KR101009316B1 - An error detection method based on network analysis of induction motor winding - Google Patents

Abstract

The present invention relates to a method for detecting a failure of an induction motor through circuit analysis of an induction motor winding. In the present invention, to detect the failure of the induction motor, extracting the power harmonics and non-power harmonics (S10), calculating the even harmonic current harmonics (S20), calculating the odd harmonic current harmonics (S30), Calculating a positive sequence harmonic current harmonic (S40), calculating a subsequence harmonic current harmonic (S50), and calculating a zero sequence harmonic current harmonic (S60).

In the present invention, the user can set the upper and lower control limits by comparing data such as the effective value of the coil, transformer, and AC motor, harmonic content rate, and gain value.

Induction motor, fault, short circuit, fault current, fault detection, fault diagnosis, winding.

Description

An error detection method based on network analysis of induction motor winding}

The present invention relates to a method for detecting a failure of an induction motor by analyzing a circuit of the induction motor winding. In particular, the present invention relates to a method of detecting a failure of an induction motor using a resistance, an impedance, an induction coefficient, and a harmonic harmonic, and a method of detecting a failure of an induction motor using a deviation of a rotor current waveform. The winding circuit analysis technology of the induction motor provides an excellent fault detection tool, and the comparison of the record between the coil, transformer, AC motor and DC motor sets the upper and lower control limits.

Representative methods for detecting a failure of a conventional induction motor are as follows:

1. Fault detection method by trend of impedance change

It is true that in equipment with alternating current 600 volts or more, short circuits develop and cause rapid errors, and in particular, short circuits gradually develop into failures. In practice, the detection of these errors depends on the peak value of the frequency spectrum and frequency section. As soon as the ball bearings begin to break away, the bearings detach very quickly and are too fast to detect using vibration analysis. If the vibration analysis cannot detect the functional deterioration of the bearing over time, this view is correct.

Conventionally, reference values, such as gain and harmonic content of a specific frequency section, are compared with each reference value for each type and capacity of a motor current analysis (MCA) motor to compare actual values such as gain and harmonic content of collected data with each other. It can be used to set upper and lower control limits for manufacturing and acceptance test.

As such, the MCA used for trends of impedance and for analysis purposes is a comparison tool that uses the difference between the percent imbalance method and the experimental method. In this method of percent imbalance, the difference between the coils (i.e., the phase difference in a three-phase motor) takes advantage of the fact that over time it is oriented in either direction. For example, the resistance value is affected by temperature, while the relative difference between phases is not affected by temperature, and using this fact, the user or software does not need to perform temperature correction calculations, resulting in impedance It can be seen that (Z) and induction coefficient (L) are not significantly affected by temperature. Thus, the percentage imbalance method has been recognized as the most convenient way to detect errors over time.

On the other hand, in the percent imbalance method, the extremes are more visual than numbers. Sudden changes in the graph indicate that an error is occurring and should be addressed immediately. In this method, a slight change over time indicates that the error is inclined in some direction, which should be taken into account regularly, and a change in the imbalance of the resistance component R usually indicates that the connection is loose.

Also, when induction coefficients and impedance imbalances occur due to rotor position, the relative imbalances show similar values (ie, L = 11%, Z = 12% vs. L = 5%, Z = 50%). If there is a large difference in relative unbalance between L and Z, this indicates an electrical breakdown over time and action should be taken.

On the other hand, when detecting the failure of the induction motor, the trend of failure over time can be examined by using the graphic technique. Sudden changes in readings in the records indicate that severe errors are occurring and that emergency measures should be taken. For example, as scheduled, records requiring repairs will gradually change from green to yellow and then red. This would indicate a change of record as follows. Table 1 shows the fault detection criteria by the trend of impedance and phase angle change.

record Change of reference value Accuracy R.Z.L. <3% green R.Z.L. <3%, <5% yellow R.Z.L. > 5% Red

2. Fault detection method by rotor current deviation

The failure detection method by the rotor current deviation is a method suitable for the trend for performing the initial evaluation of the rotor condition. This method is based on the difference in relative average currents between each sinusoidal waveform. That is, the deviation value of the average current is calculated from the average value, and the higher the current deviation, the higher the deviation value. Table 2 shows the criteria for rotor current deviation.

