Multi-dimensional electrical fault diagnosis method for dual three-phase motor system based on magnetic motive force equilibrium trajectory

Abstract

This invention relates to a multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory, belonging to the field of electrical fault diagnosis. The method includes: preliminary judgment of the phase with an open-circuit fault based on image similarity theory, identifying whether the open-circuit fault occurs in phases ABC or DEF; if the fault occurs in phases ABC, calculating the theoretical real-time MMF balance value after the fault occurs in phases ABC respectively, and using SSIM to compare the actual balance of phases ABC with the theoretical MMF balance to locate the faulty bridge arm; the same applies to the DEF phase bridge arm; determining that the fault is located at the switching transistor S. j1 / S j2 Or S j3 / S j4 When in the high modulation region, S is distinguished based on the average current value. j1 With S j2 Fault, or S j3 With S j4 Fault; When in the low modulation region, a three-to-two level hybrid modulation scheme is used to distinguish S. j1 With S j2 Fault, or S j3 With S j4 Fault.

Description

Technical Field

[0001] This invention belongs to the field of electrical fault diagnosis technology for dual- and three-phase motors, and relates to a multi-dimensional electrical fault diagnosis method for dual- and three-phase motor systems based on magnetomotive force balance trajectory. Background Technology

[0002] Dual three-phase permanent magnet synchronous electric drive systems driven by three-level voltage source inverters are widely used in high-power-density, high-reliability applications such as high-speed rail, electric vehicles, and ships due to their high efficiency and strong fault tolerance. However, long-term operation in complex and harsh environments can accelerate the aging of the drivers and motors due to factors such as overvoltage, overcurrent, corrosion, and dust. Simultaneously, load disturbances, ambient temperature, electromagnetic interference, and mechanical vibration can cause solder joint detachment, ultimately leading to converter failure. Although the high degree of control freedom of dual three-phase systems provides a certain degree of passive fault tolerance or non-diagnostic fault tolerance, these methods still require determining whether a fault has occurred. Furthermore, due to the lack of specific fault information, the output voltage vector deviates from the command vector, limiting control accuracy. Obtaining appropriate fault information sources can trigger remedial fault-tolerant control strategies to maintain system functionality as much as possible while preventing fault propagation and system damage. Therefore, fault detection functionality is crucial.

[0003] Among all electrical faults in a drive system, open-circuit faults are the most common. Fault diagnosis strategies mainly include: analytical model-based strategies, but their diagnostic effectiveness is highly dependent on model accuracy and observer parameters, and they carry the risk of misdiagnosis under dynamic conditions such as sudden load / torque changes, making them difficult to adapt to the complex and variable operating conditions of dual-phase and three-phase systems. Furthermore, the observer requires additional parameter tuning and time convergence, increasing the difficulty of algorithm application. Data-driven fault diagnosis strategies, however, still require high computational power, and the diagnostic delay typically exceeds one fundamental frequency cycle, making them slower than dedicated algorithms optimized for specific topologies and unable to meet the rapid diagnostic needs of dual-phase and three-phase systems under complex operating conditions. Signal processing-based fault diagnosis strategies, however, have potential fault risks in auxiliary circuits, increase system cost and complexity, and their scalability is inferior to pure software methods. Additionally, eigenvalues ​​based on phase current slope and zero-crossing time can also be used for rapid diagnosis of SOPF in dual-phase and three-phase motors. However, these eigenvalues ​​are similar to normal signals during dynamic response, resulting in insufficient algorithm robustness.

[0004] Therefore, traditional fault diagnosis methods rely on the timing characteristics of current / voltage in a single fault phase, which results in low information density, difficulty in distinguishing between normal dynamics and fault states, and open circuit faults in the upper and lower bridge arms and the midpoint bridge arm. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a multi-dimensional electrical fault diagnosis method for dual- and three-phase motor systems based on the magnetomotive force (MMF) balance deviation trajectory. The fault diagnosis strategy based on the similarity comparison of multi-source MMF balance deviation trajectory graphics can not only accurately locate the faulty three phases, but also effectively distinguish between sensor faults and open-circuit faults through the extreme differences in MMF imbalance, and precisely locate the three-phase subsystem where the sensor fault is located. This method can effectively solve the problem of fault feature aliasing and fault differentiation under complex topologies, and significantly improve the accuracy and robustness of multi-dimensional fault identification.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory includes the following steps: S1: Based on image similarity theory, make a preliminary judgment on the phase with open circuit fault, and identify whether the single-open-switch-fault (SOSF) occurs in the ABC phase bridge arm or the DEF phase bridge arm. S2: If a single switch open-circuit fault occurs in the ABC phase bridge arm, the actual fault characteristic curve of the ABC phase bridge arm is sampled, and then the structural similarity index SSIM is used to compare the similarity between the actual fault characteristic curve of the ABC phase and the reference fault characteristic curve, thereby locating the faulty bridge arm; if a single switch open-circuit fault occurs in the DEF phase bridge arm, the actual fault characteristic curve of the DEF phase bridge arm is sampled, and then the structural similarity index SSIM is used to compare the similarity between the actual fault characteristic curve of the DEF phase and the reference fault characteristic curve, thereby locating the faulty bridge arm. S3: Based on the time of the fault occurrence, it is preliminarily determined that the fault is located in the switching transistor S. j1 / S j2 Or S j3 / S j4 Where j represents the j-th phase, j =A~F, where 1~4 represent the serial numbers of the switching transistors in the j-th phase; S4: When in the high modulation region, S is distinguished based on the average current value. j1 With S j2 Fault, or S j3 With S j4 Fault; When in the low modulation region, using a three-to-two level hybrid modulation scheme, if the system recovers to normal, the fault type is switch S. j2 or S j4 A fault occurs if the fault is not detected, and vice versa if the fault is detected by the switching transistor S. j1 or S j3 The fault was ultimately located by identifying an open circuit fault in a single switch.

