 is realized.
  • the power conversion program may be provided after being fixedly recorded on a non-temporary recording medium such as a CD-ROM, a DVD-ROM, or a semiconductor memory.
  • the power conversion program may be provided via a communication network as a data signal superimposed on a carrier wave.
  • the power conversion system includes a power conversion device that supplies power to the motor of the machine.
  • the power conversion system produces a current distribution generator that generates current distribution information regarding the current distribution, which is a probability distribution of the frequency components of the current signal, based on the current signal indicating the current operation of the machine, and a past operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal shown.
  • the power converter supplies power to the motor of the machine.
  • the power conversion device generates current distribution information regarding the current distribution, which is a probability distribution of the frequency components of the current signal, based on the current signal indicating the current operation of the machine, and a current distribution generator that generates the past operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal shown.
  • the state estimation device includes a current distribution generator that generates current distribution information regarding a current distribution, which is a probability distribution of frequency components of the current signal, based on a current signal indicating the current operation of the machine. It is provided with an estimation unit that estimates the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of the past signal indicating the past operation of the machine.
  • the power conversion method is executed by a power conversion system having a power conversion device that supplies power to the motor of the machine.
  • the power conversion method includes a step of generating current distribution information regarding a current distribution, which is a probability distribution of frequency components of the current signal, based on a current signal indicating the current operation of the machine, and a past signal indicating the past operation of the machine. It includes a step of estimating the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of.
  • the power conversion program causes a computer system to function as a power conversion system having a power conversion device that supplies power to a machine motor.
  • the power conversion program has a step of generating current distribution information regarding the current distribution, which is a probability distribution of the frequency component of the current signal, based on the current signal indicating the current operation of the machine, and a past signal indicating the past operation of the machine.
  • the computer system is made to perform a step of estimating the current state of the machine based on the difference between the reference distribution information and the current distribution information regarding the reference distribution, which is a probability distribution based on the frequency component of.
  • the current state of the machine is estimated based on the difference between the probability distribution indicating the current operation of the machine and the probability distribution based on the past operation of the machine.
  • the probability distribution By using the probability distribution, the current state and the past state are compared with each other, including the certainty of estimation regarding the state of the machine, so that the current state of the machine can be estimated appropriately.
  • the current state of a machine can be estimated without being affected by the characteristics of each machine. This estimation can improve the efficiency of work such as repairing machines and investigating the causes of machine abnormalities.
  • the power conversion system may further include a power conversion control unit that operates the power conversion device in a predetermined pattern.
  • the current distribution generation unit may generate current distribution information based on the current signal when the power conversion device is operated in a predetermined pattern, and the estimation unit operates the power conversion device in a predetermined pattern.
  • the current state of the machine may be estimated based on the difference between the reference distribution information and the current distribution information. By comparing two probability distributions corresponding to the same pattern, the accuracy of estimation can be improved.
  • the power conversion system corresponds to a control mode setting unit that sets at least one control mode from a plurality of control modes of the power conversion device corresponding to a plurality of types of machines, and at least one control mode.
  • a reference distribution selection unit for selecting reference distribution information may be further provided.
  • the estimator may estimate the current state of the machine based on the difference between the selected reference distribution information and the current distribution information. Since the reference distribution is prepared so as to correspond to each of the plurality of control modes of the power converter, the current state of the machine can be appropriately estimated according to the actual control mode of the power converter.
  • the power conversion system may further include a storage unit that stores reference distribution information and an adjustment unit that adjusts reference distribution information based on machine information.
  • the estimator may estimate the current state of the machine based on the difference between the adjusted reference distribution information and the current distribution information.
  • the power conversion device may further include a storage unit that stores reference distribution information and an adjustment unit that adjusts reference distribution information based on machine information.
  • the estimator may estimate the current state of the machine based on the difference between the adjusted reference distribution information and the current distribution information.
  • a reference distribution that takes into account the characteristics of the machine is prepared, so the accuracy of estimation can be improved.
  • the power conversion device may include a current distribution generation unit and an estimation unit.
  • the power conversion system may further include a reference device that transmits reference distribution information to the power conversion device. Since the power converter performs the generation of current distribution information and the estimation of the current state of the machine, there is no need to introduce separate hardware resources for that estimation. Therefore, the power conversion system can be constructed inexpensively and easily.