Deviation(%) evaluation <5 Good > 5, <15 Bad > 15 Severe

These deviation values are suitable to represent the trend for the error condition of the rotor. In other words, by setting the reference line of the average rated current of the rotor, and periodically comparing the deviation value of the rotor current with respect to the reference line of the operating time can be given a change trend value for the rotor error. A change of 5% or more in the rotor current deviation indicates that an error of the rotor is occurring, and the detected error means a rotor eccentricity, a broken rotor rod, or breakage of the rotor.

3. Fault detection method by harmonic current harmonic

In this method, since the equations for detecting motor defects are all calculated from the rotational frequency, the number of revolutions that most affects the reliability of the diagnostic algorithm is the rotational speed. In this method, if the exact number of revolutions is not known, the error of the rotor rod defect frequency is larger, as well as the static and dynamic eccentric frequencies which are analyzed as high frequencies.

4.Stator winding short circuit detection by spectrum analysis

The stator windings of induction motors are stressed by various factors, such as thermal overload mechanical vibrations acting on the insulation system. They affect insulation in different ways, but they interact with each other in such a way that the degradation caused by one cause increases the degradation caused by the rest.

Line short is one of the most common types of stator failures, especially motors with irregular winding styles. In most cases, this type of fault develops between layers, phases, or coils and ground faults, eventually leading to breakage of the motor.

Degradation of the winding insulation usually starts with line faults that are related to some number of windings. Fault currents cause severe local heating and rapidly propagate to much of the windings. Insulation failures between the windings and the ground can cause large ground currents that cause irreparable damage to the iron core of the motor.

Table 3 and Figure 1 summarize the common failure modes and patterns. Regardless of the cause of the failure, the actual failure mode can be divided into five groups. The most common type of stator winding short circuit failure is a line short within the coil.

Failure Mode Type of failure Short circuit between coils Symmetry Interlayer shorting on the same bed Open phase: single phase Interphase Asymmetrical form with ground Phase-to-ground short Other asymmetrical forms without grounding Opening in the Hansang Other asymmetrical forms without grounding

In general, the fault progression process involves turn-to-turn → coil-to-coil → phase-to phase or coil and phase-to-ground → final failure. Proceeds in the order of. For example, if the motor is frequently started, minute line short circuits occur in one coil due to excessive coil movement. As this state progresses, excessive heat is generated in the shorted coil, which in turn causes insulation deterioration. Depending on the type of protective device of the motor, the motor may continue to operate, but the increase in heat occurs at the damaged site until the phase or ground insulation is broken. This causes phase to phase defects and ground faults.

6 compares the frequency spectra of the stator current signals of a good motor without failure under full load conditions and a motor with a fault in the line. It can be seen that even in a good motor, the third harmonic component of the power supply frequency (60 Hz) appears. In the full load condition in FIG. 6A, frequency components appear in 30 Hz and around 90 Hz in a good motor, which are side-band frequencies with a width of rotational frequency. When the line defect occurs in FIG. 6B, these components are slightly changed, and it can be seen that the 180 Hz component, which is only the third harmonic harmonic component, is greatly increased.

In summary, in the prior art, the rotation speed detection method using current measurement finds the rotation frequency in the spectrum by analyzing only the current signal, not directly measuring the shaft rotation with a taco sensor. If there is no component, it may not be able to find the number of revolutions, and it is accompanied by a large error depending on the frequency resolution.

Therefore, there is a need for the appearance of a failure detection method of an induction motor that can improve these problems.

An object of the present invention is to provide a fault detection method by analyzing a circuit of an induction motor winding to solve the problems of the fault detection methods of the conventional induction motor described above.

