Dual three-phase permanent magnet synchronous motor inverter switching tube open-circuit fault diagnosis method

Abstract

The invention discloses a dual three-phase permanent magnet synchronous motor inverter switching tube open-circuit fault diagnosis method, belongs to the field of motor control and fault diagnosis, and aims at solving the problem that damage is continuously caused in the existing dual three-phase motor open-circuit fault diagnosis process. The method comprises the steps of collecting six-phase stator current, and mapping the six-phase stator current to a torque plane, a harmonic plane and a zero-sequence plane through vector space decoupling transformation; the torque plane current is decomposed into two independent sub-plane currents which are respectively corresponding to two sets of windings; and monitoring the harmonic plane current, and triggering a fault diagnosis process when a starting condition is met. In the diagnosis process, first-level and supplementary fault flag bits are calculated according to the characteristics of two torque sub-plane currents in a fundamental wave period so as to judge the fault type; positioning a fault bridge arm through the secondary fault flag bit; and finally, determining a specific fault switch tube by using the three-level fault flag bit. At the same time, the harmonic current reference value is updated to suppress torque ripples during the diagnosis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of motor control and fault diagnosis, and particularly relates to a method for diagnosing open-circuit fault of switch tube of inverter of dual three-phase permanent magnet synchronous motor. BACKGROUND

[0002] Multi-phase permanent magnet synchronous motor, especially dual three-phase permanent magnet synchronous motor, has been widely used in the field of high reliability requirement such as aerospace and electric vehicle due to its high power density, small torque ripple and strong fault tolerance. The stable operation of the motor depends largely on its driving inverter. However, statistics show that the power switch tube in the inverter is the most vulnerable link in the drive system, and its failure rate accounts for a large proportion of the total failure rate.

[0003] When the power switch tube has a short-circuit fault, it is usually quickly cut off by a hardware protection circuit or converted into an open-circuit fault. Subsequently, the system needs to rely on accurate fault diagnosis results to start the fault-tolerant control strategy to maintain the continued operation of the system. Therefore, fast and accurate fault diagnosis is a key prerequisite for achieving high reliability of the system.

[0004] Traditional fault diagnosis methods often analyze based on the motor model information after the fault, which has some limitations. For example, some methods analyze the sub-plane current trajectory and combine neural networks to diagnose faults, but when facing complex faults such as multiple switch tubes open circuit, the fault features will become unobvious, leading to diagnosis failure or misdiagnosis. In addition, the diagnosis needs to take one or more fundamental periods of time, and during the fault diagnosis process, the motor is still in operation. If the control strategy is not adjusted in time, the torque fluctuation caused by the fault may be aggravated, causing secondary damage to the motor and the load system.

[0005] Therefore, there is an urgent need in the art for a diagnosis method that can quickly and accurately diagnose single-tube and multi-tube open-circuit faults, and can suppress torque fluctuations and reduce damage to the system during diagnosis. SUMMARY

[0006] In order to solve the problem of continuous damage in the existing open-circuit fault diagnosis process of dual three-phase motor, the present application provides a method for diagnosing open-circuit fault of switch tube of inverter of dual three-phase permanent magnet synchronous motor.

[0007] The method for diagnosing open-circuit fault of switch tube of inverter of dual three-phase permanent magnet synchronous motor provided by the present application comprises the following steps:

[0008] Collecting six-phase stator currents of the motor;

[0009] Mapping the six-phase stator currents to a torque plane, a harmonic plane and a zero sequence plane through a vector space decoupling transformation;

[0010] The current of the torque plane is decomposed into two independent torque sub-plane currents, each of the torque sub-planes corresponding to a set of three-phase windings;

[0011] The harmonic current of the harmonic plane is monitored, and when a fault diagnosis starting condition is met, a fault diagnosis process is started; the process of diagnosing the open-circuit fault of the inverter is:

[0012] According to the characteristics of the two torque sub-plane currents in a fundamental period when the open-circuit fault occurs, a primary fault flag and a supplementary fault flag are calculated, and the type of the open-circuit fault of the inverter is preliminarily judged;

[0013] A secondary fault flag is calculated to locate the fault bridge arm;

[0014] A tertiary fault flag is calculated to locate the specific switch tube in the fault bridge arm;

[0015] During the diagnosis process, the amplitude and phase of the harmonic current are extracted by using a phase-locked loop based on a second-order generalized integrator, and the harmonic current reference value is dynamically updated after time delay compensation to suppress motor torque fluctuation.