[0007] Furthermore, in step S1, the initial judgment of the phase with open-circuit fault based on image similarity theory is made to identify whether the single-switch open-circuit fault occurs in the ABC phase bridge arm or the DEF phase bridge arm, based on the following: When a single-switch open-circuit fault occurs in phases ABC, the reference value of the composite fundamental MMF of phases ABC during the fault half-cycle is:

[0008] When a single switch open-circuit fault occurs in phases ABC, the combined MMF reference value for phases DEF during the fault half-cycle is:

[0009] When a single-switch open-circuit fault occurs in phase DEF, the reference value of the fundamental MMF of phases ABC during the fault half-cycle is:

[0010] When a single-switch open-circuit fault occurs in a phase of the DEF phase, the reference value of the combined MMF of the DEF phase during the fault half-cycle is:

[0011] in The MMF coefficient is determined by the motor topology and is a constant; j =A~F are phases, The spatial angle is measured in electrical degrees. As the initial phase angle variable, when j When =A~F, φ They are 0, 2π / 3, 4π / 3, π / 6, 5π / 6, and 3π / 2, respectively; The current of the faulty phase. The number of turns in the motor stator winding. This refers to the current amplitude. ω Represents electric angular velocity. t For time; when the switching transistor fault is located in phases A through F respectively, θ 0 is equal to 0, 2 / 3π, -2 / 3π, -1 / 3π, 1 / 3π and -π respectively.

[0012] Furthermore, in step S1, the MMF balance of the two windings is used as a characteristic value for fault judgment, specifically including: The following explanation uses an SOSF (SO2-SO2) fault in the A-phase power switch as an example. θs = ωt = θe ,in θe It is the electrical angle of the rotor position; and Simplified to onlyωt An expression with a single independent variable:

[0013] Wherein, the subscript OCF-a indicates the result when a phase a experiences a SOSF fault, and the superscript... θ s = ωt Indicates ωt Replace the original expression θ s Similarly, simplify the expression for faults in other phases; Establish fault determination feature quantities F η1 and F η2 To determine whether a fault has occurred and to locate the three-phase winding where the fault is located;

[0014] Under normal circumstances F η1 and F η2 The ideal value is 1; when a SOSF switch failure occurs in the ABC phase bridge arm... F η1 Less than 1; when the DEF phase bridge arm switch SOSF fails, F η2 Less than 1.

[0015] Furthermore, in step S2, the actual fault characteristic value curve is determined based on the structural similarity index SSIM. f η1 and f η2 and F η1 and F η2 The degree of similarity between the resulting reference fault characteristic value curves is used to locate the faulty bridge arm, specifically including: Once the fault has been located in the ABC three-phase bridge arm, subsequent judgments will use... F η1 By setting a sliding window W Similarity determination is performed, window length W L As the fundamental frequency of the phase current changes, the time scale of the sampling window remains 1 / 8 of the fundamental frequency period, satisfying the following relationship:

[0016] In the formula, f f It is the fundamental frequency. f sThis refers to the sampling frequency; the data in the sampling window is updated in real time with the control, and the stored data is:

[0017] In the formula, The actual current data stored in the sampling window. The curve representing the actual fault characteristic value of the i-th sample. The theoretical values ​​of each phase are stored in the sampling window. For the first i Secondary sampling j The reference fault curve for the phase; Average of actual current data Theoretical values ​​of each phase (mean values) Variance of actual current data Variance of theoretical values ​​for each phase and covariance The calculation is as follows:

[0018]

[0019]

[0020] Similarity SM j Using the intermediate variable contrast measurement function C jr and structural measurement function S jr The calculation is as follows:

[0021]

[0022] in These represent the similarity scores of A, B, and C, respectively. This indicates the candidate phase for faults in the current similarity comparison; Based on the physical meaning of similarity, the maximum SM j corresponding j The value indicates the phase arm of the bridge where the fault is located, and the fault alarm flag is in position. F flag Set to 1, after an alarm fault occurs, SM j The judgment result is output after a delay of 0.05 fundamental frequency cycles; the maximum SM within 0.05 cycles is statistically analyzed. j The phase that appears most frequently in each phase is selected as the final faulty phase. The determination logic is as follows:

[0023]

[0024] in This indicates the number of times candidate phase j is determined to be a faulty phase within the statistical period L, where L represents the data length used for statistical and delay output. This represents the maximum value among the statistical counts. This indicates the fault type code.