Other objects and advantages of the present invention will become more apparent from the following detailed description of the invention and the accompanying drawings.

The fault detection method by circuit analysis of the induction motor winding according to the present invention,

To detect the failure of an induction motor,

Extracting power harmonics and non-power harmonics (S10);

Calculating an even harmonic current harmonic (S20);

Calculating odd harmonic current harmonics (S30);

Calculating a positive sequence harmonic current harmonic (S40);

Calculating a sub-sequence harmonic current harmonic (S50); And

Calculating a zero sequence harmonic current harmonic (S60),

The above steps (S10 ~ S60) is characterized in that it is performed sequentially.

In this invention, the step of extracting the power source harmonics and non-power harmonics (S10),

Reading frequency CP1 and level CL1 values of peak components in the spectrum (S10-1);

Confirming that an integer multiple of 60 Hz (within ± 0.2%) of the frequencies of the peak components read above (S10-2);

The part to be an integer is harmonic PH1 of the integer component and the peak level PL1 of the integer component, and if not, converting to NPH1 and NPL1 (S10-3);

Calculating TPH by adding the square of PL1, which is the level of the integer component PH1, to continue adding T1;

Calculating TNPH by adding a square of NPL1, which is a non-integer level of the component (S10-5);

Repeating the above steps (S10-1 to S10-5) to a predetermined number a1 (S10-6);

Completion of the above predetermined number of times, calculating the total harmonic distortion (%) by dividing the TPH and TNPH values by a power of 1/2 (root) by the level (60) at 60 Hz (S10-7). ); And

And storing (S10-8) displaying and displaying "" danger "," "warning" "by storing the value at a predetermined time interval and comparing the level with the reference value.

At this time, the frequency analysis is characterized in that it is processed in high resolution.

In this invention, the step of calculating the even harmonic current harmonic (S20),

Reading the frequency CP1 and the level CL1 of the component having the peak in the spectrum (S20-1)

Setting H = 60x2, and if CP1 is H = 120, square the sum value CL1, add a2 = a2 + 1, and set x1 = x1 + 2 to repeat S20-1 (S20-2);

If H = 120, compare CP1 to less than H = 120, repeat CP2, and if CP1 is greater than 120, change to H = 60x4 and repeat (S20-3);

After repeating the above steps (S20-1 to S20-3) to a predetermined number a2, the squared sum (Het) is added to the power of 1/2 (root) and divided by the component (60) of 60 Hz (total even harmonic distortion) (Total Obtaining Even Harmonic distortion (%) (S20-4); And

And a step (S20-5) of storing the value at each predetermined time interval (The) and comparing the level with the reference value and displaying "danger" and "warning".

In the present invention, the step (S30) of calculating the odd harmonic current harmonics,

Reading the frequency CP1 and the level CL1 of the component having the peak in the spectrum (S30-1);

If H = 60x1 (x = 1), CP1 is H = 60, the value CL1 is squared and added, and a3 = a3 + 1, x2 = x2 + 2, and the above steps (S30-1) are repeated. Step S30-2;

If H = 60, compare CP1 to less than H = 60. Read CP2 and repeat step S30-2. If CP1 is greater than 60, change to H = 60x3 and go to step S30-1. Return step (S30-3);

After repeating up to a3 a3 times, multiplying the added value (Het) by 1/2 power (root) to divide by 60Hz component (L60) to obtain the total odd harmonic distortion (%) (S30). -4); And

And storing the value at each predetermined time interval (Tho) and comparing the level with the reference value (“30”) to display “danger” and “warning”.

