Method for estimating circuit constants in induction motors

Abstract

To set a circuit constant even when automatic tuning of an induction motor (IM) is stopped.SOLUTION: A rated impedance Z is derived from a rated voltage V and a rated current I of circuit constant design information of a plurality of IMs. A circuit constant (primary resistance (R1), secondary resistance (R2), primary leakage inductance (l1), secondary leakage inductance (l2), mutual inductance (M)) is derived using the unit method by using the rated impedance Z. By using the rated information (rated output, rated current, rated voltage, rated frequency, rated rotation speed, number of poles, and rated impedance Z) as constraints, a trend and a correlation between the circuit constant of each electric induction machine and the data derived from the unit method are analyzed. The constraint is analyzed by comparing the rated information and the constraint of an induction motor to be driven with the analysis result.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a technique for estimating circuit constants required to drive an induction motor (hereinafter referred to as an IM). [Background technology]

[0002] As is well known, inverters use an autotuning method that automatically measures the circuit constants of the IM (such as the constants of the motor windings and the magnitude of the moment of inertia of the load) and stores the measurement results. Examples of this method are known from Patent Documents 1 and 2.

[0003] <Patent Document 1> Patent Document 1 proposes a method for measuring the second-order time constant from the measurement results of IM. This method is carried out by steps (1) to (3).

[0004] (1) Set the excitation current command value to a constant value (i q * = constant), and the torque current command value is set to zero (i q * = 0).

[0005] (2) From the steady state to "i q * " is changed in steps.

[0006] (3) "i q * The step change in the primary side q-axis voltage (V q ) decay waveform is measured and stored, and the second-order time constant is determined from the results.

[0007] <Patent Document 2> In Patent Document 2, the control gain (slip gain) "K S " is used to calculate the IM circuit constant and torque gain "K t A method for deriving " has been proposed. This method is carried out by steps (1) to (4).

[0008] (1) Excitation current (i d ) is constant, and the torque current (i q ) is subjected to an impulse change, the stator d-axis voltage (V d Based on the measurement results of the initial slip gain "K S (0)" is derived.

[0009] (2) d-axis secondary magnetic flux (φ 2d ) and "V d "About "K S (0)" and the error between the derived value and the measured value is "(φ 2d err ) → (0,V d_err )→0” S " is derived.

[0010] (3) "K S After deriving ", (i q ) is changed stepwise, and the rotor angular velocity (ω r ) and "V d ", the leakage inductance (L σ =L1-(M 2 / L2), stator inductance (L1), and rotor inductance (L2) are derived using the least squares method.

[0011] (4) Based on the results of the above derivation, the remaining circuit constants, secondary resistance (R2), mutual inductance (M), and torque gain (K t )" is derived. [Prior art documents] [Patent documents]

[0012] [Patent Document 1] Patent Publication No. 2017-005920 [Patent Document 2] Patent Publication No. 07-298697 Summary of the Invention [Problem to be solved by the invention]

[0013] However, the methods of Patent Documents 1 and 2 may cause the following problems.

[0014] (1) The method of Patent Document 1 measures the secondary time constant via mutual inductance, so it is susceptible to errors in other circuit constants. Furthermore, although it is said to be possible to determine motor faults, it is necessary to reset the measurement and storage time of the decay waveform while assuming variations in the secondary time constant value for each motor, which is cumbersome.

[0015] (2) In the method of Patent Document 2, the optimum slip gain "K2" and circuit constant "L б ,L1,L2" and "V" with multiple circuit constants as variables d When deriving ", it is necessary to calculate multidimensional and multivariable equations using the least squares method, so the values may not converge and calculation results may not be obtained.

[0016] (3) As a result, in the methods of Patent Documents 1 and 2, depending on the IM standard, the accuracy of autotuning may decrease or the autotuning may stop midway. This may cause the autotuning to fail, making it impossible to set the IM circuit constants.