[0025] Furthermore, in step S3, it is determined whether the fault belongs to S at this time based on the time when the fault occurred. j1 / S j2 Or S j3 / S j4 Types, specifically including: If S j1 / S j2 Fault, characteristic value of actual current calculation f η1 Appeared 0~π , F type Determined to be 1; otherwise, S j3 / S j4 In case of type fault, π~2π Within the range f η1 Reaching the threshold T h1 , F type The value is determined to be 4.

[0026] Furthermore, in step S4, when in the high modulation region, for S j1 / S j2 Type of fault, S j1 A fault will cause the current to approach zero, S j2 During a fault, the current is between zero and its normal value. Utilizing this characteristic, a length of [length missing] is used... W L The average current Δ of the sliding window calculation i j Thus distinguishing S j1 and S j2 Fault:

[0027] in This represents the instantaneous current value of faulty phase j at the i-th sampling time; In the early stages of sliding window generation, it was not fully achieved. W L The length is calculated based on the actual sliding window length; a threshold is set. T h2 =0.1 i s Distinguish between the two types of faults, is This is the stator current reference value:

[0028] Under normal operating conditions, the fault switch indicator F sw Keep the value zero; when S j1 and S j2 When open circuit faults occur separately, F sw The corresponding values ​​are 1 and 2; The same method is used to distinguish S in the high modulation region j3 With S j4 Fault.

[0029] Furthermore, in step S4, when in the low modulation region, a three-to-two-level hybrid modulation scheme is used for differentiation, that is, the faulty bridge arm uses two-level modulation, while the normal other bridge arms maintain three-level modulation; the switching function of the three-level system simulating two-level modulation is as follows: The switch state is P. S j1 =1, S j2 S j3 S j4 The value is 0, and the voltage is [missing information]. U dc / 2, where U dc This refers to the DC bus voltage of the inverter. The switch state is N. S j4 S is 1. j1 , S j2 S j3 The value is 0, and the electrode voltage is - U dc / 2; Simulating two-level operation using a three-level real system S j1 and S j4 Driven by a set of complementary PWM, and S j2 and S j3 Keep it closed; if S j2 or S j3 In case of a fault, the system will recover to normal operation in two-level mode. (Characteristic value) f η1 and f η2It will also recover to around 1, combined with the already determined S j1 / S j2 Type of fault: ① If f η1 and f η2 If the value recovers to around 1, then the fault switch transistor is S. j2 ; ② Conversely, if f η1 and f η2 If the fault is not resolved, then the faulty switch is S. j1 ; This method effectively distinguishes S in the low-modulation region. j1 With S j2 Fault; use the same method to distinguish S in the low modulation region j3 With S j4 Fault.

[0030] The beneficial effects of this invention are as follows: (1) Although this scheme is slightly slower than the hardware scheme in absolute speed, it can trigger an alarm within several control cycles and within 0.15 hours. T f Internal fault location offers significant advantages over non-hardware solutions.

[0031] (2) The proposed solution did not cause false alarms in steady-state / dynamic conditions (including current distortion caused by neutral point imbalance), and the fault can still be accurately located in the dynamic process, showing better robustness.

[0032] (3) The proposed method is based on the MMF balance. After a switch failure occurs, other multi-level multi-N phase systems will also have an MMF balance shift. Therefore, it can be applied and the fault can be reduced to the corresponding phase.

[0033] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0034] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein: Figure 1 The theoretical composite MMF trajectory generated by two sets of three-phase windings under normal and switching transistor fault conditions (×) N Im ); Figure 2 The image shows the fault feature values, where (a) is... F η1 Image (b) is F η2 Images; Figure 3 The flowchart shows a multidimensional electrical fault diagnosis method for dual- and three-phase motor systems based on magnetomotive force balance trajectory. Figure 4 The results of the diagnostic test for SOSF occurring under rated operating conditions are shown; where (a) is the result under rated operating conditions. S A1 The phase current waveform during an open-circuit fault, (b) is under rated operating conditions. S A1 The theoretical and actual fault characteristic values ​​of open circuit faults, (c) is the rated operating condition. S A1 The magnetomotive force trajectory of an open-circuit fault, (d) is the magnetomotive force trajectory under rated operating conditions. S A1 The average current of the fault phase in the event of an open-circuit fault, (e) is the current under rated operating conditions. S A2 The phase current waveform during an open-circuit fault, (f) is the phase current waveform under rated operating conditions. S A2 The theoretical and actual fault characteristic values ​​of open circuit faults, (g) is under rated operating conditions. S A2 The magnetomotive force trajectory of an open-circuit fault, (h) is under rated operating conditions. S A2 Average current of the faulty phase in which an open-circuit fault occurs. Detailed Implementation

[0035] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0036] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0037] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0038] Example 1: This invention provides a multidimensional electrical fault diagnosis method for dual three-phase motor systems based on magnetomotive force balance trajectory.