In this invention, the step (S40) of calculating the above positive sequence harmonic current harmonic,

Obtaining both sequence harmonic harmonics (4,7,10 ..) components from the stored peak list (S40-1);

Checking and storing whether the harmonic component is within an error range (S40-2); And

Obtaining and storing the total amount of sequence harmonic components, and comparing the set risk and warning level to determine the failure (S40-3),

At this time, in the above step (S40-3), if the rotational force acts in the direction of increasing the RPM of the motor, heat is generated in the stator, or if the danger or warning level is exceeded, it is determined as a failure,

The rotation speed of the induction motor is calculated by using an interpolation method of a window function used for frequency analysis.

In the present invention, the step (S50) of calculating the sub-sequence harmonic current harmonic,

Obtaining a subsequence harmonic harmonic (3,5,8 ..) component from the stored peak list (S50-1);

Checking and storing whether the harmonic harmonic component is within an error range (S50-2); And

And calculating (S50-3) to obtain and store the total bush sequence harmonic harmonic component and determine the failure by comparing with the set risk and warning level.

At this time, in the above step (S50-3), if the total bushing harmonic harmonic component obtained exceeds the danger or warning level, the torque acts in the reverse direction to the motor, the current increases, or the heat is generated,

Frequency analysis is processed in high resolution,

The rotation speed of the induction motor is calculated by using an interpolation method of a window function used for frequency analysis.

In the present invention, the step (S60) of calculating the zero sequence harmonic current harmonic,

Obtaining a zero sequence harmonic harmonic (3, 6, 9.) component from the stored peak list (S60-1);

Checking and storing whether the harmonic harmonic component is within an error range (S60-2); And

Obtaining and storing the total zero sequence harmonic harmonic component and comparing the set risk and warning levels to determine a failure (S60-3).

At this time, if the danger or warning level is exceeded in the above step (S60-3), or if the heat is generated from the motor but there is no influence on the rotation, or the overcurrent and the circulating current flow through the neutral wire due to the nonlinear load, it is determined as a failure. and,

Frequency analysis is processed in high resolution,

The rotation speed of the induction motor is calculated by using an interpolation method of a window function used for frequency analysis.

The effects expected by this invention are as follows:

First, an excellent fault detection tool can be provided by analyzing the winding circuit of the induction motor.

Second, the user can set the control limits of the upper and lower limits by comparing data such as the effective value of the coil, transformer, AC motor, harmonic flux rate, and gain value.

The conventional method of detecting rotation speed by measuring current does not directly measure shaft rotation with taco sensor but finds the rotation frequency in the spectrum that analyzes the current signal only.If there is no rotation frequency component in the spectrum, the rotation speed cannot be found. Also, accompanied by a large error according to the frequency resolution. Therefore, in order to solve this problem, in the present invention, the frequency analysis is made high resolution, and the accurate rotational speed is calculated by applying the window function interpolation method used in the frequency analysis to increase the reliability of defect detection.

On the premise that the rotational speed of the electric motor calculated by this invention is exactly the same as the rotational speed measured directly, fault detection criteria were set based on empirical data. The fault detection method by circuit analysis of the induction motor winding according to the present invention consists of the following six algorithms as shown in FIG.

end. Power Harmonics and Non-Power Harmonic Extraction Algorithms (S10)

3 is a detailed flowchart of an extraction algorithm S10 of power harmonics and non-power harmonics in the fault detection method by circuit analysis of an induction motor winding according to the present invention. The processing of this algorithm is as follows:

1) Read CP1, CL1 (a = 1): Reads the frequency CP1 and level CL1 of the peak component from the spectrum (S10-1).

2) CP1 / 60: Check that it is an integer multiple of 60 Hz (within ± 0.2%) of the read peak frequencies (S10-2).

3) The portion to be an integer is the harmonic PH1 of the constant component and the peak level PL1 of the constant component. If the constant is not, the portion is stored as NPH1 and NPL1 (S10-3).

4) TPH: PL1 which is the level of the integer component PH1 is squared and added continuously (S10-4).

5) TNPH: Squares and adds NPL1, which is a non-integer component level (S10-5).

6) Repeat this process with a = a + 1 and run the loop until a = 100 (S10-6).