[0016] Preferably, the vector space decoupling transformation is completed by using the following formula:

[0017]

[0018] wherein, and represent the three-phase currents of the first set of windings ABC and the second set of windings UVW of the double three-phase motor, respectively;

[0019] is the current of the torque plane ;

[0020] is the harmonic current of the harmonic plane ;

[0021] is the current of the zero sequence plane ;

[0022] The torque plane is split into two torque sub-planes and by Clark transformation, and the currents of the torque sub-planes are , m=1, 2, and there is a relationship: , .

[0023] Preferably, the fault diagnosis starting condition is:

[0024] .

[0025] Preferably, the process of updating the harmonic current reference value by using the phase-locked loop based on the second-order generalized integrator is as follows:

[0026] The harmonic current is input into the second-order generalized integrator, and two orthogonal signals and are generated.

[0027] Park transformation is performed on the orthogonal signals to obtain direct-axis current and quadrature-axis current .

[0028] The direct-axis current is taken as the amplitude of the updated harmonic current reference value.

[0029] The quadrature-axis current is compared with the reference current =0, and the error thereof generates the reference frequency adjustment amount by a PI regulator. The motor speed is compared with the reference frequency adjustment amount , and the error thereof generates the phase information of the harmonic current reference value by integration.

[0030] The new harmonic current reference value is synthesized by using the amplitude and the phase information. .

[0031] Preferably, the process of obtaining the primary fault flag is as follows:

[0032] The initial primary fault flag is calculated as follows:

[0033]

[0034] is the starting moment of fault diagnosis, is the fundamental period, and m=1, 2. and are the initial primary fault flags corresponding to the first and second sets of winding pairs, respectively.

[0035] is normalized to obtain the primary fault flag :

[0036] , and are the primary fault flags corresponding to the first and second sets of winding pairs, respectively.

[0037] Preferably, the supplementary fault flag is obtained as follows:

[0038]

[0039] wherein M m is a supplementary fault flag, is the time when the first-level fault flag is first 0 after the fault diagnosis begins.

[0040] Preferably, the preliminary judgment of the open-circuit fault type of the inverter is determined according to the numerical combination of the first-level fault flag and the supplementary fault flag , and the fault type includes:

[0041] when , , and , it is the single-pipe open-circuit fault of the first set of inverter;

[0042] when , , and , it is the single-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 120° in the first set of inverter;

[0043] when , , and , it is the single-pipe open-circuit fault of the second set of inverter;

[0044] when , , and , it is the single-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 120° in the second set of inverter;

[0045] when , , it is the parallel two-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 30°, the cross two-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 150°, or the single-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 90°;

[0046] when , , it is the cross two-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 30°, the parallel two-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 150°, or the single-pipe open-circuit fault of the two-phase bridge arms with a phase angle difference of 90°.

[0047] Preferably, the second-level fault flag is calculated, and the process of locating the fault bridge arm is as follows:

[0048] First, the initial second-level fault flag is calculated according to the following formula , :

[0049]

[0050] Then, numerical processing: Time, ; else ;

[0051] Then, the secondary fault flag bit is calculated as follows :

[0052]

[0053] Finally, according to the fault type determined by the primary fault flag bit, the phase with the minimum value in the selected is the fault phase bridge arm.

[0054] Preferably, the tertiary fault flag bit is calculated, and the process of locating the specific switch tube in the fault bridge arm is as follows:

[0055] First, the current of the determined fault phase is calculated to obtain the initial tertiary fault flag bit :

[0056]

[0057] In the formula, is the phase current of the fault phase;

[0058] The is normalized to obtain the tertiary fault flag bit :

[0059]

[0060] Then, according to the tertiary fault flag bit determine the specific switch tube position in the fault phase:

[0061] .