  • the power conversion system may further include a reference distribution generation unit that generates reference distribution information based on past signals and current signals.
  • the reference distribution to be compared with the current distribution is generated considering not only the past signal but also the current signal. Therefore, the difference between the two probability distributions can be appropriately obtained, and the current state of the machine can be continuously estimated with high accuracy.
  • the reference distribution generator generates the reference distribution information based on the trained model which is generated based on the past signal and outputs the reference distribution information when the current signal is input. You may. Since the accuracy of the reference distribution is improved by using the trained model, the current state of the machine can be estimated accurately.
  • the reference distribution generator is generated based on the frequency component of the past signal, and is based on a trained model that outputs the reference distribution information when the frequency component of the current signal is input.
  • Reference distribution information may be generated. By considering the frequency component of the current signal, reference distribution information that reflects the tendency of the change of the current signal is generated. The accuracy of estimation can be improved by using this reference distribution information.
  • the power conversion system corresponds to a control mode setting unit that sets at least one control mode from a plurality of control modes of the power conversion device corresponding to a plurality of types of machines, and at least one control mode. It may further include a model selection unit that selects a trained model.
  • the reference distribution generator may generate reference distribution information based on the selected trained model. Since the reference distribution information is generated using the trained model corresponding to the control mode of the power converter, the current state of the machine can be estimated accurately according to the actual control mode of the power converter.
  • the estimator determines whether the machine is normal or abnormal based on the comparison of the difference with the threshold set based on the generation of the trained model.
  • the result of the determination may be estimated as the current state of the machine. It can be said that the threshold value set based on the generation of the trained model is a reference value set objectively without depending on the human empirical rule. By using this threshold value, it is possible to objectively determine whether or not the machine is normal.
  • the estimation unit determines the product life of the machine based on the degree of deviation of the difference from the threshold value set based on the generation of the trained model, and the result of the determination is the machine. It may be estimated as the current state of. It can be said that the threshold value set based on the generation of the trained model is a reference value set objectively without depending on the human empirical rule. By using this threshold value, the product life of the machine can be objectively determined.
  • the current distribution generation unit may generate the current distribution information by a calculation method having a lower calculation cost than the reference distribution generation unit.
  • the estimation unit may estimate the abnormal factor corresponding to the difference as the current state of the machine based on the given relationship between the difference and the abnormal factor of the machine. With this configuration, the cause of machine abnormality can be estimated in detail.
  • the current distribution generator may execute Bayesian estimation based on the frequency component of the current signal to generate the current distribution information.
  • Bayesian inference By using Bayesian inference, the current distribution can be obtained accurately even when the sample data is small.
  • the probability distribution can be calculated on limited hardware resources.
  • the accuracy of the calculation can be improved as the processing is repeated.
  • the state estimation device may be applied to a device different from the power conversion device.
  • the state estimation device may be applied to a configuration in which the state to be estimated appears at the frequency of the signal.
  • the power conversion device described in the above embodiment is a device directly related to the current state of the machine from the viewpoint of supplying electric power to the motor of the machine. Therefore, by applying the state estimation device to the power conversion device, it is possible to accurately estimate the current state of the machine while saving the hardware resources in the power conversion system as a whole.
  • the configuration of the power conversion system is not limited to the above embodiment.
  • the reference device may reside in a computer system or control system separate from the power conversion system.
  • the power conversion system and the other system are connected via a communication network, and the same functions and processes as those in the above embodiment are executed.
  • the function of the reference device may be incorporated into the power conversion device or the state estimation device.
  • the power conversion device 10 includes an adjusting unit 18, but this adjusting unit may be omitted. When there is only one control mode of the power converter, it is not necessary to separate the processing for each control mode.
  • the hardware configuration of the system is not limited to the mode in which each functional module is realized by executing the program.
  • the functional module in the above embodiment may be configured by a logic circuit specialized for the function, or may be configured by an ASIC (Application Specific Integrated Circuit) in which the logic circuit is integrated. ..
  • the processing procedure of the method executed by at least one processor is not limited to the example in the above embodiment. For example, some of the steps (processes) described above may be omitted, or each step may be executed in a different order. In addition, any two or more steps of the above-mentioned steps may be combined, or a part of the steps may be modified or deleted. Alternatively, other steps may be performed in addition to each of the above steps.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Inverter Devices (AREA)

