Permanent magnet synchronous motor turn-to-turn short circuit fault diagnosis method based on current residual error dual normalization

Abstract

The invention discloses a permanent magnet synchronous motor turn-to-turn short circuit fault diagnosis method based on current residual error dual normalization, and the method comprises the following steps: calculating a dq-axis reference voltage and a dq-axis current according to collected three-phase voltage, three-phase current and electrical angle information; calculating an estimated value of dq-axis current according to a linear expansion observer; obtaining a dq-axis current residual error according to the actual value and the estimated value of the dq-axis current; calculating a negative-sequence direct-current component of a dq-axis current residual error by utilizing negative-sequence rotation coordinate transformation; performing amplitude normalization processing on the negative-sequence direct-current component of the dq-axis current residual error to obtain an amplitude-normalized negative-sequence direct-current component; performing data credibility check on the signal with the normalized amplitude, and latching the data when a trigger condition is met; performing frequency normalization processing on the dq-axis current residual error to obtain a negative sequence direct current component with double normalization of amplitude and frequency; calculating the quadratic sum of the normalized dq-axis negative sequence direct-current component so as to construct fault features; and comparing the fault feature with a fault threshold, and defining a fault indicator to map the state of the permanent magnet synchronous motor. According to the invention, the diagnosis of the turn-to-turn short circuit fault under the servo working condition can be realized without additional detection equipment.

Description

Technical Field

[0001] The present invention relates to the technical field of motor fault diagnosis, and in particular to a method for diagnosing inter-turn short-circuit faults of permanent magnet synchronous motors based on current residual double normalization. Background Art

[0002] High precision and fast response have always been the goals of high-performance permanent magnet synchronous motor (PMSM) servo systems. However, long-term operation in challenging environments, such as inappropriate temperatures, excessive humidity, severe vibration, heavy loads, and rapidly switching operating states, makes PMSM servo systems more susceptible to failure. When PMSM insulation components age or wear out more rapidly due to these factors, inter-turn short-circuit faults (ISCFs) become inevitable. Due to the high circulating current passing through the short-circuited winding, increasing losses and creating localized hot spots, ISCFs can severely damage the PMSM servo system, accelerating fault propagation and rapidly leading to secondary failures. In this process, the PMSM servo system may shut down, suffer damage, or even suffer catastrophic failures. Therefore, to ensure the stability of PMSM servo systems, effective fault diagnosis is crucial.

[0003] However, the inter-turn short-circuit fault characteristics under variable speed and variable load servo conditions show time domain non-periodic characteristics, which makes the inter-turn short-circuit fault diagnosis method based on steady-state design likely to cause misdiagnosis under servo conditions, and its fault diagnosis reliability and robustness are poor. Summary of the Invention

[0004] Purpose of the invention: The present invention provides a method for diagnosing inter-turn short-circuit faults of permanent magnet synchronous motors based on double normalization of current residuals, which can realize the diagnosis of inter-turn short-circuit faults under servo conditions without the need for additional detection equipment.

[0005] Technical solution: The present invention provides a method for diagnosing inter-turn short-circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals, comprising the following steps:

[0006] Step 1: Calculate the dq axis reference voltage u based on the collected three-phase voltage, three-phase current and electrical angle information d ,u q , dq axis current i d ,i q ;

[0007] Step 2: Calculate the estimated value i of the dq axis current according to the linear expansion observer de ,i qe ;

[0008] Step 3: According to the actual value and estimated value of the dq axis current and the q axis current DC component, the dq axis current residual Δi is obtained. d ,Δiq ;

[0009] Step 4: Calculate the negative sequence DC component of the dq axis current residual using the negative sequence rotating coordinate transformation

[0010] Step 5: Normalize the amplitude of the negative sequence DC component of the dq axis current residual to obtain the amplitude normalized negative sequence DC component

[0011] Step 6: Check the data credibility of the amplitude-normalized signal, and latch the data when the trigger condition is met;

[0012] Step 7: Perform frequency normalization on the dq axis current residuals to obtain the negative sequence DC component with amplitude and frequency double normalization.

