A DFIG Rotor Resistance Unbalance Fault Diagnosis Method and System

Abstract

The present invention discloses a method and system for diagnosing the DFIG rotor resistance imbalance fault, including: establishing a decoupling model of the dq-axis fault signals of the DFIG rotor winding under the rotor resistance imbalance fault; selecting the fault diagnosis sampling signals and determining the fault characteristic parameters; establishing a method for designing the current closed-loop bandwidth; designing a method for diagnosing the rotor resistance imbalance fault; and designing a fault diagnosis process. Considering the influence of the current closed-loop control on the fault characteristic signals, while ensuring the control performance, the present invention proposes the design of the current closed-loop bandwidth, which can effectively eliminate the interference of the current closed-loop control on the diagnosis of the DFIG rotor winding resistance imbalance fault and ensure the accurate and reliable diagnosis of the rotor resistance imbalance fault; by designing reasonable fault characteristic quantities, a simple and reliable fault diagnosis process is realized, without adding new sensors, and only a simple fault parameter extraction algorithm is added to the original controller, so that the on-site real-time fault detection, location and fault degree evaluation can be realized.

Description

Technical Field

[0001] The present invention relates to a method and system for diagnosing rotor resistance imbalance faults, and particularly to a method and system for diagnosing DFIG rotor resistance imbalance faults. Background Art

[0002] As an effective way to solve the energy crisis and environmental pollution problems, wind power generation has developed rapidly worldwide. With the development of offshore wind power technology, far - sea, deep - water, and large - scale offshore wind farms have become a major trend in the development of offshore wind power. In a doubly - fed induction wind power generation system, the frequency converter only flows through the slip power, which has the advantages of small capacity, small investment loss, high power generation efficiency, and convenient harmonic absorption. Therefore, the doubly - fed induction generator (DFIG) has become one of the mainstream models of offshore wind generators. Compared with the on - shore environment, the operating environment of offshore DFIGs is more severe, with a higher failure rate, difficult on - site maintenance, and huge losses caused by fault shutdowns. Therefore, it is urgent to detect electrical defects of offshore wind generators early to avoid catastrophic accidents caused by the deterioration of faults.

[0003] Winding resistance imbalance fault is one of the common electrical faults of DFIGs. It is found that winding resistance imbalance faults will lead to an increase in motor torque ripple, a decrease in average torque, and current and voltage imbalance. Continuous winding resistance imbalance fault states may lead to overheating, which further increases the connection resistance. If these positive feedback loops are not broken, winding resistance imbalance faults may spread to uncontrolled consequences and even lead to catastrophic failures. Therefore, it is of great significance to detect winding resistance imbalance faults and prevent potential serious damage.

[0004] At present, various fault characteristics based on voltage and current signals have been proposed for diagnosing winding resistance imbalance faults. However, due to the connection of the frequency converter on the rotor side and the direct grid connection of the stator winding in DFIGs, the different designs of closed - loop controllers increase the difficulty of identifying rotor winding resistance imbalance faults with voltage and current as fault characteristics. Therefore, it is necessary to explore a diagnostic method for DFIG rotor winding resistance imbalance faults suitable for closed - loop control with a frequency converter. Summary of the Invention

[0005] Object of the Invention: The object of the present invention is to provide a method and system for diagnosing DFIG rotor resistance imbalance faults, which can effectively eliminate the interference of current closed - loop control on the diagnosis of DFIG rotor winding resistance imbalance faults and ensure the accurate and reliable diagnosis of rotor resistance imbalance faults.

[0006] Technical solution: The present invention includes the following steps: establishing a decoupling model for dq-axis fault signals of the DFIG rotor winding under rotor resistance imbalance faults; selecting fault diagnosis sampling signals and determining fault characteristic parameters; establishing a design method for the current closed-loop bandwidth; designing a rotor resistance imbalance fault diagnosis method; and designing a fault diagnosis process.