[0062] The beneficial effects of the present application: through the synergistic effect of dynamic optimization of harmonic plane reference value and multi-level fault diagnosis, the unity of diagnosis accuracy and system safety is realized. The traditional method adheres to the original control target after the fault, which leads to torque fluctuation and characteristic distortion. The present application actively adjusts the control strategy at the beginning of diagnosis, guides the system to enter a new stable transient state, which not only suppresses the torque fluctuation, but also provides a pure fault signal for subsequent diagnosis. On the basis of this stable signal, the multi-level fault flag bit can be accurately extracted and progressively analyzed, so as to quickly locate the fault switch tube. This mechanism of "stabilizing and diagnosing at the same time" forms a seamless transition from fault occurrence to fault-tolerant operation, which maximizes the risk reduction of the diagnosis process itself.

[0063] Traditional diagnostic schemes work independently between the control system and the diagnostic module after the fault occurs: the control system still tries to maintain the operating state as before the fault (e.g. zero forced harmonic current), while the diagnostic module struggles to extract the features from the distorted current. This control strategy, which insists on the original state, is contrary to the objective fact that the fault has occurred, not only leading to a dramatic torque fluctuation, but also making the fault features submerged in the distortion, increasing the difficulty and time of diagnosis.

[0064] The present application breaks through this limitation. The core is that once the signs of failure are monitored, a dynamic optimization process of the harmonic plane current reference value is immediately started. The process tracks the amplitude and phase of the harmonic current after the fault in real time based on the phase-locked loop of the second-order generalized integrator, and takes it as the new control target. The direct effect of this operation is that the control system no longer resists the harmonic component that inevitably appears after the fault, but guides the system into a new electromagnetic balance state that allows the existence of the component. This brings key benefits at the system level: significantly suppresses the torque pulsation during diagnosis, avoids the secondary impact of the fault on the system, and provides a stable and reliable current signal source for subsequent diagnosis.

[0065] It is on the basis of this stabilized current that the subsequent multi-stage fault diagnosis can exhibit its high precision and high robustness.

[0066] The primary fault flag is effectively stripped of common-mode interference such as load changes through periodic integration and normalization processing of the stabilized torque sub-plane current, so that the value of the flag is clearly associated with the fault mode (such as single tube, multiple tubes and their phase angle difference).

[0067] The supplementary fault flag uses the predictability of the stabilized waveform on the time axis to solve the ambiguity problem of specific types of single tube and multiple tube faults in macroscopic features by analyzing the energy distribution in different intervals of the fault period.

[0068] According to the numerical combination of the primary fault flag and the supplementary fault flag, the fault type is preliminarily judged.

[0069] Further, the secondary fault flag uses the deterministic mathematical relationship between the harmonic plane and the phase current after the fault to realize the positioning of the fault phase bridge arm.

[0070] Finally, the tertiary fault flag performs periodic integration on the stabilized current waveform, and can accurately determine whether the upper tube or the lower tube has an open circuit according to the sign and size of the integral value.

[0071] Therefore, the advantages of this invention form an organic whole: the dynamic optimization at the front end creates the necessary conditions for accurate diagnosis at the back end, while the rapid and accurate diagnosis at the back end enables the system to switch to the final fault-tolerant control strategy as early as possible, forming a seamless transition from fault occurrence to fault-tolerant operation. This method not only achieves rapid fault location but also minimizes the impact of the diagnostic process itself on the system by actively managing the system state after a fault, achieving a balance between diagnostic performance and system security. Attached Figure Description

[0072] Figure 1 This is a diagram of a dual three-phase permanent magnet synchronous motor control system using the diagnostic method of this invention;

[0073] Figure 2 This is a topology diagram of a dual three-phase permanent magnet synchronous motor inverter;

[0074] Figure 3 This is a diagram showing the relationship between the operating time zone and the switching transistor;

[0075] Figure 4 This is a schematic diagram of a phase-locked loop based on a second-order generalized integrator;

[0076] Figure 5 This is a schematic diagram of the principle of the open circuit fault diagnosis method for the winding of a dual three-phase permanent magnet synchronous motor described in this invention;

[0077] Figure 6 This is a control system diagram for a traditional dual three-phase permanent magnet synchronous motor. Detailed Implementation

[0078] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0079] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0080] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but this is not intended to limit the scope of the invention.