[0017] The present invention has been made to solve such conventional problems, and aims to make it possible to set circuit constants even when IM auto-tuning is stopped midway. [Means for solving the problem]

[0018] (1) The present invention provides a method for estimating circuit constants required to drive an electric induction motor, comprising: an analysis stage based on circuit constant design information of multiple electric induction machines prepared in advance; an estimation stage for estimating the circuit constants using the analysis results of the analysis stage; and The analysis stage includes a first step of deriving a rated impedance Z from a rated voltage V and a rated current I of each of the circuit constant design information; A second step of deriving circuit constants (primary resistance (R1), secondary resistance (R2), primary leakage inductance (l1), secondary leakage inductance (l2), mutual inductance (M)) using the rated impedance Z in a unit method; a third step of analyzing the tendency and correlation between the circuit constants of each electric induction machine and the derived data of the unit method using rated information (rated output, rated current, rated voltage, rated frequency, rated rotation speed, number of poles, and rated impedance Z) as constraints; and The estimation stage is characterized by estimating the circuit constants by comparing the rated information and constraints of the induction motor to be driven with the analysis results.

[0019] (2) In one aspect, the analysis results are tabulated for each constraint condition, The circuit constants are stored in the inverter in advance. If the circuit constants cannot be set by autotuning the inverter, the circuit constants can be set using the information stored in the table.

[0020] The initial value of the auto-tuning of the inverter can also be set using the information stored in the table. If the result of the auto-tuning of the inverter deviates beyond a preset threshold, the value can be corrected based on the information stored in the table. [Effects of the Invention]

[0021] According to the present invention, it is possible to set the circuit constants even when autotuning of the IM is stopped midway. [Brief explanation of the drawings]

[0022] [Figure 1] 10 shows an example of an analysis result in the circuit constant estimation method according to the embodiment of the present invention. [Figure 2] 1 is a flowchart showing a processing procedure according to the first embodiment. [Figure 3] 10 is a flowchart showing a processing procedure according to a second embodiment. [Figure 4]FIG. 11 is an image diagram of an approximation line showing a threshold value in the third embodiment. [Figure 5] 10 is a flowchart showing the processing procedure. DETAILED DESCRIPTION OF THE INVENTION

[0023] The following describes a method for estimating the circuit constants of an IM according to an embodiment of the present invention. This estimation method is used to provide redundancy to the inverter autotuning and ensure that the IM is always ready to operate. Here, it has been found that the relationships between the circuit constant parameters of similar IMs are roughly linear, and this is utilized in the autotuning of the parameters.

[0024] The reasons for failure of current autotuning methods are: (1) When the error in the value obtained by auto-tuning and you are trying to set is obviously large, for example, when the primary resistance (R1) and secondary resistance (R2) are several times the rated impedance, (2) Divergence and oscillation of values due to initial values in convergence calculations, and an increase in the number of iterations. (3) Hardware failure is one of the factors.

[0025] The estimation method proposes solutions to the above factors (1) and (2). That is, for the IM circuit constants, the actual values of the temperature during operation, etc., are more likely to be correct as a result of auto-tuning. Therefore, in cases where auto-tuning has failed but there are no major problems with the hardware, it is possible to set more appropriate values for the circuit constants.

[0026] Specifically, the estimation method is executed by an estimation device (not shown) configured by a computer, which implements an analysis unit and an estimation unit as a result of cooperation between hardware resources (CPU, RAM, ROM, etc.) and software resources (OS, applications, etc.).

[0027] The analysis unit executes an analysis stage to analyze trends and correlations of circuit constants based on circuit constant design information of a plurality of electric induction machines prepared in advance, and the estimation unit executes an estimation stage to estimate the circuit constants using the analysis results of the analysis stage.

[0028] Analysis Stage First, the analysis stage will be described. This analysis stage may use a known machine learning method, but here we will explain an example of analysis based on the results of a large number of measurements of IMs of similar types (such as the same series).

[0029] That is, as mentioned above, the circuit constants of many IMs are analyzed for each similar system to confirm the trends and correlations of the circuit constants. At this time, it is expected that the accuracy of the trends and correlations will be improved by imposing constraints using rating information. The processing contents of the analysis stage (S01 to S04) are explained below.

[0030] S01: First, circuit constant design information (such as a test contact list) for a large number of IMs prepared in advance is stored in a computer storage device.

[0031] S02: Extract the rated voltage V and rated current I from the circuit constant design information stored in S01. Derive the rated impedance Z from the extracted rated voltage V and rated current I.

[0032] S03: Using the rated impedance Z derived in S02, derive the circuit constants (primary resistance (R1), secondary resistance (R2), primary leakage inductance (l1), secondary leakage inductance (l2), mutual inductance (M)) using the unit method (pu method).