[0007] The establishment of the decoupling model for dq-axis fault signals of the DFIG rotor winding under rotor resistance imbalance faults specifically includes:

[0008] Using the Park transformation, the dq-axis voltage equations of the DFIG rotor winding are obtained as:

[0009]

[0010] In the formula: u dr , u qr are the dq-axis voltages of the rotor; i dr , i qr are the dq-axis currents of the rotor; R r represents the resistance of the rotor winding, and R is the additional resistance connected in series in the fault phase to simulate the rotor resistance imbalance fault; σ = 1 - L 2 m / (L s L r ) is the leakage magnetic coefficient of the DFIG; L s = L ls + 3 / 2L ms ; L r = L lr + 3 / 2L ms ; L m = 3 / 2L ms ; L ls is the stator winding leakage inductance; L lr is the rotor winding leakage inductance; L ms is the maximum mutual inductance between the stator and the rotor; ω e = ω s - ω r , ω s is the electrical angular velocity of the stator magnetic field, and ω r is the electrical angular velocity of the rotor; e WRId and e WRIq are the additional terms introduced due to the rotor resistance imbalance fault, and θ e is the electrical angle of the rotor relative to the stator magnetic field;

[0011] Adopting stator magnetic field orientation and introducing feedforward compensation to achieve decoupling of the dq-axis on the rotor side, the decoupling model of the fault signal is obtained as:

[0012]

[0013] The fault diagnosis sampling signal is the voltage signal u PId , the voltage signal u PId is used as the fault diagnosis sampling signal, and the fault feature parameter u PId,2 is extracted.

[0014] The specific method for establishing the current closed-loop bandwidth design is as follows:

[0015] Construct a fault feature parameter model:

[0016]

[0017] In the formula: I dr,2 and θ idr,2 are the amplitude and initial phase of i dr,2 respectively, E WRId,2 and θ eWRId,2 are the amplitude and initial phase of e WRId,2 , respectively, U PId,2 and θ uPId,2 are the amplitude and initial phase of u PId,2 respectively, and there is

[0018]

[0019] In the formula: K p and K i are the proportional gain and integral gain respectively.

[0020] The amplitude and phase of the fault feature parameter u PId,2 satisfy

[0021]

[0022] The criterion for establishing the current closed-loop bandwidth design is: K p and K i take larger values within the range of excellent control performance.

[0023] The specific method for designing the rotor resistance unbalance fault diagnosis is as follows:

[0024] Extract the second-order component u PId online from the u PId,2 signal:

[0025]

[0026] Extract the amplitude of u PId,2 :

[0027]

[0028] Extract the initial phase of u PId,2 :

[0029]

[0030] Define C m as the current characteristic quantity, and its value is:

[0031]

[0032] In the formula: when m = a, n = 0; when m = b, n = 1; when m = c, n = -1; I cm and θ cm are respectively the amplitude and the initial phase of C m and are respectively

[0033]

[0034]

[0035] By calculating θ eWRId,2 , and comparing it with θ ca , θ cb and θ cc respectively, to judge the faulty phase;

[0036] The degree of the fault can be judged by calculating the additional resistance ΔR of the faulty phase:

[0037]

[0038] In the formula: I cm is the characteristic current amplitude of the faulty phase.

[0039] The online extraction of the second-order component u PId from the u PId,2 signal adopts a frequency tracking algorithm.

[0040] The designed fault diagnosis process, the diagnosis process specifically includes:

[0041] Fault detection: Using the frequency tracking algorithm to online extract the second-order component of u PId . If the amplitude U PId,2 is greater than the preset detection threshold U th , then the rotor resistance imbalance fault occurs;

[0042] Fault location: After detecting the rotor resistance imbalance fault, calculate e WRId,2 and the characteristic currents C a , C b and C c , and compare their initial phases. Among them, the one closest to θ e WRId,2 is the faulty phase;

[0043] Fault degree evaluation: After the faulty phase is located, through ΕWRId,2 The additional resistance value ΔR of the faulty phase is calculated from the amplitude of the characteristic current of the faulty phase.

[0044] A DFIG rotor resistance unbalance fault diagnosis system is used to implement the above DFIG rotor resistance unbalance fault diagnosis method.