[0039] A dual three-phase motor is a high-order nonlinear system, and the phase current needs to be converted to a spatially decoupled coordinate system (VSD). αβ, xy and o 1 o 2. Decoupling control is performed on three orthogonal subplanes. Among them, αβ The subplane is the mapping subplane of the fundamental current, which participates in electromechanical energy conversion. xy The subplane is a harmonic subplane, which only generates copper loss. o 1 o Subplane 2 is an uncontrollable, generalized zero-sequence subplane. Thanks to the motor's dual-neutral-point topology, the current in this subplane is always zero. Typically, to minimize harmonic copper losses, the harmonic axis current... I xy The reference value is zero. The fundamental current... I αβ ， It reflects the size and orientation of the stator MMF, its αβ The reference values ​​of the axes are orthogonally distributed with equal amplitude, thus generating a spatial rotation MMF with constant amplitude, the expression of which is as follows: (1) In the formula, They are respectively shaft current, I m The current amplitude, ω It represents electric angular velocity.

[0040] Substitute the current reference values ​​of the three orthogonal subplanes into the inverse VSD coordinate transformation matrix in equation (2). T 6s2s This allows us to obtain the reference phase current under normal conditions.I j The expression is shown in equation (3).

[0041] (2) (3) In the formula, φ It is the initial phase angle variable, when j When =A~F, φ The values ​​are 0, 2π / 3, 4π / 3, π / 6, 5π / 6, and 3π / 2, respectively.

[0042] When the current in equation (3) flows through a coil, each turn of the coil generates a ring-shaped magnetic field. The magnetic fields of multiple turns of the coil are superimposed to form the total MMF. Taking the axis of phase a winding as the origin of the spatial coordinate system, and rotating counterclockwise (the spatial angle) θ s ( θ s It is the positive direction (measured by an electric angle meter). At a certain instant... t Distance from the axis of phase a winding θ s At each location, the fundamental reference MMF for each phase is as follows: (4) in, N The MMF coefficient is expressed as follows: (5) In the formula, N This indicates the total number of turns in series for each phase. k wl The winding factor of the MMF fundamental frequency. P n It is the number of pole pairs of the motor. N It is a constant value determined solely by the motor topology.

[0043] The ABC three-phase composite MMF is a spatially rotating wave with constant amplitude. It is distributed sinusoidally in space and has an angular frequency... ω Rotate counterclockwise. ABC three-phase combined MMF reference value. F The magnitude of 1 always appears at the winding axis position where the phase current reaches its peak. Similarly, the combined MMF reference value for the three phases DEF can be calculated. F 2. The MMF of the two sets of three-phase windings are always equal.

[0044] Taking a SOSF (Suspension of Flow Rate) event in the IGBT of phase A as an example, the fault phase current will approach 0 in the half-cycle of the fault. According to the matrix... T 6s2sThe coordinate transformation relationship in the first row represents the actual phase current during the fault half-cycle. i A With reference current I′ A Satisfy constraint (6): (6) In the formula, I′ α , I′ x and I′ o1 They represent α、x and o Reference current for axis 1.

[0045] To ensure that the phase current has no DC bias, the harmonic current reference value after the fault is... I ′ xy It can be used αβ Shaft reference current I′ αβ This is indicated. Simultaneously, it is combined with the zero-sequence current. i o1 , i o2 The condition for it to always be 0 and the constraints in equation (6), I ′ xy It can be written in the following general form: (7) In the formula, k i ( i =1~4) are the coefficients to be determined.

[0046] Fundamental current after fault I′ αβ Still determined by torque and flux linkage, therefore the same as under normal conditions: (8) Typically, before the fault is diagnosed, xy The reference value for the shaft current remains 0. Although this will lead to a decrease in control degrees of freedom... xy Shaft current control and αβ The shaft current control is conflicted, and the MMF after the fault will be indirectly affected by the mutual inductance, but the motor's output torque is controlled by... αβ Determined by shaft current, therefore αβ Axis current tracking is prioritized by closed-loop control, and the output of the harmonic axis current loop is typically limited. Therefore, considering these factors, xy The axis control is weaker, and the system still operates close to the typical minimum copper loss mode after a fault, that is: (9) The simultaneous equations (6)~(9) and the inverse transformation matrix T 6s2s The reference current expression for the fault half-wave period after an SOSF occurs in the IGBT of phase A bridge arm can be obtained as follows: (10) Based on the reference current expression after the fault in equation (10), the reference value of the composite fundamental MMF of phases ABC during the fault half-cycle after a SOSF fault occurs in phase A can be calculated. for (11) Wherein, the subscript OCF-A indicates the result when a SOSF fault occurs in phase A. Similarly, the combined MMF reference value of phases DEF after a SOSF fault occurs in phase A can be obtained. for (12) Similarly, the MMF reference value expression for the IGBTs of the other bridge arms after SOSF occurs during the fault half-wave period can be obtained. The derivation results are summarized in Table 1.

[0047] Table 1

[0048] In the table, , When the switching transistor faults are located in phases A through F, θ 0 is equal to 0, 2 / 3π, -2 / 3π, -1 / 3π, 1 / 3π and -π respectively.

[0049] The table above shows the synthesized MMF of the open-circuit fault system, which provides a theoretical basis for various subsequent fault identification methods.

[0050] Firstly, an open-circuit fault phase identification method based on image similarity theory is used. After a fault, the MMF (Multi-Functional Flow Model) of the two sets of three-phase windings no longer maintains the constant amplitude rotating circular characteristic. Instead, the instantaneous amplitude of the composite MMF of the normal three-phase windings will change with phase. ωt and θ s The changes form a set of MMF-phase sequences. Based on this, the MMF balance of the two sets of windings can be used as a characteristic value for fault alarm.