Abstract

Le système de conversion de puissance selon un exemple de la présente invention comprend un dispositif de conversion de puissance pour alimenter un moteur d'une machine. Le système de conversion de puissance comprend : une unité de génération de distribution de courant pour générer, sur la base d'un signal de courant indiquant le fonctionnement actuel d'une machine, des informations de distribution de courant concernant une distribution de courant qui est la distribution de probabilité d'une composante de fréquence du signal de courant ; et une unité d'estimation pour estimer l'état actuel de la machine sur la base de la différence entre les informations de distribution de référence et les informations de distribution de courant, lesdites informations de distribution de référence concernant une distribution de référence qui est une distribution de probabilité sur la base d'une composante de fréquence d'un signal antérieur indiquant le fonctionnement antérieur de la machine.
PCT/JP2020/044906 2020-02-17 2020-12-02 Système de conversion de puissance, dispositif de conversion de puissance, dispositif d'estimation d'état, procédé de conversion de puissance et programme de conversion de puissance WO2021166366A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202062977414P 2020-02-17 2020-02-17
US62/977,414 2020-02-17

Publications (1)

Publication Number Publication Date
WO2021166366A1 true WO2021166366A1 (fr) 2021-08-26

Family

ID=77391730

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/044906 WO2021166366A1 (fr) 2020-02-17 2020-12-02 Système de conversion de puissance, dispositif de conversion de puissance, dispositif d'estimation d'état, procédé de conversion de puissance et programme de conversion de puissance

Country Status (1)

Country Link
WO (1) WO2021166366A1 (fr)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015079554A1 (fr) * 2013-11-29 2015-06-04 株式会社日立製作所 Dispositif permettant d'estimer l'état d'un réseau électrique, son procédé d'estimation d'état et système de commande de réseau électrique
JP2019028765A (ja) * 2017-07-31 2019-02-21 株式会社安川電機 電力変換装置、サーバ、及びデータ生成方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015079554A1 (fr) * 2013-11-29 2015-06-04 株式会社日立製作所 Dispositif permettant d'estimer l'état d'un réseau électrique, son procédé d'estimation d'état et système de commande de réseau électrique
JP2019028765A (ja) * 2017-07-31 2019-02-21 株式会社安川電機 電力変換装置、サーバ、及びデータ生成方法

Similar Documents

Publication Publication Date Title
JP6774637B2 (ja) 制御装置及び制御方法
JP6193961B2 (ja) 機械の送り軸の送りの滑らかさを最適化する機械学習装置および方法ならびに該機械学習装置を備えたモータ制御装置
JP6243385B2 (ja) モータ電流制御における補正値を学習する機械学習装置および方法ならびに該機械学習装置を備えた補正値計算装置およびモータ駆動装置
US10331104B2 (en) Machine tool, simulation apparatus, and machine learning device
JP6469065B2 (ja) 機械学習装置及び加工時間予測装置
US11822308B2 (en) Polishing tool wear amount prediction device, machine learning device, and system
JP2018186610A (ja) システムおよび評価装置ならびに評価方法
JP2019139755A (ja) 研磨工具摩耗量予測装置、機械学習装置及びシステム
CN108628188A (zh) 控制系统、控制器以及控制方法
US11960251B2 (en) Parameterization of a component in an automation system
CN110286612B (zh) 控制装置
US20190317472A1 (en) Controller and control method
US20210049506A1 (en) Learning assistance device, processing system, learning assistance method, and storage medium
DK2339416T3 (en) System and method of operating a machine
JP2019185742A (ja) 制御装置及び制御方法
JP2019150932A (ja) 衝突位置推定装置及び機械学習装置
Shaker et al. Design and implementation of fuzzy logic system for DC motor speed control
Schenke et al. Finite-set direct torque control via edge computing-assisted safe reinforcement learning for a permanent magnet synchronous motor
WO2021166366A1 (fr) Système de conversion de puissance, dispositif de conversion de puissance, dispositif d'estimation d'état, procédé de conversion de puissance et programme de conversion de puissance
KR20210013971A (ko) Pso를 이용한 모터 제어기의 자동 학습 튜닝 시스템
JP2020035159A (ja) パラメータ調整装置
CN116569120A (zh) 信息处理装置及信息处理方法
US10684594B2 (en) Machine learning device, servo motor controller, servo motor control system, and machine learning method
CN110888398A (zh) 控制装置、cnc装置以及控制方法
US20220072708A1 (en) Motion model calculation device, control device, joint mechanism, and motion model calculation method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20919829

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

NENP Non-entry into the national phase

Ref country code: JP

122 Ep: pct application non-entry in european phase

Ref document number: 20919829

Country of ref document: EP

Kind code of ref document: A1