7) If a = 101, the TPH and TNPH values are divided by half (root), divided by the level at 60 Hz (L60), and multiplied by 100 to obtain the total harmonic distortion (%) (S10-7). ).

8) The value is stored at each predetermined time interval (Tipd) and the level is compared with the reference value to display and display "DANGER" and "Warning" (S10-8).

I. Even harmonic current harmonics algorithm (S20)

4 is a detailed flowchart of an even harmonic current harmonic calculation algorithm S20 in the fault detection method by circuit analysis of an induction motor winding according to the present invention. The processing of this algorithm is as follows.

1) Read CPa, CLa (a = 1): Read the frequency CP1 and level CL1 of the component having the peak in the spectrum (S20-1).

2) Let H = 60x2. If CP1 is H = 120, the value CL1 is squared and added, and 1) is repeated with a = a + 1 and x = x + 2 (S20-2).

3) If H = 120, compare CP1 to less than H = 120 and repeat CP2. If CP1 is greater than 120, change to H = 60x4 and repeat (S20-3).

4) After repeating up to a = 5000, the squared sum (Het) is divided by half power (root), divided by 60Hz component (L60), and multiplied by 100 to get total Even Harmonic distortion (%). (S20-4).

5) The value is stored at each predetermined time interval (The), and the level is compared with the reference value to display and display “DANGER” and “Warning” and go to the odd harmonic calculation routine (S20-5).

All. Odd harmonic current harmonic calculation algorithm (S30)

5 is a detailed flowchart of an odd harmonic current harmonic calculation algorithm S30 in the fault detection method by circuit analysis of an induction motor winding according to the present invention. The processing of this algorithm is as follows.

1) Read CPa, CLa (a = 1): Read the frequency CP1 and level CL1 of the component having the peak in the spectrum (S30-1).

2) Let H = 60x1 (x = 1). If CP1 is H = 60, the value CL1 is squared and added, and 1) is repeated with a = a + 1 and x = x + 2 (S30-2).

3) If H = 60, compared to CP1 value, if CP is less than H = 60, CP2 is read and 2) is repeated. If CP1 is greater than 60, H = 60x3 is changed to 1) (S30-3).

4) After repeating up to a = 5000, multiply the added value (Het) by 1/2 power (root), divide by 60Hz component (L60), and multiply by 100 to get the total odd harmonic distortion (%) (S30-4).

5) Store the value at the predetermined time interval (Tho), compare the level with the reference value, display and display “Hazard”, “Warning” and go to the H + ve calculation routine (S30-5).

la. Positive sequence harmonic current harmonic calculation algorithm (S40)

The processing procedure of this algorithm S40 is as follows.

1) Obtain both sequence harmonic harmonics (4,7,10 ..) from the stored peak list (S40-1).

2) Check that the harmonic component is within the error range and store it (S40-2).

3) The total amount of sequence harmonics is obtained, stored, and the failure is determined by comparing with the set danger and warning levels (S40-3). At this time, if the danger or warning level is exceeded, the rotational force acts in the direction in which the RPM of the motor increases, or heat is generated in the stator.

hemp. Negative sequence harmonic current harmonic algorithm (S50)

The processing procedure of this algorithm S50 is as follows.

1) The subsequence harmonic harmonic (3,5,8 ..) component is obtained from the stored peak list (S50-1).

2) Check if the harmonic harmonic component is within the error range and store it (S50-2).

3) Obtain and store the total Bush Harmonic Harmonic Harmonic Component and determine the failure by comparing with the set danger and warning level (S50-3). If the total Busquency Harmonic Harmonic component obtained at this point exceeds the danger or warning level, the rotational force acts in the reverse direction to the motor, increases current, generates heat, or damages to bearings, couplings and rotors. Therefore, it is considered a failure.

bar. Zero sequence harmonic current harmonic calculation algorithm (S60)

The processing procedure of this algorithm S60 is as follows.