[0081] Specific Implementation Method 1: The following is combined with... Figures 1 to 6 For a description of this implementation method, please refer to [link / reference]. Figure 6 The control system diagram of a traditional dual three-phase permanent magnet synchronous motor first measures the basic parameters of the motor stator current and electrical angle: the six-phase stator current is collected by current sensors. and The rotor phase of the dual three-phase motor 6 is measured using encoder 7. Then, a Clark transformation is performed to... and The torque plane current in the two-phase stationary coordinate system is obtained through Clark converter 8. With harmonic plane current Next, a Park transformation is performed, using Park transformer 9 to transform... Transformed into a rotating coordinate system At the same time, the motor phase The actual motor speed n is output by the speed calculator 10 and fed back to the front end of the control system.

[0082] At the front end of the control system, the following reference value is given: Speed ​​reference value d-axis current reference value Set to zero, q-axis current reference value Harmonic current reference value in the harmonic current plane m=1,2 is set to 0.

[0083] Control process: Set the d-axis current reference value With the direct-axis current fed back to the front end The difference is then processed by PI controller 1 to obtain the control voltage that needs to be applied to the d-axis. q-axis current reference value Reference value for rotational speed The difference between the actual rotational speed n and the input to PI controller 1 is used to obtain the control voltage required to be applied to the q-axis. This voltage is then passed through another PI controller 1. , Output after inverse Park converter Input the vector space decoupling inverse transformation matrix 3; harmonic current reference value in the harmonic current plane. With harmonic current Differential, harmonic current reference value With harmonic current The difference between these two values ​​is then processed by the deadbeat controller 2 to generate the voltage vector command to be applied. In traditional control, the reference value of harmonic current in the harmonic current plane Set to 0; The common input vector space decoupling inverse transformation matrix 3 outputs dual three-phase voltage commands. and The dual zero-sequence injection PWM module 4 is used to inject PWM according to dual three-phase voltage commands. and Six PWM signals for controlling the on-off of power switch tubes are calculated, and the six PWM signals control the operation of the double three-phase motor 6 through the six-phase inverter 5.

[0084] When the motor has an inverter open circuit fault, due to the closed-loop control of the d-q plane, the current amplitude on the torque plane is increased in a direction that can compensate for the missing magnetic motive force. Meanwhile , the current on the harmonic plane has a certain quantitative relationship with the current on the harmonic plane . Therefore, if the given current reference value on the harmonic plane is still set to zero, the increase in the current amplitude on the harmonic plane will be hindered , which in turn hinders the compensation for the missing magnetic motive force, resulting in a direct consequence of the motor having a severe fluctuation, and in severe cases, causing the motor to be damaged. Diagnosing the motor when it has a fault requires a certain length of time, and allowing the motor torque to fluctuate during the diagnosis process is not conducive to the service life of the motor.

[0085] In view of this, the present application proposes a new diagnosis method, as shown in Figures 1 to 5 , which comprises the following steps:

[0086] Collecting the six-phase stator current of the motor

[0087] Mapping the six-phase stator current to the torque plane , the harmonic plane and the zero sequence plane through a vector space decoupling transformation ; is the current on the torque plane ; is the harmonic current on the harmonic plane ; is the current on the zero sequence plane

[0088] Decomposing the current on the torque plane into two independent torque sub-plane currents, each of the torque sub-planes corresponding to a set of three-phase windings; the torque plane is split into two torque sub-planes , through a Clark transformation, and the current on the torque sub-plane is , m = 1, 2, and there is a relationship: , . Wherein is the current on the torque sub-plane of the first set of windings (ABC), is the current on the torque sub-plane of the second set of windings (UVW).

[0089] Monitoring the harmonic current amplitude of the harmonic plane, when the harmonic current amplitude exceeds a preset threshold, starting the fault diagnosis process; the process of diagnosing the winding open circuit fault is:

[0090] According to the characteristics of the two torque sub-plane currents in a fundamental cycle, the first-level fault flag and the supplementary fault flag are calculated, and the inverter open circuit fault type is preliminarily judged;

[0091] The second-level fault flag is calculated, and the fault bridge arm is positioned;

[0092] The third-level fault flag is calculated, and the specific switch tube in the fault bridge arm is positioned.