[0045] Beneficial effects: Considering the influence of current closed-loop control on the fault characteristic signal, while ensuring the control performance, the present invention proposes a current closed-loop bandwidth design, which can effectively eliminate the interference of current closed-loop control on the DFIG rotor winding resistance unbalance fault diagnosis, and ensure the accurate and reliable diagnosis of rotor resistance unbalance faults; by designing reasonable fault characteristic quantities, a simple and reliable fault diagnosis process is realized, without adding new sensors, and only a simple fault parameter extraction algorithm is added to the original controller, and real-time fault detection, location and fault degree evaluation can be realized on-site. Description of the Drawings

[0046] Figure 1 is a schematic diagram of DFIG rotor resistance unbalance fault;

[0047] Figure 2 is a schematic diagram of dq-axis decoupling under DFIG rotor resistance unbalance fault;

[0048] Figure 3 is a schematic diagram of the influence of current closed-loop on DFIG rotor fault characteristic signal;

[0049] Figure 4 is a flowchart of DFIG rotor resistance unbalance fault diagnosis. Detailed Embodiments

[0050] The present invention will be further described below with reference to the accompanying drawings.

[0051] As Figure 1 shown, the DFIG rotor resistance unbalance fault diagnosis method of the present invention includes the following steps:

[0052] Step S1, establish a dq-axis fault signal decoupling model of the DFIG rotor winding under rotor resistance unbalance fault; specifically:

[0053] Using the Park transformation, the dq-axis voltage equations of the DFIG rotor winding in the stator magnetic field synchronous rotating coordinate system are obtained as

[0054]

[0055] where: u dr 、u qr are the dq-axis voltages of the rotor; i dr 、i qr are the dq-axis currents of the rotor; R rRepresents the resistance of the rotor winding, and R is the additional resistance connected in series in the faulty phase to simulate the rotor resistance unbalance fault; σ = 1 - L 2 m / (L s L r ) is the leakage magnetic coefficient of the DFIG; L s = L ls + 3 / 2L ms ; L r = L lr + 3 / 2L ms ; L m = 3 / 2L ms ; L ls is the leakage inductance of the stator winding; L lr is the leakage inductance of the rotor winding; L ms is the maximum mutual inductance between the stator and rotor; ω e = ω s - ω r , ω s is the electrical angular velocity of the stator magnetic field, ω r is the electrical angular velocity of the rotor; e WRId and e WRIq are the additional terms introduced due to the rotor resistance unbalance fault, and θ e is the electrical angle of the rotor relative to the stator magnetic field.

[0056] By adopting stator magnetic field orientation and introducing feedforward compensation to achieve decoupling of the dq axes on the rotor side, the decoupled model of the fault signal is obtained as:

[0057]

[0058] Step S2, select the fault diagnosis sampling signal and determine the fault characteristic parameters; among them, the fault diagnosis sampling signal is the voltage signal u PId . For a healthy DFIG, e WRId and e WRIq are equal to 0, and the ideal dq-axis currents and voltages are constant values. When a rotor resistance unbalance fault occurs, e WRId and e WRIq will introduce harmonics in the dq-axis currents and voltages. According to Equation (1), e WRId and e WRIq are mainly composed of second-order components and DC components. Therefore, the second-order components will appear in the dq-axis signals, such as the current signals i dr and i qr , as well as the voltage signals u PId and u PIq .

[0059] Since rotor resistance imbalance can introduce second-order torque into the DFIG, disturbances of the same frequency will be induced in the speed signal. Since the input i qr * of the current loop is the output of the speed closed-loop controller, the i qr * signal will also contain a second-order component i qr,2 *, which will be affected by the speed closed-loop controller and the mechanical parameters of the motor. Therefore, i qr,2 and u PIq,2 will also be disturbed by the speed closed-loop controller and the mechanical parameters of the motor, etc.

[0060] The DFIG current controller design adopts i dr * = 0 control, and there is no second-order component in the i dr * signal. Obviously, i dr,2 and u PId,2 only contain the second-order component e WRId in e WRId,2 , and are independent of the mechanical parameters of the motor. Therefore, using i dr,2 and u PId,2 for rotor resistance imbalance fault diagnosis can eliminate the interference of speed, etc. In order to avoid the influence of the asymmetry problem on the power grid introduced into the stator winding, measures should be taken to suppress i dr,2 while amplifying u PId,2 . Therefore, the voltage signal u PId is selected as the fault diagnosis sampling signal to extract the fault characteristic parameter u PId,2 .