[0013] Step 8: Calculate the square sum of the normalized dq axis negative sequence DC components to construct the fault signature

[0014] Step 9: Compare the fault signature with the fault threshold, and define a fault indicator to map the state of the permanent magnet synchronous motor.

[0015] Furthermore, in step 1, according to the DC bus voltage u dc , three-phase space vector pulse width modulation switching signal S a ,S b ,S c , the three-phase voltage u is obtained based on the following formula a ,u b ,u c ,

[0016] u x =u dc (2S x -S x+1 -S x+2 ) / 3

[0017] Among them, u x is the phase voltage of phase x, where x represents one of phases a, b, and c; x+1 represents the phase that lags behind phase x by 120 degrees, and x+2 represents the phase that lags behind phase x by 240 degrees;

[0018] According to the motor three-phase current i a ,i b ,i c , three-phase voltage u a ,u b ,u c , using coordinate transformation to obtain the dq axis current i d ,i q , dq axis voltage ud ,u q .

[0019] Furthermore, in step 2, the estimated value i of the dq axis current is obtained according to the following formula: de ,i qe ,

[0020]

[0021] in, and Respectively represent the expanded state variables of dq axis current; R s Indicates the stator resistance of the permanent magnet synchronous motor, L d and L q They represent the dq axis inductance of the permanent magnet synchronous motor respectively; ω e is the electrical angular velocity of the permanent magnet synchronous motor; β1 and β2 are the gain coefficients of the linear extended state observer. When the bandwidth adjustment method is used, the gain coefficient of the linear extended state observer satisfies β1=2ω o , ω o is the bandwidth adjustment parameter.

[0022] Furthermore, in step 3, according to the estimated value i of the dq axis current de ,i qe , the dq axis current residual Δi is obtained based on the following formula d ,Δi q ,

[0023]

[0024] Among them, i d ,i q is the dq axis current.

[0025] Furthermore, in step 4, the negative sequence DC component of the dq axis current residual is calculated by negative sequence rotation transformation

[0026]

[0027] Among them, θ e is the rotor position angle.

[0028] Furthermore, in step 5, the negative sequence DC component is amplitude normalized to obtain the negative sequence DC component with normalized amplitude

[0029]

[0030] Where k is a constant that scales the voltage amplitude to the same order of magnitude as the current harmonics; U baseis the amplitude of the phase voltage, and its value can be calculated from the dq axis voltage.

[0031]

[0032] Furthermore, in step 6, when the permanent magnet synchronous motor is in a stable state, the fluctuation of the signal amplitude data is less than the threshold of the data latch, and the data of the current sampling cycle is directly output; when the permanent magnet synchronous motor has a state step, the fluctuation of the signal amplitude data is greater than the threshold of the data latch, and the data latch is the output of the previous sampling cycle.

[0033] Furthermore, in step 7, the current residual is frequency normalized to obtain the negative sequence DC component with amplitude and frequency double normalization.

[0034]

[0035] Where X is the number of equal divisions of the discretized integration window.

[0036] Furthermore, in step 8, the negative sequence DC component is normalized by amplitude and frequency. Structural fault characteristics

[0037]

[0038] Furthermore, in step 9, based on the comparison between the fault characteristics and the fault threshold, a fault indicator (FI) is defined to map the operating status of the permanent magnet synchronous motor. When the fault characteristics are less than the fault threshold, the fault indicator FI outputs 0, indicating that the permanent magnet synchronous motor is in a healthy state; when the fault characteristics are greater than the fault threshold, the fault indicator FI outputs 1, indicating that the permanent magnet synchronous motor is in a turn-to-turn short circuit fault state.

[0039] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: (1) Fault feature construction based on closed-loop state observer: The current residual is generated by subtracting the healthy current estimated by LESO from the actual fault current, and the negative sequence DC component in the harmonics is extracted using the current residual as the fault feature. The negative sequence DC component of the current residual is used for ISCF diagnosis, which does not rely on additional detection equipment, has little dependence on system parameters, and has strong adaptability; (2) Double normalization technology of amplitude and frequency: In order to improve the reliability and robustness of diagnosis under servo conditions, the double normalization technology is adopted to reduce the influence of speed and load changes on fault features, which is very suitable for online fault diagnosis in variable speed and variable load servo applications; (3) Signal reliability detection and data latching technology: The signal reliability detection and data latching technology are adopted to reduce the probability of misdiagnosis when the motor state changes step by step. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Schematic diagram of the inter-turn short circuit fault involved in the present invention.