[0093] In the diagnosis process, the amplitude and phase of the harmonic plane current are extracted by using the phase-locked loop 11 based on the second-order generalized integrator, and the harmonic current reference value is dynamically updated after time delay compensation, so as to suppress the motor torque fluctuation.

[0094] A complete and systematic fault diagnosis solution is provided, which integrates fault detection, type identification, accurate positioning and online protection in one process.

[0095] The synchronous diagnosis and torque fluctuation suppression are realized, and the key problem that the motor is damaged due to torque fluctuation during diagnosis in the traditional method is solved.

[0096] The foundation of the overall beneficial effects of the present application, such as rapidness, accuracy and safety, is laid.

[0097] Referring to Figure 1 The vector space decoupling transformation is completed by the following formula:

[0098] (1)

[0099] Wherein, And The three-phase currents of the first set of windings ABC and the three-phase currents of the second set of windings UVW of the double three-phase motor are represented by I ABC and I UVW respectively.

[0100] An accurate mathematical tool is provided, which decouples the complex six-phase system into independent subspaces, so that the fault characteristics can be separated and extracted.

[0101] The correspondence between the torque sub-plane and the physical winding (one set of windings corresponds to one sub-plane) is determined, which lays a theoretical foundation for subsequent fault phase positioning by analyzing the sub-plane state.

[0102] The effectiveness and feasibility of the subsequent fault feature extraction algorithm are ensured.

[0103] During the operation of the motor, the six-phase stator currents , the harmonic current of the harmonic plane obtained by Park transformation and Clark transformation , , the current of the torque plane , , and the current of two torque sub-planes , , , , judging whether the condition of starting fault diagnosis is reached

[0104] (2)

[0105] When or any harmonic current reaches the above condition, start the fault diagnosis, and at the same time, the harmonic current reference value is no longer zero, but is updated to a new value.

[0106] A reliable and sensitive fault detection trigger is set. The harmonic current amplitude is close to zero in normal operation and significantly increases in fault, and the condition can effectively distinguish between normal and fault states.

[0107] False starts are avoided, and it is ensured that the diagnosis process is only triggered when a real fault or a serious anomaly occurs, improving the reliability of the system.

[0108] The premise for fast response is provided, and the diagnosis is started as soon as the fault is detected, shortening the fault response time.

[0109] Referring to FIG. 11, the process of updating the harmonic current reference value by using the phase-locked loop 11 based on the second-order generalized integrator is as follows: Figure 4 The harmonic currents

[0110] , m = 1, 2 are input into the second-order generalized integrator to generate two orthogonal signals and ; the second-order generalized integrator has good dynamic performance and strong filtering ability. Park transformation is performed on the orthogonal signals to obtain direct-axis current

[0111] and quadrature-axis current ; The direct-axis current

[0112] is taken as the amplitude of the updated harmonic current reference value; The quadrature-axis current

[0113] is compared with the reference current = 0, and the error thereof generates the reference frequency adjustment amount by a PI regulator , and the motor speed The reference frequency adjustment amount The error is integrated to generate phase information of the harmonic current reference value ;

[0114] Or The amplitude and the phase information are respectively synthesized to generate new harmonic current reference values .

[0115] (3)

[0116] The present application adjusts the harmonic current reference value quickly when the motor winding has an open circuit fault , and obtains stable output torque and stable torque sub-plane in a short time, and maintains relatively stable torque output during fault diagnosis, thereby avoiding further damage to the motor caused by the fault during diagnosis. During the open circuit fault diagnosis process, the current reference value is optimized on the z1-z2 plane, so that the closed-loop current control can drive the current on the plane to the ideal value, thereby realizing fault-tolerant control of the winding open circuit fault. The values of z1 and z2 are collected in real time, and after time delay compensation, they are used as the new current reference value on the z1-z2 plane. This is continuously iterated until the current amplitude on the z1-z2 plane reaches the ideal amplitude that can compensate for the missing magnetomotive force. The phase-locked loop 11 based on the second-order generalized integrator realizes real-time collection of z1 and z2, and iterates the current on the z1-z2 plane to the ideal value.