[0061] Step S3, establish a current closed-loop bandwidth design method; specifically:

[0062] According to the action mechanism of the current closed-loop controller on the fault characteristic parameter u PId,2 , construct a fault characteristic parameter model as:

[0063]

[0064] Where: I dr,2 and θ idr,2 are the amplitude and initial phase of i dr,2 respectively, E WRId,2 and θ eWRId,2 are the amplitude and initial phase of e WRId,2 respectively, U PId,2 and θ uPId,2 are the amplitude and initial phase of u PId,2 respectively, and there is

[0065]

[0066] Where: K p and K iThey are the proportional gain and the integral gain respectively.

[0067] The amplitude and phase of the fault characteristic parameters satisfy

[0068]

[0069] According to Equation (4), |T id | and |T ud | are greatly affected by the values of K p and K i For a current closed-loop controller with high bandwidth, that is, when K p and K i are relatively large, the magnitude of |T id | is close to 0, while the magnitude of |T ud | is close to 1. According to Equation (5), i dr,2 is well suppressed by the current closed-loop with high bandwidth and only has a relatively small magnitude. While u PId,2 has a greatly increased magnitude in this case. On the contrary, for a current closed-loop controller with low bandwidth, the magnitude of u PId,2 is small, while i dr,2 has a relatively large magnitude.

[0070] Establish the design criterion for the bandwidth of the current closed-loop: K p and K i take larger values within the range of excellent control performance to ensure that the magnitude of |T id | is close to 0, while the magnitude of |T ud | is close to 1.

[0071] Step S4, design the rotor resistance unbalance fault diagnosis method; specifically:

[0072] Adopt the frequency tracking algorithm to extract the second-order component u PId online from the u PId,2 signal:

[0073]

[0074] Extract the magnitude of u PId,2 :

[0075]

[0076] Select the magnitude U PId,2 as the fault characteristic quantity to indicate whether the rotor resistance unbalance fault occurs. For a healthy DFIG, it is theoretically 0. Extract the initial phase of u PId,2 :

[0077]

[0078] Define Cm is the current characteristic quantity, where the subscript m can be a, b, c, representing the three phases of ABC respectively, and its value is:

[0079]

[0080] In the formula: when m = a, n = 0; when m = b, n = 1; when m = c, n = -1. I cm and θ cm are respectively the amplitude and initial phase of C m , and are respectively

[0081]

[0082]

[0083] By calculating θ eWRId,2 , and comparing it with θ ca , θ cb and θ cc respectively to judge the faulty phase. If a fault occurs in phase A, then θ ca = θ eWRId,2 , while θ cb and θ cc are not equal to θ eWRId,2 ; if a fault occurs in phase B, then θ cb = θ eWRId,2 , while θ ca and θ cc are not equal to θ eWRId,2 ; if a fault occurs in phase C, then θ cc = θ eWRId,2 , while θ ca and θ cb are not equal to θ eWRId,2 .

[0084] The degree of the fault can be judged by calculating the additional resistance ΔR of the faulty phase:

[0085]

[0086] In the formula: I cm is the amplitude of the characteristic current of the faulty phase.

[0087] Step S5, design an executable fault diagnosis process, and the diagnosis steps are as Figure 4 shown as:

[0088] S51: Fault detection. Use the frequency tracking algorithm to extract the second-order component of u PId online. Once its amplitude U PId,2 is greater than the preset detection threshold U th , it is considered that a rotor resistance imbalance fault has occurred;

[0089] S52: Fault location. After detecting the rotor resistance imbalance fault, calculate e WRId,2 and three characteristic currents C a , C b and C c , and compare their initial phases. The one closest to θ eWRId,2 is the faulty phase;

[0090] S53: Fault degree assessment. After the faulty phase is located, calculate the additional resistance value ΔR of the faulty phase through Ε WRId,2 and the amplitude of the characteristic current of the faulty phase.