[0041] Figure 2 This is a fault diagnosis flow chart of the present invention.

[0042] FIG3( a ) is a schematic diagram of d-axis current estimation according to the present invention.

[0043] FIG3( b ) is a schematic diagram of q-axis current estimation according to the present invention.

[0044] FIG4( a ) is a speed diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0045] FIG4( b ) is a phase current diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0046] FIG4( c ) is a fault characteristic diagram of the PMSM of the present invention when it changes from a healthy state to a fault state.

[0047] FIG4( d ) is an output diagram of the fault indicator FI when the PMSM of the present invention changes from a healthy state to a fault state.

[0048] FIG5( a ) is a speed diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0049] FIG5( b ) is a phase current diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0050] FIG5( c ) is a fault characteristic diagram of the PMSM of the present invention when it changes from a healthy state to a fault state.

[0051] FIG5( d ) is an output diagram of the fault indicator FI when the PMSM of the present invention changes from a healthy state to a fault state.

[0052] FIG6( a ) is a speed diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0053] FIG6( b ) is a phase current diagram of the PMSM of the present invention when the PMSM changes from a healthy state to a fault state.

[0054] FIG6( c ) is a fault characteristic diagram of the PMSM of the present invention when it changes from a healthy state to a fault state.

[0055] FIG6( d ) is an output diagram of the fault indicator FI when the PMSM of the present invention changes from a healthy state to a fault state. DETAILED DESCRIPTION

[0056] When a permanent magnet synchronous motor experiences an inter-turn short-circuit fault, an additional second harmonic component is introduced into the dq-axis current. Leveraging this characteristic, a linear extended state observer is used to calculate the estimated dq-axis currents based on the motor's three-phase voltages and currents. The dq-axis current residuals are then calculated based on the actual and estimated dq-axis currents. The negative-sequence DC component of the dq-axis current residuals is calculated using a negative-sequence rotating coordinate transformation, and amplitude and frequency dual normalization is applied to the negative-sequence DC component. A signal credibility check and latching mechanism are employed to mitigate misdiagnosis caused by motor state steps. Finally, a fault signature is constructed based on the negative-sequence DC component of the dq-axis current residuals and compared with a fault threshold to determine whether the motor has failed.

[0057] The present invention provides a method for diagnosing inter-turn short-circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals, comprising the following steps:

[0058] Step 1: Calculate the dq axis reference voltage u based on the collected three-phase voltage, three-phase current and electrical angle information d ,u q , dq axis current i d ,i q ;

[0059] Step 2: Calculate the estimated value of the dq axis current i according to the linear expansion observer de ,i qe ;

[0060] Step 3: According to the actual value and estimated value of the dq axis current and the q axis current DC component, the dq axis current residual Δi is obtained. d ,Δi q ;

[0061] Step 4: Calculate the negative sequence DC component of the dq axis current residual using the negative sequence rotating coordinate transformation

[0062] Step 5: Normalize the negative sequence DC component of the dq axis current residual to obtain the negative sequence DC component with normalized amplitude.

[0063] Step 6: Check the data credibility of the amplitude-normalized signal, and latch the data when the trigger condition is met;

[0064] Step 7: Perform frequency normalization on the dq axis current residuals to obtain the negative sequence DC component with amplitude and frequency double normalization.

[0065] Step 8: Calculate the square sum of the normalized dq axis negative sequence DC components to construct the fault signature

[0066] Step nine: compare the fault signature with the fault threshold, and define a fault indicator to map the state of the permanent magnet synchronous motor.