[0033] S04: The IM's rated information (rated output, rated current, rated voltage, rated frequency, rated rotation speed, number of poles, rated impedance Z) is acquired and stored in the computer's memory device. This rated information is used as a constraint, and the IM's circuit constants (primary resistance (R1), secondary resistance (R2), primary leakage inductance (l1), secondary leakage inductance (l2), mutual inductance (M)) are calculated. and Its unit system notation (%R1,%R2,%l1,%l2,%M) to Trends and Correlations Is there Analyze.

[0034] Details will be explained based on the analysis results in Figure 1. Here, the values are narrowed down based on the rating information, i.e., constraint conditions, and the relationship between (%R1) and (%R2) is plotted, and a linear approximation line (dashed line) is drawn from the results. This can be expressed as in equation (1).

[0035]

number

[0036] In formula (1), "a" is (%R2 * ) and (%R1 * ) and "b" indicates the slope of the fitted line (%R2 * ) and (%R1 * ) and the intercepts of the fitted line are shown. * " indicates that it is a variable of the approximate line drawn from the plot.

[0037] Then, "a" and "b" derived from equation (1) are tabulated for each constraint, and the table is stored in the inverter beforehand, completing the analysis stage.

[0038] <Estimated Stage> Next, the estimation stage will be explained. Here, the rated information of the IM to be driven is compared with the constraint conditions to search for analysis results with similar trends, and the circuit constants are estimated from the values in the table of search results. Details will be explained below based on Examples 1 to 3.

[0039] (1) Example 1 Example 1 shows the setting of circuit constants when auto-tuning is interrupted due to the factors (1) and (2). If the trend and correlation of the circuit constants can be confirmed from the analysis results of the analysis stage and the rating information of the IM to be driven, auto-tuning can be made redundant. The processing procedure (S11 to S16) of Example 1 will be explained with reference to FIG. 2.

[0040] S11 to S13: When the process starts, auto-tuning of the inverter is performed (S11). After that, it is confirmed whether all the circuit constants can be set by the auto-tuning in S11. If it is confirmed that all the circuit constants can be set, the process proceeds to S13, where the circuit constants are set based on the calculation results of the auto-tuning, and the process ends. On the other hand, if all the circuit constants cannot be set, the process proceeds to S14.

[0041] S14: For circuit constants that have not been set by auto-tuning, it is checked whether the information required to set the circuit constants exists in the table. If the result of the check is that the information exists in the table, proceed to S15, and if not, proceed to S16.

[0042] S15: Circuit constants that have not been set by auto-tuning are set based on the approximation line present in the table, and the process ends.

[0043] S16: An auto-tuning error is displayed and the process ends.

[0044] According to the first embodiment, if there is a trend or correlation in the IM circuit constants, that information is stored in the inverter, so that it is possible to set the circuit constants without performing complex calculations when auto-tuning is performed, even when auto-tuning is stopped midway.

[0045] (2) Example 2 In the second embodiment, a value derived using a table is set as the initial value for auto-tuning, which allows a value closer to the true value to be set compared to when an arbitrary value is set as the initial value.

[0046] This is particularly useful in that convergence can be expected in the auto-tuning method in which calculation results are updated as needed, as in Patent Document 2. The processing procedure (S21 to S28) of the second embodiment will be described with reference to FIG.

[0047] S21: When the process starts, it is checked whether the IM circuit constants can be set from the table stored in the inverter. If the result of the check is that they can be set, the process proceeds to S22, and if they cannot be set, the process proceeds to S23.

[0048] S22: A circuit constant is estimated from the analysis result and the rating information of the IM to be driven by referring to the table, and the estimated result is set as the initial value for auto-tuning.

[0049] S23 to S28: The same processes as S11 to S16 are performed.

[0050] According to the second embodiment, when an auto-tuning method is used in which the calculation results are updated as needed using the least squares method, the initial value can be set to a value close to the true value, improving convergence and making it possible to shorten the time required for auto-tuning.

[0051] (3) Example 3 In the third embodiment, if it is confirmed from the table analysis results that there is a trend / correlation in the circuit constants of the IM to be driven, and if the auto-tuning results deviate from the trend / correlation due to a convergence error or the like, the values of the circuit constants are corrected.

[0052] A: Processing example First, a processing example of this embodiment will be described based on the analysis results in Fig. 1. Here, equation (2) is derived from equation (1) and the auto-tuning result (%R1).

[0053]

number

[0054] At this time, formula (3) is applied to (%R2) calculated by autotuning.

[0055]

number

[0056] However, "α1" and "β1" must be set by the system designer. An example of a method for designing "α1" and "β1" will be explained below.