1) The zero sequence harmonic harmonic (3,6,9 ..) component is obtained from the stored peak list (S60-1).

2) Check and store the harmonic harmonic components within the error range (S60-2).

3) Obtain and store the total zero sequence harmonic harmonic components and determine the failure by comparing with the set danger and warning levels (S60-3). At this time, if the danger or warning level is exceeded, it is determined as a failure. The symptom of this time is that heat is generated in the motor, but the rotation is not affected.

The six algorithms described above are executed in sequence to detect faults by analyzing the circuit of the inductively conductive winding.

This invention may be variously modified and may take various forms and only detailed embodiments thereof are described in the detailed description of the invention. It is to be understood, however, that the invention is not limited to the specific forms referred to in the description of the invention, but rather includes all modifications, equivalents, and substitutions within the spirit and scope of the invention as defined by the appended claims. It should be understood to do.

The failure detection method by circuit analysis of the stator winding using the technical idea of the present invention can be applied to the failure prediction of the induction motor.

1 is an exemplary view showing a general failure mode and pattern of an induction motor.

2 is a flow chart of a fault detection method by circuit analysis of an induction motor winding according to the present invention;

3 is a flowchart illustrating an extraction algorithm of power harmonics and non-power harmonics in the fault detection method by circuit analysis of an induction motor winding according to the present invention;

4 is a flowchart showing an even harmonic current harmonic calculation algorithm in a fault detection method by circuit analysis of an induction motor winding according to the present invention;

5 is a flowchart illustrating an odd harmonic current harmonic calculation algorithm in a fault detection method by circuit analysis of an induction motor winding according to the present invention;

6A is a frequency spectrum diagram of a stator current signal of a good induction motor without failure under full load conditions.

6b is a frequency spectrum diagram of a stator current signal of an induction motor having a line fault under full load conditions.

Claims (7)