[0117] Referring to FIG. 6, the updated harmonic current reference value is fed back to the control front end. Since the harmonic plane reference value optimization requires a high dynamic response speed controller, the collected amplitude and phase information are processed and input into the deadbeat controller for control. The core principle of the deadbeat controller is that, according to the discrete mathematical model of the controlled object (motor harmonic plane), an optimal voltage vector command is directly calculated based on the actual harmonic current and the given reference value at the current sampling time. Figure 1

[0118] The above is the core of the function of suppressing torque fluctuation during diagnosis. The amplitude and phase of the harmonic current are extracted in real time based on the phase-locked loop 11 of the second-order generalized integrator, and are used as new reference values, thereby realizing closed-loop control of the harmonic current.​​​​​​​​​​​​​​​

[0119] This enables the control system to actively compensate for the missing magnetomotive force due to open circuit faults, thereby effectively suppressing fluctuations in torque and speed.

[0120] The second-order generalized integrator 11 has excellent filtering performance, can accurately extract the fundamental component, reduce the impact of noise interference on the update of the reference value, and ensure the stability of the torque ripple suppression effect.

[0121] This minimizes the damage to the motor and drive system caused by faults during the diagnostic process.

[0122] The following explains the process of diagnosing open-circuit faults in motor inverters.

[0123] Figure 3 The document specifies the switching transistors involved in modulation when the motor operates in different time zones, with each 60° phase interval representing one time zone. When only a single transistor is open-circuited, the phase current waveform exhibits normal operation for half of the electrical cycle and fault operation for the other half. In this case, one electrical cycle is divided into two large time zones: normal and fault. Similarly, when single-transistor open-circuit faults occur in two different bridge arms, the switching transistors involved in modulation are... Figure 3 As shown, a complete electrical cycle is divided into four time zones: the normal operating time zone with no faulty switching transistors, the time zone with the first faulty switching transistor, the time zone with the second faulty switching transistor, and the time zone with both faulty switching transistors. Therefore, when two switching transistors fail, the current will be affected by four time zones in succession. Taking the simultaneous open circuit of S1 and S3 as an example, the current of the first set of three-phase windings should be represented by four time zones. The first time zone is the impact of the fault of only S1 being open on the current, the second time zone is the impact of the fault of both S1 and S3 being open on the current, the third time zone is the impact of the fault of only S3 being open on the current, and the fourth time zone is the normal operating state with no faulty switching transistors. The specific expression of the current of the first set of windings is shown in equations (4) to (6).

[0124] (4)

[0125] (5)

[0126] (6)

[0127] In the formula, This represents the amplitude of the phase current under normal operating conditions.

[0128] like Figure 5The flow chart of the open-circuit fault diagnosis method for dual three-phase permanent magnet synchronous motor inverter. Taking A-phase bridge arm fault, AU-phase bridge arm fault, AW-phase bridge arm fault, and AB-phase bridge arm fault as examples, single tube open-circuit fault, single bridge arm complete open-circuit, and phase angle difference of 30° or 150°, 90°, 120°, and two tube open-circuit fault in two bridge arms are illustrated. Other combinations can be diagnosed by using the same diagnosis method. When single tube or two tube open-circuit fault occurs in the A-phase bridge arm, the α-β axis current in the two independent torque planes in the fault time zone is:

[0129] (7)

[0130] When two tube open-circuit faults occur in the AU two bridge arms, the α-β axis current in the two independent torque planes in the fault time zone where both switch tubes are open-circuit is:

[0131] (8)

[0132] When AW two-phase winding open-circuit, the α-β axis current in the two independent torque planes in the fault time zone where both switch tubes are open-circuit is:

[0133] (9)

[0134] When AB two-phase winding open-circuit, the α-β axis current in the two independent torque planes in the fault time zone where both switch tubes are open-circuit is:

[0135] (10)

[0136] Wherein is the current in the torque sub-plane corresponding to the first set of windings is the current in the torque sub-plane corresponding to the second set of windings, is the q-axis current.

[0137] After meeting the formula (2) start diagnosis condition, the diagnosis starts, and the diagnosis process is shown in Figure 5 .