[0057] Here, the plot point with the largest negative deviation from the approximation line is called the "%R 2β1 ” and the plot point with the maximum positive error from the approximation line is called “%R 2α1 " and "α1" and "β1" are defined as in equations (4) and (5).

[0058]

number

[0059]

number

[0060] The upper and lower limits of equation (3) are found on an approximation line with a slope of "a" that passes through the maximum error points in the positive and negative directions by applying "α1" and "β1" from equations (4) and (5). Figure 4 shows an image of the approximation line that indicates the threshold used to determine whether to correct the numerical values of the circuit constants. If the results of auto-tuning do not satisfy equation (3), equations (6) and (7) are applied to correct the auto-tuning results (%R2).

[0061]

number

[0062]

number

[0063] It should be noted that even if the two circuit constants are not (%R1) and (%R2), they can be similarly applied as long as the relationship between the two circuit constants satisfies linearity.

[0064] B: Processing Procedure The processing content of the third embodiment will be described with reference to FIG.

[0065] S31, S32: Perform the same processing as S11, S12.

[0066] S33: Check whether the circuit constants set in S32 exceed the thresholds ("α1", "β1"). If the check result shows that they do not exceed the thresholds, proceed to S34; if they do exceed the thresholds, proceed to S35.

[0067] S34: The same process as S13 is performed.

[0068] S35: The circuit constants obtained by auto-tuning are corrected based on the threshold value, the corrected results are set as the circuit constants, and the process ends. For example, if the results of auto-tuning do not satisfy equation (3) as described above, equations (6) and (7) are applied to correct the auto-tuning result (%R2).

[0069] S36 to S38: The same processes as S14 to S16 are performed.

[0070] According to the third embodiment, if the auto-tuning results deviate significantly from the analysis results in the table, the values of the circuit constants are reset based on the threshold values, thereby reducing the error and enabling the IM to be operated with high accuracy after auto-tuning is completed.

[0071] The present invention is not limited to the above-described embodiment, and can be modified and implemented within the scope of the claims. For example, the device configuration is not limited to the analysis unit and the estimation unit, and it is sufficient if S01 to S04, S11 to S16, S21 to S28, and S31 to S38 can be executed by a computer.

Claims

1. A method for estimating circuit constants required to drive an electric induction motor, comprising: an analysis stage based on circuit constant design information of multiple electric induction machines prepared in advance; an estimation stage for estimating the circuit constants using the analysis results of the analysis stage; and The analysis stage includes a first step of deriving a rated impedance Z from a rated voltage V and a rated current I of each of the circuit constant design information; A second step of deriving circuit constants (including primary resistance (R1), secondary resistance (R2), primary leakage inductance (l1), secondary leakage inductance (l2), and mutual inductance (M)) using the rated impedance Z in a unit method; a third step of analyzing whether there is a trend or correlation between the circuit constants and the unit system notations (%R1, %R2, %l1, %l2, %M) of the circuit constants of each electric induction machine using rated information (including rated output, rated current, rated voltage, rated frequency, rated rotation speed, and rated impedance Z) as constraints; and The third step is The values are narrowed down based on the constraints, and the relationship between (%R 1 ) and (%R 2 ) is plotted. From the plotted results, a linear approximation line expressed by formula (1) is drawn. [Equation 1] a: Slope of the approximation line between (%R 1 ) and (%R 2 ) b: Intercept of the approximation line between (%R 1 ) and (%R 2 ) *: Variables for the approximate line drawn from the plot A table is created for each constraint condition using "a" and "b" derived from the formula (1) as an analysis result, and the created table is stored in the inverter. The estimation stage compares the rating information and constraints of the induction motor to be driven with the analysis results to search the table for analysis results with similar tendencies, and estimates the circuit constants from the searched values in the table. A method for estimating circuit constants in an induction motor, comprising:

2. If the circuit constants cannot be set by auto-tuning the inverter, The circuit constants are set using the information stored in the table.

2. The method for estimating circuit constants in an induction motor according to claim 1.

3. The initial value of the autotuning of the inverter is set as follows:

2. The method for estimating circuit constants in an induction motor according to claim 1, wherein the circuit constants are set using information stored in said table.

4. If the auto-tuning result of the inverter deviates beyond a preset threshold, 2. The method for estimating circuit constants according to claim 1, wherein the circuit constants are corrected based on the information stored in the table.

Images