[0091] Embodiment

[0092] This method is illustrated by taking the case where the resistance of phase A of the DFIG rotor is unbalanced as an example:

[0093] According to Figure 1 , when the rotor resistance imbalance fault of the DFIG occurs in phase A, the phase voltage equation in the three-phase stationary coordinate system can be expressed as

[0094]

[0095] In the formula: [u s = [u as u bs u cs T is the stator voltage matrix; [u r = [u ar u br u cr T ; [i s = [i as i bs i cs T is the stator current matrix; [i r = [i ar i br i cr T ; [R s = R s dig[1 1 1] T is the stator winding resistance matrix; [R r = dig[R r + R R r R r T ;

[0096]

[0097] ​​​​​L ls and L lr are the leakage inductances of the stator and rotor windings respectively; L ms is the maximum mutual inductance between the stator and rotor;

[0098]

[0099] Using Park transformation, Eq. (1) is transformed into the synchronous rotating coordinate system, and the dq-axis voltage equations of the DFIG can be obtained:

[0100]

[0101] where: u dr and u qr are the dq-axis voltages of the rotor; i dr and i qr are the dq-axis currents of the rotor; R r represents the resistance of the rotor winding, and R is the additional resistance connected in series in the faulty phase to simulate the rotor resistance unbalance fault; σ = 1 - L 2 m / (L s L r ) is the leakage magnetic coefficient of the DFIG; L s = L ls + 3 / 2L ms ; L r = L lr + 3 / 2L ms ; L m = 3 / 2L ms ; L ls is the leakage inductance of the stator winding; L lr is the leakage inductance of the rotor winding; L ms is the maximum mutual inductance between the stator and rotor; ω e = ω s - ω r , ω s is the electrical angular velocity of the stator magnetic field, and ω r is the electrical angular velocity of the rotor; e WRId and e WRIq are the additional terms introduced due to the rotor resistance unbalance fault, and θ e is the electrical angle of the rotor relative to the stator magnetic field.

[0102] With stator magnetic field orientation, the voltage equation on the rotor side of the DFIG can be simplified to

[0103]

[0104] As Figure 2 , a feedforward compensation is introduced in the current loop to achieve decoupling of the dq axes on the rotor side, and the decoupling model of the fault signal is obtained as:

[0105]

[0106] After the dq-axis decoupling, the DFIG current controller with rotor resistance unbalance fault can be simplified, as Figure 3 shown. For a healthy DFIG, e WRId and e WRIq are equal to 0, and the ideal dq-axis currents and voltages are constant values. After the rotor resistance unbalance fault occurs, e WRId and e WRIq will introduce harmonics into the dq-axis currents and voltages. According to Equation (2), e WRId and e WRIq are mainly composed of second-order components and DC components. Therefore, the second-order components will appear in the dq-axis signals, such as the current signals i dr and i qr , and the voltage signals u PId and u PIq .

[0107] Since the rotor resistance unbalance can introduce second-order torque into the DFIG, a perturbation with the same frequency will be induced in the speed signal. Since the input i qr * of the current loop is the output of the speed closed-loop controller, the i qr * signal will also contain the second-order component i qr,2 *, which will be affected by the speed closed-loop controller and the motor mechanical parameters. Therefore, i qr,2 and u PIq,2 will also be disturbed by the speed closed-loop controller and the motor mechanical parameters, etc.

[0108] The DFIG current controller design adopts the i dr * = 0 control, and there is no second-order component in the i dr * signal. Obviously, i dr,2 and u PId,2 only contain the second-order component e WRId in e WRId,2 , and are independent of the motor mechanical parameters. Therefore, using i dr,2 and u PId,2 for rotor resistance unbalance fault diagnosis can eliminate the interference of speed, etc. In order to avoid the influence of the asymmetry problem on the power grid introduced into the stator winding, measures should be taken to suppress i dr,2 while amplifying u PId,2 . Therefore, the voltage signal u PId is selected as the fault diagnosis sampling signal to extract the fault characteristic parameter u PId,2 .