[0051] Taking the SOSF generated by the A-phase power switch as an example, to facilitate the study of the amplitude characteristics of the MMF, we take... θs = ωt = θe ( θe It is the rotor position electrical angle. and Simplified to only ωt (Right nowθ e An expression with a single independent variable: (13) Wherein, the subscript OCF-a indicates the result when a SOSF fault occurs in phase a, and the superscript θ s = ωt Indicated by ωt Replace the original expression θ s Similarly, the expressions for other phase faults can also be simplified. Figure 1 It shows the polar coordinate system and Follow θ e The trajectory of change.

[0052] in accordance with Figure 1 Based on the characteristics in the data, establish fault determination feature quantities. F η1 and F η2 This allows for quick determination of whether a fault has occurred and to locate the three-phase winding where the fault is located.

[0053] (14) Under normal circumstances F η1 and F η2 The ideal value is 1. When the switching transistor SOSF occurs in the ABC phase bridge arm, F η1 It will be less than 1. When the DEF phase bridge arm switch transistor is SOSF, F η2 It will be less than 1.

[0054] Figure 2 (a) and (b) show F η1 and F η2 The theoretical curve. When the fault is located in phases ABC, F η1 The image exhibits a sinusoidal change, with values ​​ranging from 0 to 1, and the amplitude changes relatively slowly; however, when the fault is located in the DEF phase, F η1 The amplitude changes dramatically. F η2 The pattern is similar. These changes manifest as three sets of periodic sinusoidal curves with equal phase differences. Borrowing from the Structural Similarity Index (SSIM) used in image processing to compare the similarity of two images, the actual fault characteristic value curve can be determined. fη1 and f η2 Compared with the reference fault curve F η1 and F η2 The degree of similarity is used to locate the faulty bridge arm.

[0055] Still with S A1 Taking SOSF as an example, the fault has been located in the ABC three-phase bridge arm, and the subsequent judgment will be made by using F η1 By setting a sliding window W Similarity determination is performed, window length W L As the fundamental frequency of the phase current changes, the time scale of the sampling window remains 1 / 8 of the fundamental frequency period, satisfying the following relationship: (15) In the formula, f f It is the fundamental frequency. f s It is the sampling frequency. The data W of the sampling window. d The data is updated in real time with the control system and is stored as follows: (16) In the formula, the subscript is marked with " r The data shown is the actual current, indicated by the subscript "". j The numbers “” represent the theoretical values ​​of each phase, and the subscripts in the following formulas have the same meaning.

[0056] Other parameters, mean μ ,variance σ 2 Covariance σ jr The calculation is as follows (17) (18) (19) Similarity SM j The intermediate variable contrast measurement function needs to be used. C jr and structural measurement function S jr The calculation is as follows: (20) (twenty one) Based on the physical meaning of similarity, the maximum SM j correspondingj The value indicates the faulty bridge arm / phase. To ensure robustness of the determination, in F flag Set to 1, after an alarm fault occurs, SM j The judgment result needs to be output after a delay of 0.05 fundamental frequency cycles. This is achieved by statistically analyzing the maximum SM within 0.05 cycles. j The phase that appears most frequently in each phase is selected as the final faulty phase. The determination logic is as follows: (twenty two) (twenty three) SM determined j After reaching the maximum value, the bridge arm containing the faulty switch can be located. Based on the time of the fault occurrence, it can be determined whether it belongs to S. j1 / S j2 Or S j3 / S j4 Type. If S A1 / S A2 Fault, characteristic value of actual current calculation f η1 It will inevitably appear 0~π , F type It can be determined to be 1. Conversely, S A3 / S A4 In case of type fault, π~2π Within the range f η1 It will reach the threshold T h1 , F type It can be determined to be 4.

[0057] Once the fault has been located to a specific type, the faulty switching transistor can be further identified using a hybrid modulation-based method to enhance the open-circuit fault characteristics of different bridge arms. (Using S...) j1 / S j2 Taking a type of fault as an example, in the high modulation region, S j1 A fault will cause the current to approach zero, while S j2 During a fault, the current is between zero and its normal value. Utilizing this characteristic, a length of [length missing] is still used. W L The average current Δ of the sliding window calculation i j Thus distinguishing S j1 and S j2 Fault: (twenty four) In the early stages of sliding window generation, it was not fully achieved. W LThe length is calculated based on the actual sliding window length. A threshold is set. T h2 =0.1 i s ( i s (This is the stator current reference value) to distinguish between two types of faults: (25) The above formula shows that, under normal operating conditions, the fault switch indicator... F sw Keep the value zero; when S j1 and S j2 When open circuit faults occur separately, F sw The corresponding values ​​are 1 and 2.

[0058] In the low modulation region, S j1 / S j2 Fault types are difficult to distinguish. Therefore, a three-to-two-level hybrid modulation scheme is adopted: the faulty bridge arm uses two-level modulation, while the normal bridge arms maintain three-level modulation. The switching function of the three-level system simulating two-level modulation is shown in Table 2: Table 2

[0059] When using a three-level real system to simulate two-level operation S j1 and S j4 Driven by a set of complementary PWM, and S j2 and S j3 Keep it closed. Therefore, if S j2 or S j3 In case of a fault, the system will recover to normal operation in two-level mode. (Characteristic value) f η1 and f η2 It will also recover to around 1. Combined with the already determined S... j1 / S j2 Type of fault: ① If f η1 and f η2 If the value recovers to around 1, then the fault switch transistor is S. j2 ② Conversely, if f η1 and fη2 If the fault is not resolved, then the faulty switch is S. j1 .