[0138] First, calculate the primary fault flag bit, the primary fault flag bit The acquisition process is:

[0139] Calculate the initial primary fault flag bit :

[0140] (11)

[0141] is the moment when the fault diagnosis starts,​​​ is the fundamental period, m = 1, 2; are the initial first fault flag bits corresponding to the first and second sets of winding pairs, respectively;

[0142] The values of the initial first fault flag bits obtained for the open-circuit fault of the switch tube in different types of bridge arms are shown in Table 1:

[0143] Table 1 Switch tube open-circuit fault type (initial first fault flag bit)

[0144]

[0145] In the table is the amplitude of the quadrature-axis current.

[0146] In an ideal case, when the open-circuit of the switch tube occurs, the initial first fault flag bit should satisfy the above table. However, considering that the two switch tube open-circuit faults will bring complex working states, the system is in normal operation, two-tube open-circuit operation and other non-full fault operation in a fundamental period, and the working states repeatedly switch, the controller cannot complete the dynamic adjustment of the harmonic plane reference value in a limited time zone in some cases.

[0147] For example, when the upper switch tube of the A bridge arm and the lower switch tube of the U bridge arm are open-circuited, the positive half cycle of the A-phase current is missing, the negative half cycle of the U-phase current is missing, and the working time zone of the two switch tubes being open-circuited is only The time zone of the two switch tubes being open-circuited has a relatively short impact time, and the initial first fault flag bit mainly shows the single-tube open-circuit fault information. When the upper switch tube of the A bridge arm and the upper switch tube of the U bridge arm are open-circuited, the positive half cycle of the A-phase current is missing, the positive half cycle of the U-phase current is missing, and the working time zone of the two switch tubes being open-circuited is The time zone of the two switch tubes being open-circuited has a relatively long impact time, and the initial first fault flag bit can show the double-tube open-circuit fault information. As described in the foregoing winding fault analysis, the time zone of the two switch tubes of the bridge arm being open-circuited at a phase angle difference of 30° or 150° is similar to that of the two-phase winding being open-circuited, and this fault condition is relatively serious. Only these two types of faults have this problem, and the first fault flag bits of the remaining faults are not affected. Therefore, the diagnosis table of the first fault flag bit needs to be supplemented as shown in Table 2:

[0148] Table 2 Switch tube open-circuit fault type (initial first fault flag bit)

[0149]

[0150] ​The open-circuit fault of the switch tube in several different bridge arms can be distinguished by the initial first fault flag. In order to improve the anti-interference ability of the diagnosis method, the threshold is normalized by selecting appropriate parameters, and the normalization formula is:

[0151] (12)

[0152] The first and second sets of windings are respectively indicated by the first fault flag.

[0153] The open-circuit fault type of the switch tube is judged by the first fault flag after normalization as shown in Table 3:

[0154] Table 3 Open-circuit fault of switch tube (first fault flag)

[0155]

[0156] It can be found that due to the interference of the non-full fault zone, the single bridge arm switch tube fault and the two-tube open-circuit fault of the bridge arm with a phase angle difference of 120° cannot be distinguished, which will lead to the failure to effectively distinguish the single-tube open-circuit fault and the double-tube open-circuit fault, and will affect the subsequent further diagnosis. Therefore, after detecting the single bridge arm open-circuit fault of the switch tube or the two-tube open-circuit fault of the bridge arm with a phase angle difference of 120°, a supplementary fault flag is set to distinguish them. The supplementary fault flag is set as:

[0157] (13)

[0158] wherein, is the first time when the first fault flag is 0 after the start of fault diagnosis.

[0159] The diagnosis table of the supplementary fault flag is shown in Table 4:

[0160] Table 4 Supplementary fault flag

[0161]

[0162] In the table, only the single-tube open-circuit and the two-tube open-circuit in the same set of inverters need to be further distinguished by the supplementary fault flag. Among them, is the supplementary fault flag for distinguishing the single-tube open-circuit and the two-tube open-circuit fault in the first set of inverters, is the supplementary fault flag for distinguishing the single-tube open-circuit and the two-tube open-circuit fault in the first set of inverters. × indicates that this supplementary fault flag is not needed.