To detect the failure of an induction motor, Extracting power harmonics and non-power harmonics (S10); Calculating an even harmonic current harmonic (S20); Calculating odd harmonic current harmonics (S30); Calculating a positive sequence harmonic current harmonic (S40); Calculating a sub-sequence harmonic current harmonic (S50); And Calculating a zero sequence harmonic current harmonic (S60), The above step (S10 ~ S60) characterized in that performed sequentially, fault detection method by the circuit analysis of the induction motor winding. The method of claim 1, wherein the extracting the power source harmonics and non-power source harmonics (S10), Reading frequency CP1 and level CL1 values of peak components in the spectrum (S10-1); Confirming that an integer multiple of 60 Hz (within ± 0.2%) of the frequencies of the peak components read above (S10-2); The part to be an integer is harmonic PH1 of the integer component and the peak level PL1 of the integer component, and if not, converting to NPH1 and NPL1 (S10-3); Calculating TPH by adding the square of PL1, which is the level of the integer component PH1, to continue adding T1; Calculating TNPH by adding a square of NPL1, which is a non-integer level of the component (S10-5); Repeating the above steps (S10-1 to S10-5) to a predetermined number a1 (S10-6); Completion of the above predetermined number of times, calculating the total harmonic distortion (%) by dividing the TPH and TNPH values by a power of 1/2 (root) by the level (60) at 60 Hz (S10-7). ); And And storing (S10-8) displaying and displaying "" danger "," "warning" "by storing the value at a predetermined time interval and comparing the level with the reference value. At this time, the frequency analysis is characterized in that the high-resolution processing, fault detection method by the circuit analysis of the induction motor winding. The method of claim 1, wherein the calculating of the even harmonic current harmonic (S20), Reading the frequency CP1 and the level CL1 of the component having the peak in the spectrum (S20-1) Setting H = 60x2, and if CP1 is H = 120, square the sum value CL1, add a2 = a2 + 1, and set x1 = x1 + 2 to repeat S20-1 (S20-2); If H = 120, compare CP1 to less than H = 120, repeat CP2, and if CP1 is greater than 120, change to H = 60x4 and repeat (S20-3); After repeating the above steps (S20-1 to S20-3) to a predetermined number a2, the squared sum (Het) is added to the power of 1/2 (root) and divided by the component (60) of 60 Hz (total even harmonic distortion) (Total Obtaining Even Harmonic distortion (%) (S20-4); And And storing the value at each predetermined time interval (The) and comparing the level with the reference value (“20”) to display “danger” and “warning”. At this time, the frequency analysis is characterized in that the high-resolution processing, fault detection method by the circuit analysis of the induction motor winding. The method of claim 1, wherein calculating the odd harmonic current harmonic (S30), Reading the frequency CP1 and the level CL1 of the component having the peak in the spectrum (S30-1); If H = 60x1 (x = 1), CP1 is H = 60, the value CL1 is squared and added, and a3 = a3 + 1, x2 = x2 + 2, and the above steps (S30-1) are repeated. Step S30-2; If H = 60, compare CP1 to less than H = 60. Read CP2 and repeat step S30-2. If CP1 is greater than 60, change to H = 60x3 and go to step S30-1. Return step (S30-3); After repeating up to a3 a3 times, multiplying the added value (Het) by 1/2 power (root) to divide by 60Hz component (L60) to obtain the total odd harmonic distortion (%) (S30). -4); And And a step (S30-5) of storing a value at a predetermined time interval (Tho) and comparing the level with a reference value and displaying “danger” and “warning”. At this time, the frequency analysis is characterized in that the high-resolution processing, fault detection method by the circuit analysis of the induction motor winding. The method of claim 1, wherein the calculating of the above positive sequence harmonic current harmonic (S40), Obtaining both sequence harmonic harmonics (4,7,10 ..) components from the stored peak list (S40-1); Checking and storing whether the harmonic component is within an error range (S40-2); And Obtaining and storing the total amount of sequence harmonic components, and comparing the set risk and warning level to determine the failure (S40-3), At this time, in the above step (S40-3), when the rotational force acts in the direction of increasing the RPM of the motor, heat is generated in the stator, or if the danger or warning level is exceeded, it is determined as a failure, Frequency analysis is processed in high resolution, A method for detecting faults by circuit analysis of induction motor windings, comprising calculating the rotational speed of an induction motor using an interpolation method of a window function used for frequency analysis. The method of claim 1, wherein the calculating of the subsequence harmonic current harmonic (S50), Obtaining a subsequence harmonic harmonic (3,5,8 ..) component from the stored peak list (S50-1); Checking and storing whether the harmonic harmonic component is within an error range (S50-2); And And calculating (S50-3) to obtain and store the total bush sequence harmonic harmonic component and determine the failure by comparing with the set risk and warning level. At this time, in the above step (S50-3), if the total bushing harmonic harmonic component obtained exceeds the danger or warning level, the torque acts in the reverse direction to the motor, the current increases, or the heat is generated, Frequency analysis is processed in high resolution, A method for detecting faults by circuit analysis of induction motor windings, comprising calculating the rotational speed of an induction motor using an interpolation method of a window function used for frequency analysis. The method of claim 1, wherein the calculating of the zero sequence harmonic current harmonic (S60), Obtaining a zero sequence harmonic harmonic (3, 6, 9.) component from the stored peak list (S60-1); Checking and storing whether the harmonic harmonic component is within an error range (S60-2); And Obtaining and storing the total zero sequence harmonic harmonic component and comparing the set risk and warning levels to determine a failure (S60-3). At this time, if the danger or warning level is exceeded in the above step (S60-3), or if the heat is generated from the motor but there is no influence on the rotation, or the overcurrent and the circulating current flow through the neutral wire due to the nonlinear load, it is determined as a failure. and, Frequency analysis is processed in high resolution, A method for detecting faults by circuit analysis of induction motor windings, comprising calculating the rotational speed of an induction motor using an interpolation method of a window function used for frequency analysis.

Images