[0067] In a specific embodiment of the present invention, a schematic diagram of a permanent magnet synchronous motor generating ISCF is shown as follows: Figure 1 ISCF will cause the stator winding of the permanent magnet synchronous motor to be unbalanced. Usually an additional short-circuit branch is used to simulate ISCF, where the fault resistance R f To simulate the degree of winding insulation degradation, the fault current i f ISCF will cause the phase current amplitude to be unbalanced, thereby generating additional second harmonic components in the dq axis current. Figure 1 to simulate.

[0068] In an embodiment of the present invention, a motor drive system includes a DC voltage source, an inverter circuit, a drive circuit, a current and voltage sampling circuit, a controller, and a protection circuit. The DC voltage source provides the system with DC bus voltage. The inverter circuit converts DC power into the three-phase AC power required to control the motor. The drive circuit drives the power devices within the inverter. The sampling circuit monitors voltage and current signals in real time. The controller generates PWM signals based on the sampled data to control the motor. The protection circuit ensures safe operation of the system in the event of an abnormality.

[0069] The specific parameters of the PMSM in this embodiment include: stator phase resistance R s =3Ω, direct-axis inductance L d =30mH, quadrature-axis inductance L q =30mH, permanent magnet fundamental flux ψ f =0.82Wb, moment of inertia J = 0.1337kg·m 2 , pole pair number n p =4.

[0070] The specific experimental conditions are as follows: the DC bus voltage of the PMSM is 510V; the sampling frequency and switching frequency are 10kHz; the speed under steady-state conditions is 300r / min, and the load torque is 5.2N·m; the servo operating conditions include variable speed and variable load conditions: under variable speed conditions, the speed steps from 300r / min to 500r / min and then returns to 300r / min, and the load torque is constant at 5.2N·m; under variable load conditions, the load torque steps from 5.2N·m to 10.2N·m and then returns to 5.2N·m, and the speed is constant at 300r / min.

[0071] Fault diagnosis flow chart as follows Figure 2 As shown, the specific steps included in the embodiment are as follows:

[0072] The three-phase voltage can be calculated from the DC bus voltage and switching signals of the inverter as follows:

[0073] u x =u dc (2S x -S x+1 -S x+2 ) / 3 (1)

[0074] Among them, u x is the phase voltage of phase x, x represents one of phases a, b, and c, x+1 represents the phase that lags behind phase x by 120 degrees, and x+2 represents the phase that lags behind phase x by 240 degrees; u dc is the DC bus voltage; S x Represents the switching signal of space vector pulse width modulation in the x-th phase.

[0075] The phase voltage and phase current in the three-phase stationary abc coordinate system can be converted into the dq axis voltage and current in the two-phase rotating dq coordinate system through the transformation matrix. The transformation matrix is ​​as follows:

[0076]

[0077] Among them, θ e is the rotor position angle of the PMSM. Therefore, the dq axis voltage and current can be calculated as:

[0078]

[0079] Among them, u dq =[u d ,u q ] T ,u abc =[u a ,u b ,u c ] T ;i dq =[i d ,i q ] T ,i abc =[i a ,i b ,i c ] T .

[0080] The estimated values ​​of the dq-axis currents are calculated based on the linear extended state observer (LESO). The schematic diagrams of the LESO structure are shown in Figure 3(a) and Figure 3(b).

[0081] In LESO, the estimated values ​​of the dq-axis currents are calculated as follows:

[0082]

[0083] in, and They represent the expanded state variables of dq axis current respectively; R s Indicates the stator resistance of the permanent magnet synchronous motor, L d and L q They represent the dq axis inductance of the permanent magnet synchronous motor respectively; ω e is the electrical angular velocity of the permanent magnet synchronous motor; β1 and β2 are the gain coefficients of the linear extended state observer. When the bandwidth adjustment method is used, the gain coefficient of the linear extended state observer satisfies β1=2ω o , ω o is the bandwidth adjustment parameter.

[0084] Once the estimated dq axis currents are obtained, the harmonic components of the dq axis current residuals can be calculated by subtracting the estimated values ​​from the actual values, i.e.:

[0085]

[0086] Among them, Δi d and Δi q are the dq axis current residuals respectively.