[0109] Considering the influence of the current loop, according to Figure 3 the DFIG fault characteristic parameter i dr,2and u PId,2 The expression of

[0110]

[0111] In the formula: I dr,2 and θ idr,2 are respectively the amplitude and initial phase of i dr,2 , E WRId,2 and θ eWRId,2 are respectively the amplitude and initial phase of e WRId,2 , U PId,2 and θ uPId,2 are respectively the amplitude and initial phase of u PId,2 , and there is

[0112]

[0113] In the formula: K p and K i are respectively the proportional gain and integral gain.

[0114] The amplitude and phase of the fault characteristic parameters satisfy

[0115]

[0116] According to Equation (6), |T id | and |T ud | are greatly affected by the values of K p and K i For a current closed-loop controller with high bandwidth, that is, when K p and K i are relatively large, the amplitude of |T id | is close to 0, while the amplitude of |T ud | is close to 1. According to Equation (7), i dr,2 is well suppressed by the high-bandwidth current closed-loop and only has a relatively small amplitude. And u PId,2 has a greatly increased amplitude in this case. On the contrary, for a current closed-loop controller with low bandwidth, the amplitude of u PId,2 is small, while i dr,2 has a relatively large amplitude.

[0117] Establish the bandwidth design criterion for the current closed-loop bandwidth: K p and K i take relatively large values within the range of excellent control performance to ensure that the amplitude of |T id | is close to 0, while the amplitude of |T ud | is close to 1.

[0118] Adopt a frequency tracking algorithm to extract the second-order component u PId from the u PId,2 signal online:

[0119]

[0120] Extract u PId,2 Amplitude:

[0121]

[0122] Select the amplitude U PId,2 As a fault feature quantity to indicate whether the rotor resistance imbalance fault occurs. For a healthy DFIG, it is theoretically 0. Extract u PId,2 Initial phase:

[0123]

[0124] According to Equation (7), i d,2 and e WRId,2 The amplitudes and phases are:

[0125]

[0126] According to Equation (2), when the rotor resistance imbalance fault occurs in phase A of the rotor, e WRId,2 Can also be expressed as

[0127]

[0128] In the formula, i dr,0 and i qr,0 Are the DC components of i dr and i qr Define C a As the current feature quantity of phase A, and its value is

[0129]

[0130]

[0131]

[0132] By combining Equations (5), (12) and (13), we can obtain

[0133]

[0134] If the rotor resistance imbalance fault occurs in phase B of the rotor, then

[0135]

[0136] Similarly, if the rotor resistance imbalance fault occurs in phase C of the rotor, then

[0137]

[0138] I cb and θ cb are the amplitude and initial phase of C b respectively. I cc and θ cc are the amplitude and initial phase of C c respectively. If a fault occurs in phase A, then θ ca = θ eWRId,2 , while θ cb and θ cc are not equal to θ eWRId,2 . Therefore, by calculating θ eWRId,2 , and comparing it with θ ca , θ cb and θ cc respectively, the fault phase can be determined. If a fault occurs in phase A, the fault severity can be judged by calculating the additional resistance ΔR of the fault phase:

[0139]

[0140] Figure 4 is the flow chart for diagnosing the unbalanced rotor resistance fault of DFIG. The on-line diagnosis steps for the unbalanced rotor resistance fault are as follows:

[0141] Step 1: Fault detection. Use the frequency tracking algorithm to extract the second-order component of u PId online. Once its amplitude U PId,2 is greater than the preset detection threshold U th , it is considered that the unbalanced rotor resistance fault has occurred.

[0142] Step 2: Fault location. After detecting the unbalanced rotor resistance fault, calculate e WRId,2 and the three characteristic currents C a , C b and C c , and compare their initial phases. The one closest to θ eWRId,2 is the fault phase.

[0143] Step 3: Fault severity assessment. After the fault phase is located, calculate the additional resistance value ΔR of the fault phase through Ε WRId,2 and the amplitude of the characteristic current of the fault phase.