[0060] To ensure robustness of the judgment, for f η1 and f η2 The determination time is 0.1 fundamental frequency cycles. In this way, S in the low modulation region can be effectively distinguished. j1 / S j2 Fault.

[0061] The dual three-phase electric drive system has good passive fault tolerance. During fault detection, open-loop control of the harmonic shaft ensures that even in the most severe zero-output condition, the motor drive system will not collapse. The instantaneous expression of the phase current after the A-phase current sensor experiences three faults—disconnection, DC bias, and gain error—is as follows: (26) In the formula, the current marked with an abbreviation "'" is the current collected by the sensor after the fault, while the current without "'" is the actual phase current.

[0062] Analysis shows that the B-phase current remains the true sampled value, while a fault in the A-phase sensor will cause the reconstructed current to deviate from the actual value. In a dual three-phase system, the torque mitigation current (MMF) is generated jointly by the ABC and DEF windings. When a sensor fails, the closed-loop control system will still attempt to maintain the torque conservation corresponding to the sampled current. Although the faulty phase current has an error, unlike an actual open-circuit fault, its corresponding winding continues to provide MMF, and there is no period when the MMF is zero. Since there is no electrical fault in the actual circuit topology, although the MMF becomes unbalanced due to the control action, its unbalance will not drop to zero. This characteristic can serve as a key basis for distinguishing between current sensor faults and open-circuit faults, and the MMF imbalance can also serve as an alarm signal for sensor faults.

[0063] After a fault alarm is detected, the discontinuous change in the instantaneous current during the fault (represented by the current derivative) can be used to locate the fault. This method, combining MMF unbalance alarm and current derivative detection, can effectively avoid misdiagnosis caused by sudden current changes during the dynamic response of the system.

[0064] The specific process of the proposed fault diagnosis strategy is as follows: Figure 3 As shown. During normal system operation, only step (a) is executed, constructing MMF information through real-time phase current and electrical angle, and based on the MMF unbalance (F... η1 and F η2 The comparison result between F and the threshold Th1 triggers a fault alarm. η1 or F η2If the fault exceeds Th1, it indicates that the fault occurred in phases A, B, C, or D, respectively. The subsequent diagnostic process for phases D and D is the same as that for phases A, B, and C, and is not shown again in the figure. For current sensor faults, only step (a) needs to be executed, and it is distinguished from the switching transistor fault by the unbalance amplitude. After detecting a phase A, B, or C fault, steps (b) and (c) are executed sequentially. In step (b), the theoretical real-time MMF balance value after a fault occurs in phases A, B, and C is first calculated. F η1 | OCF-a , F η1 | OCF-b and F η1 | OCF-c (See Equation 13), and then the SSIM algorithm is used to compare the actual balance. F η1 The balance with the theoretical MMF (see Equations 15-22) is used to accurately locate the faulty bridge arm and preliminarily determine the fault type as S. j1 / S j2 or S j3 / S j4 Type. Finally, in step (c), distinguish S j1 With S j2 Fault process and differentiation S j3 With S j4 The fault process is the same; the only difference in the diagram is the S. j1 With S j2 The following example demonstrates how to locate a fault. Specifically, fault location needs to be discussed in different categories based on the modulation index: In the high modulation region, the S value can be quickly identified directly based on the average current. j1 With S j2 Fault. When in the low modulation region, using a three-to-two level hybrid modulation scheme, if the system recovers to normal, the fault type is S. j2 Fault, or S if not. j1 The fault ultimately allows for precise location of the SOSF (Synchronous Phase Loss Fault). A phase loss fault can be equivalent to the switching transistor group S... j1 / S j4 At the same time, special circumstances of the fault should be determined.

[0065] To verify the effect, the dual three-phase electric drive system was subjected to rated operation. S A1 and S A2 Open circuit fault, such as Figure 4 As shown in (a)-(g), before the fault occurred, the six-phase currents were symmetrical, and the MMF trajectories generated by phases ABC and DEF overlapped and were circular (as shown in Figure 1). Figure 4 (As shown in (c) and (g)). At this time, the normal state is f η1Approaching 1, fault flag bit F flag The normal indication remains that the diagnostic algorithm has not yet been activated.

[0066] S A1 After the open circuit fault occurs, the currents of phases ABC and DEF are distorted during the positive half-cycle of phase A current. The current of phase A drops to near zero in a short time, and the MMF trajectory no longer maintains a circle. The MMF trajectory of phases ABC collapses, and the radius of the MMF trajectory of phase DEF expands. The MMF of the two windings becomes significantly unbalanced.