[0087] To diagnose faults, the second harmonic component in the dq axis current residuals needs to be extracted as a fault feature. The second harmonic component is extracted by applying a negative sequence rotation transform at the second harmonic frequency and converting it into a DC component. The negative sequence rotation transformation matrix is ​​as follows:

[0088]

[0089] However, the fault signature is highly dependent on the operating state of the PMSM. The amplitude and frequency of the electrical signal vary with time during the PMSM servo motion. In this case, the filtering performance of traditional filters is poor. To address this problem, this paper proposes a double normalization technique to effectively extract fault signatures.

[0090] For ISCF, the second harmonic component generated by the fault is proportional to the amplitude of the fault current, and the amplitude of the fault current is proportional to the voltage of the fault phase. Therefore, the amplitude of the current residual can be normalized by the phase voltage to eliminate the influence of the motor state change. The amplitude normalization calculation method is as follows:

[0091]

[0092] Where k is a constant that scales the voltage amplitude to the same order of magnitude as the current harmonics; U base is the amplitude of the phase voltage, and its value can be calculated from the dq axis voltage as follows:

[0093]

[0094] Then, the signal is checked for credibility and latched, including: when the permanent magnet synchronous motor is in a stable state, the fluctuation of the signal amplitude data is less than the data latch threshold, and the data of the current sampling cycle is directly output; when the permanent magnet synchronous motor has a state step, the fluctuation of the signal amplitude data is greater than the data latch threshold, and the data is latched as the output of the previous sampling cycle.

[0095] Since the time-varying frequency of the electrical signal remains constant relative to the angular position of the motor, frequency normalization can be performed in the angular domain. The DC component can be obtained by integrating all other harmonic components within one electrical cycle. Therefore, the integration result can be expressed as:

[0096]

[0097] in, is the dq axis negative sequence DC component with double normalization of amplitude and frequency. In order to realize it in a digital processor, the continuous integral needs to be transformed into a discrete form for practical implementation, that is:

[0098]

[0099] where the integration window is divided into equal X parts, and θ j is the electrical angle sampled at time j=1,…,X.

[0100] Finally, the normalized fault signature based on the DC component and the negative sequence DC component is constructed as follows:

[0101]

[0102] Among them, M dcn This is a fault characteristic of ISCF.

[0103] Finally, a fault indicator (FI) is defined to map the operating status of the permanent magnet synchronous motor. When the fault characteristic is less than the fault threshold, the fault indicator FI output is 0, indicating that the permanent magnet synchronous motor is in a healthy state; when the fault characteristic is greater than the fault threshold, the fault indicator FI output is 1, indicating that the permanent magnet synchronous motor is in a turn-to-turn short-circuit fault state.

[0104] The present invention implements a turn-to-turn short-circuit fault diagnosis method for permanent magnet synchronous motors based on double normalization of current residuals, without the need for additional detection equipment, by utilizing LESO and double normalization technology to achieve turn-to-turn short-circuit fault diagnosis under servo conditions. The experimental results of fault diagnosis under steady-state conditions are shown in Figures 4(a)-(d); the experimental results of fault diagnosis under speed variation are shown in Figures 5(a)-(d); and the experimental results of fault diagnosis under load variation are shown in Figures 6(a)-(d). The experimental results demonstrate that the present invention can achieve highly reliable and robust diagnosis of turn-to-turn short-circuit faults in permanent magnet synchronous motors under variable speed and load servo conditions.

[0105] The present invention first estimates the dq-axis current through LESO, calculates the negative-sequence DC component of the current residual, and combines signal credibility check, latching mechanism and double normalization technology to reduce the influence of the motor operating state on the fault characteristics, thereby achieving accurate diagnosis of inter-turn short-circuit faults without the need for additional detection equipment, and has good applicability and robustness.