[0144] A DFIG rotor resistance unbalance fault diagnosis system is used to implement the above DFIG rotor resistance unbalance fault diagnosis method.

Claims

1. A method for diagnosing the unbalanced fault of the DFIG rotor resistance, characterized in that, Including the following steps: Establish a decoupling model for the dq-axis fault signals of the DFIG rotor winding under rotor resistance imbalance faults, specifically including: Using the Park transformation, the dq-axis voltage equations of the DFIG rotor winding are obtained as: where: u dr , u qr are the dq-axis voltages of the rotor; i dr , i qr are the dq-axis currents of the rotor; R r represents the resistance of the rotor winding, and R is the additional resistance connected in series with the faulty phase to simulate the rotor resistance unbalance fault; σ = 1 - L 2 m / (L s L r ) is the leakage magnetic coefficient of the DFIG; L s = L ls + 3 / 2L ms ; L r = L lr + 3 / 2L ms ; L m = 3 / 2L ms ; L ls is the leakage inductance of the stator winding; L lr is the leakage inductance of the rotor winding; L ms is the maximum mutual inductance between the stator and the rotor; ω e = ω s - ω r , ω s is the electrical angular velocity of the stator magnetic field, and ω r is the electrical angular velocity of the rotor; e WRId and e WRIq are the additional terms introduced due to the rotor resistance unbalance fault, and θ e is the electrical angle of the rotor relative to the stator magnetic field; Adopting stator magnetic field orientation and introducing feedforward compensation to achieve dq-axis decoupling on the rotor side, and the decoupling model for fault signals is obtained as: Select the fault diagnosis sampling signal, and determine the fault characteristic parameters. The fault diagnosis sampling signal is the voltage signal u PId , the voltage signal u PId As the fault diagnosis sampling signal, extract the second-order component u PId online from the u PId，2 ; Establish a method for designing the current closed-loop bandwidth, specifically: Construct a fault feature parameter model: where: I dr,2 and θ idr,2 are the amplitude and initial phase of i dr,2 respectively, E WRId,2 and θ eWRId,2 are the amplitude and initial phase of e WRId,2 , respectively, U PId,2 and θ uPId,2 are the amplitude and initial phase of u PId,2 respectively, and there is Where: K p and K i are the proportional gain and the integral gain, respectively; Design a method for diagnosing rotor resistance imbalance faults, specifically: From u PId Extract the second-order component u online from the u PId,2 : Extract u PId,2 Amplitude: Extract u PId,2 Initial phase: Define C m as the current characteristic quantity, and its value is: where: when m = a, n = 0; when m = b, n = 1; when m = c, n = -1; I cm and θ cm are the amplitude and initial phase of C m respectively, and are By calculating θ eWRId,2 , and comparing it with θ ca , θ cb and θ cc respectively to determine the faulty phase; The degree of the fault is judged by calculating the additional resistance ΔR of the faulty phase: Where: I cm is the characteristic current amplitude of the faulty phase; Design a fault diagnosis process, and the specific diagnosis process includes: Fault detection: Use the frequency tracking algorithm to extract the second-order component of u online. If the amplitude U PId is greater than the preset detection threshold U PId,2 , a rotor resistance imbalance fault occurs. th ​ Fault location: After detecting the rotor resistance imbalance fault, calculate e WRId,2 and the characteristic current C a , C b and C c , and compare their initial phases, where the one closest to θ eWRId,2 is the faulty phase; Fault degree assessment: After the fault phase location is completed, the additional resistance value ΔR of the fault phase is calculated through WRId,2 and the amplitude of the characteristic current of the fault phase.

2. The method for diagnosing the DFIG rotor resistance imbalance fault according to claim 1, characterized in that, The amplitude and phase of the fault characteristic parameter u PId,2 satisfy 3. A DFIG rotor resistance imbalance fault diagnosis method according to claim 1, characterized in that, The second-order component u PId is extracted online from the u PId,2 signal using a frequency tracking algorithm.

4. A DFIG rotor resistance imbalance fault diagnosis system, characterized in that, This system is used to implement the method for diagnosing DFIG rotor resistance imbalance faults described in any one of claims 1 to 3.

Images