[0067] 0.4ms after the fault occurs, the MMF balance characteristic value f η1 Below the threshold T h1 (0.8), F flag A fault indication is issued when the setting is 1, and the fault is determined to be located in phases A, B, and C. Because the system is in the high modulation region, the faulty switch can be directly distinguished by the average current value; the fault similarity determination and the average current value calculation algorithm are activated synchronously. After a delay of 0.05 fundamental frequency cycles, SM... A SM B SM C By combining phase information, the fault range is narrowed down to S. A1 / S A2 type, F type Set to 1; and because the absolute average value of phase A current Δ i A < T h2 =0.1 i s Fault switch indicator F sw Setting it to 1, the faulty switch transistor was ultimately located precisely as S. A1 The fault detection took 0.4ms, the fault location took 1.2ms, and the total diagnosis took 1.6ms, which is only 1 / 15 of a fundamental frequency cycle.

[0068] Example 2: An electronic device, comprising a memory and a processor; The memory is used to store computer programs; The processor is configured to implement the method described in Embodiment 1 when executing the computer program.

[0069] Example 3: A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in Embodiment 1.

[0070] Example 4: A computer program product includes a computer program that, when executed by a processor, implements the method described in Example 1.

[0071] In the above embodiments, the reference to "this embodiment" in the specification indicates that a specific feature, structure, or characteristic described in connection with the embodiment is included in at least some embodiments, but not necessarily all embodiments. Multiple appearances of "this embodiment" do not necessarily refer to the same embodiment.

[0072] In the above embodiments, although the invention has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory structures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed. The embodiments of the invention are intended to cover all such substitutions, modifications, and variations falling within the broad scope of the appended claims.

[0073] As will be understood by those skilled in the art, the computer-readable storage medium described in this embodiment allows for the implementation of all or part of the steps in the above method embodiments by computer program-related hardware. The aforementioned computer program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0074] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver, and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication between them. The memory is used to store computer programs, the communication interface is used to perform communication, and the processor and the transceiver are used to run the computer programs, so that the electronic terminal performs the steps of the above method.

[0075] In this embodiment, the memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0076] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0077] This invention can be used in a wide range of general-purpose or special-purpose computing system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0078] This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0079] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory, characterized in that: Includes the following steps: S1: Based on image similarity theory, make a preliminary judgment on the phase with open circuit fault, and identify whether the single switch open circuit fault occurs in the ABC phase bridge arm or the DEF phase bridge arm. S2: If a single switch open-circuit fault occurs in the ABC phase bridge arm, the actual fault characteristic curve of the ABC phase bridge arm is sampled, and then the structural similarity index SSIM is used to compare the similarity between the actual fault characteristic curve of the ABC phase and the reference fault characteristic curve, thereby locating the faulty bridge arm; if a single switch open-circuit fault occurs in the DEF phase bridge arm, the actual fault characteristic curve of the DEF phase bridge arm is sampled, and then the structural similarity index SSIM is used to compare the similarity between the actual fault characteristic curve of the DEF phase and the reference fault characteristic curve, thereby locating the faulty bridge arm. S3: Based on the time of the fault occurrence, it is preliminarily determined that the fault is located in the switching transistor S. j1 / S j2 Or S j3 / S j4 Where j represents the j-th phase, j =A~F, where 1~4 represent the serial numbers of the switching transistors in the j-th phase; S4: When in the high modulation region, S is distinguished based on the average current value. j1 With S j2 Fault, or S j3 With S j4 Fault; When in the low modulation region, using a three-to-two level hybrid modulation scheme, if the system recovers to normal, the fault type is switch S. j2 or S j4 A fault occurs if the fault is not detected, and vice versa if the fault is detected by the switching transistor S. j1 or S j3 The fault was ultimately located by identifying an open circuit fault in a single switch.

2. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 1, characterized in that: In step S1, the initial judgment of the phase with open-circuit fault based on image similarity theory is made to identify whether the single switch open-circuit fault occurs in the ABC phase bridge arm or the DEF phase bridge arm, based on the following: When a single-switch open-circuit fault occurs in phases ABC, the reference value of the composite fundamental MMF of phases ABC during the fault half-cycle is: When a single-switch open-circuit fault occurs in phases ABC, the combined MMF reference value for phases DEF during the fault half-cycle is: When a single-switch open-circuit fault occurs in phase DEF, the reference value of the fundamental MMF of phases ABC during the fault half-cycle is: When a single-switch open-circuit fault occurs in a phase of the DEF phase, the reference value of the combined MMF of the DEF phase during the fault half-cycle is: in The MMF coefficient is determined by the motor topology and is a constant; j =A~F are phases, The spatial angle is measured in electrical degrees. As the initial phase angle variable, when j When =A~F, φ They are 0, 2π / 3, 4π / 3, π / 6, 5π / 6, and 3π / 2, respectively; The current of the faulty phase. The number of turns in the motor stator winding. This refers to the current amplitude. ω Represents electric angular velocity. t For time; When the switching transistor faults are located in phases A through F respectively θ 0 is equal to 0, 2 / 3π, -2 / 3π, -1 / 3π, 1 / 3π and -π respectively.

3. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 2, characterized in that: In step S1, the fault is determined using the MMF balance of the two windings as a characteristic value, specifically including: The following explanation uses an SOSF (SO2-SO2) fault in the A-phase power switch as an example. θs = ωt = θe ,in θe It is the electrical angle of the rotor position; and Simplified to only ωt An expression with a single independent variable: Wherein, the subscript OCF-a indicates the result when a SOSF fault occurs in phase a, and the superscript θ s = ωt Indicates ωt Replace the original expression θ s Similarly, simplify the expression for faults in other phases; Establish fault determination feature quantities F η1 and F η2 To determine whether a fault has occurred and to locate the three-phase winding where the fault is located; Under normal circumstances F η1 and F η2 The ideal value is 1; when a SOSF switch failure occurs in the ABC phase bridge arm... F η1 Less than 1; when the DEF phase bridge arm switch SOSF fails, F η2 Less than 1.

4. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 3, characterized in that: In step S2, the actual fault characteristic value curve is determined based on the structural similarity index SSIM. f η1 and f η2 and F η1 and F η2 The degree of similarity between the resulting reference fault characteristic value curves is used to locate the faulty bridge arm, specifically including: Once the fault has been located in the ABC three-phase bridge arm, subsequent judgments will use... F η1 By setting a sliding window W Similarity determination is performed, window length W L As the fundamental frequency of the phase current changes, the time scale of the sampling window remains 1 / 8 of the fundamental frequency period, satisfying the following relationship: In the formula, f f It is the fundamental frequency. f s This refers to the sampling frequency; the data in the sampling window is updated in real time with the control, and the stored data is: In the formula, The actual current data stored in the sampling window. The curve representing the actual fault characteristic value from the i-th sample is shown. The theoretical value data of each phase stored in the sampling window. For the first i Secondary sampling j The reference fault curve for the phase; Average of actual current data Theoretical values ​​of each phase (mean values) Variance of actual current data Variance of theoretical values ​​for each phase and covariance The calculation is as follows: Similarity SM j Using the intermediate variable contrast measurement function C jr and structural measurement function S jr The calculation is as follows: in These represent the similarity scores of A, B, and C, respectively. This indicates the candidate phase for faults in the current similarity comparison; Based on the physical meaning of similarity, the maximum SM j corresponding j The value indicates the phase arm of the bridge where the fault is located, and the fault alarm flag is in position. F flag Set to 1, after an alarm fault occurs, SM j The judgment result is output after a delay of 0.05 fundamental frequency cycles; the maximum SM within 0.05 cycles is statistically analyzed. j The phase that appears most frequently in each phase is selected as the final faulty phase. The determination logic is as follows: in This indicates the number of times candidate phase j is determined to be a faulty phase within the statistical period L, where L represents the data length used for statistical and delay output. This represents the maximum value among the statistical counts. This indicates the fault type code.

5. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 4, characterized in that: In step S3, the system determines whether the fault belongs to S based on the time of its occurrence. j1 / S j2 Or S j3 / S j4 Types, specifically including: If S j1 / S j2 Fault, characteristic value of actual current calculation f η1 Appeared 0~π , F type Determined to be 1; otherwise, S j3 / S j4 In case of type fault, π~2π Within the range f η1 Reaching the threshold T h1 , F type The value is determined to be 4.

6. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 5, characterized in that: In step S4, when in the high modulation region, for S j1 / S j2 Type of fault, S j1 A fault will cause the current to approach zero, S j2 During a fault, the current is between zero and its normal value. Utilizing this characteristic, a length of [length missing] is used... W L The average current Δ of the sliding window calculation i j Thus distinguishing S j1 and S j2 Fault: in This represents the instantaneous current value of faulty phase j at the i-th sampling time; In the early stages of sliding window generation, it was not fully achieved. W L The length is calculated based on the actual sliding window length; a threshold is set. T h2 =0.1 i s Distinguish between the two types of faults, i s This is the stator current reference value: Under normal operating conditions, the fault switch indicator F sw Keep the value zero; when S j1 and S j2 When open circuit faults occur separately, F sw The corresponding values ​​are 1 and 2; The same method is used to distinguish S in the high modulation region j3 With S j4 Fault.

7. The multidimensional electrical fault diagnosis method for a dual-three-phase motor system based on magnetomotive force balance trajectory according to claim 5, characterized in that: In step S4, when in the low modulation region, a three-to-two-level hybrid modulation scheme is used for differentiation, that is, the faulty bridge arm uses two-level modulation, while the normal other bridge arms maintain three-level modulation; the switching function of the three-level system simulating two-level modulation is as follows: The switch state is P. S j1 =1, S j2 S j3 S j4 The value is 0, and the voltage is [missing information]. U dc / 2, where U dc This refers to the DC bus voltage of the inverter. The switch state is N. S j4 S is 1. j1 , S j2 S j3 The value is 0, and the electrode voltage is - U dc / 2; Simulating two-level operation using a three-level real system S j1 and S j4 Driven by a set of complementary PWM, and S j2 and S j3 Keep it closed; if S j2 or S j3 In case of a fault, the system will recover to normal operation in two-level mode. (Characteristic value) f η1 and f η2 It will also recover to around 1, combined with the already determined S j1 / S j2 Type of fault: ① If f η1 and f η2 If the value recovers to around 1, then the fault switch transistor is S. j2 ; ② Conversely, if f η1 and f η2 If the problem persists, the faulty switch is S. j1 ; This method effectively distinguishes S in the low-modulation region. j1 With S j2 Fault; use the same method to distinguish S in the low modulation region j3 With S j4 Fault.

Images