Claims

1. A method for diagnosing inter-turn short-circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals, characterized in that: The steps include: Step 1: Calculate the dq axis reference voltage u based on the collected three-phase voltage, three-phase current and electrical angle information d ,u q , dq axis current i d ,i q ; Step 2: Calculate the estimated value i of the dq axis current according to the linear expansion observer de ,i qe ; Step 3: According to the actual value and estimated value of the dq axis current and the q axis current DC component, the dq axis current residual Δi is obtained. d ,Δi q ; Step 4: Calculate the negative sequence DC component of the dq axis current residual using the negative sequence rotating coordinate transformation Step 5: Normalize the amplitude of the negative sequence DC component of the dq axis current residual to obtain the amplitude normalized negative sequence DC component Step 6: Check the data credibility of the amplitude-normalized signal, and latch the data when the trigger condition is met; Step 7: Perform frequency normalization on the dq axis current residuals to obtain the negative sequence DC component with amplitude and frequency double normalization. Step 8: Calculate the square sum of the normalized dq axis negative sequence DC components to construct the fault signature Step 9: Compare the fault signature with the fault threshold, and define a fault indicator to map the state of the permanent magnet synchronous motor.

2. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 1, according to the DC bus voltage u dc , three-phase space vector pulse width modulation switching signal S a ,S b ,S c , the three-phase voltage u is obtained based on the following formula a ,u b ,u c , u x =u dc (2S x -S x+1 -S x+2 ) / 3 Among them, u x is the phase voltage of phase x, where x represents one of phases a, b, and c; x+1 represents the phase that lags behind phase x by 120 degrees, and x+2 represents the phase that lags behind phase x by 240 degrees; According to the motor three-phase current i a ,i b ,i c , three-phase voltage u a ,u b ,u c , using coordinate transformation to obtain the dq axis current i d ,i q , dq axis voltage u d ,u q .

3. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 2, the estimated value of the dq axis current i is obtained according to the following formula: de ,i qe , in, and Respectively represent the expanded state variables of dq axis current; R s Indicates the stator resistance of the permanent magnet synchronous motor, L d and L q They represent the dq axis inductance of the permanent magnet synchronous motor respectively; ω e is the electrical angular velocity of the permanent magnet synchronous motor; β1 and β2 are the gain coefficients of the linear extended state observer. When the bandwidth adjustment method is used, the gain coefficient of the linear extended state observer satisfies β1=2ω o , ω o is the bandwidth adjustment parameter.

4. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 3, according to the estimated value of dq axis current i de ,i qe , the dq axis current residual Δi is obtained based on the following formula d ,Δi q , Among them, i d ,i q is the dq axis current.

5. The method for diagnosing inter-turn short circuit fault of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 4, the negative sequence DC component of the dq axis current residual is calculated by negative sequence rotation transformation Among them, θ e is the rotor position angle.

6. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 5, the negative sequence DC component is normalized to obtain the negative sequence DC component with normalized amplitude. Where k is a constant that scales the voltage amplitude to the same order of magnitude as the current harmonics; U base is the amplitude of the phase voltage, and its value is calculated by the dq axis voltage.

7. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 6, when the permanent magnet synchronous motor is in a stable state, the fluctuation of the signal amplitude data is less than the threshold of the data latch, and the data of the current sampling cycle is directly output; when the permanent magnet synchronous motor has a state step, the fluctuation of the signal amplitude data is greater than the threshold of the data latch, and the data latch is the output of the previous sampling cycle.

8. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 7, the current residual is frequency normalized to obtain the negative sequence DC component with amplitude and frequency double normalization. Where X is the number of equal divisions of the discretized integration window.

9. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 8, the negative sequence DC component is normalized by amplitude and frequency. Structural fault characteristics 10. The method for diagnosing inter-turn short circuit faults of a permanent magnet synchronous motor based on double normalization of current residuals according to claim 1, characterized in that: In step 9, based on the comparison between the fault characteristics and the fault threshold, a fault indicator FI is defined to map the operating status of the permanent magnet synchronous motor. When the fault characteristics are less than the fault threshold, the fault indicator FI outputs 0, indicating that the permanent magnet synchronous motor is in a healthy state; when the fault characteristics are greater than the fault threshold, the fault indicator FI outputs 1, indicating that the permanent magnet synchronous motor is in a turn-to-turn short circuit fault state.

Images