[0163] In summary, the preliminary judgment of the open-circuit fault type of the inverter is based on the first fault flag and the supplementary fault flag The fault type includes:

[0164] When , , and , it is a single tube open circuit fault of the first set of inverters;

[0165] When , , and , it is a single tube open circuit fault of the first set of inverters;

[0166] When , , and , it is a single tube open circuit fault of the second set of inverters;

[0167] When , , and , it is a single tube open circuit fault of the second set of inverters;

[0168] When , , it is a parallel two-tube open circuit fault of the two-phase bridge arms with a phase angle difference of 30°, a cross two-tube open circuit fault of the two-phase bridge arms with a phase angle difference of 150°, or a single tube open circuit fault of the two-phase bridge arms with a phase angle difference of 90°;

[0169] When , , it is a parallel two-tube open circuit fault of the two-phase bridge arms with a phase angle difference of 30°, a cross two-tube open circuit fault of the two-phase bridge arms with a phase angle difference of 150°, or a single tube open circuit fault of the two-phase bridge arms with a phase angle difference of 90°.

[0170] The above achieves rapid and parallel preliminary identification of the fault type. By querying the preset threshold interval table, the single-phase open circuit, two-phase open circuit (and different phase angle differences) fault can be distinguished in a short time.

[0171] The unique threshold interval setting (such as 0.65, 0.85, 1.15, 1.85, etc.) effectively distinguishes the fault modes that are easily confused in traditional methods (for example, distinguishing single-phase faults and certain two-phase faults), significantly reducing the misjudgment rate.

[0172] The fault range (which set of inverters) is preliminarily determined, which narrows the range for accurate positioning of the secondary diagnosis and improves the efficiency.

[0173] After determining the fault type, the secondary fault flag bit is further calculated to locate the fault phase. This step is divided into two categories: positioning of single-phase open circuit faults and positioning of two-phase open circuit faults.

[0174] The vector space decoupling transformation matrix of the double three-phase permanent magnet synchronous motor according to formula (1) is

[0175] (14)

[0176] When the inverter switch tube occurs open circuit fault, the corresponding phase current in the fault time zone is zero, which is substituted into the vector space decoupling transformation matrix, and the following formula can be obtained:

[0177] (15) It can be seen from the above formula that under different fault conditions, the harmonic plane and the torque plane have a special quantitative relationship, and the current in the harmonic plane is zero under the condition that the motor is fault-free. Six groups of expressions are obtained by processing, and the secondary fault flag bit reflecting the open circuit fault is obtained , :

[0178] The process of calculating the secondary fault flag bit and positioning the fault phase is as follows:

[0179] First, calculate the initial secondary fault flag bit according to the following formula :

[0180]

[0181] Then, numerical processing is performed: When, ; otherwise ;

[0182] Next, calculate the secondary fault flag bit according to the following formula :

[0183]

[0184] Finally, according to the fault type determined by the primary fault flag bit, identify the secondary flag bit The smallest one or two bits in the secondary fault flag bit determine the single bridge arm or two bridge arms in which the switch tube open circuit fault occurs, and the fault bridge arm is diagnosed within one fundamental period. Through the dynamic adjustment of the harmonic plane reference value, the phase current waveform is stable, and the stability of the torque output waveform is greatly improved. Combined with the primary fault flag bit and the supplementary fault flag bit, the secondary fault flag bit can accurately distinguish the bridge arm in which the switch tube open circuit fault occurs.

[0185] Further, according to the tertiary fault flag bit, it is determined which switch tube in the fault phase is faulty.

[0186] The process of calculating the tertiary fault flag bit and positioning the specific switch tube of the fault phase is as follows: ​

[0187] First, the current of the determined fault phase is calculated according to the following formula to obtain an initial three-level fault flag bit :

[0188]

[0189] In the formula, is the phase current of the fault phase;

[0190] is normalized to obtain a three-level fault flag bit :

[0191]

[0192] Then, according to the three-level fault flag bit the specific switch position in the fault phase is determined:

[0193] .

[0194] While the application has been described with reference to particular embodiments, it will be understood that the examples are merely illustrative of the principles and applications of the application. It will be further understood that numerous modifications can be made to the illustrative embodiments, and that other arrangements can be devised without departing from the spirit and scope of the present application as defined by the appended claims. It will be understood that the features described with reference to individual embodiments can be used in other described embodiments. It will be further understood that features described with reference to individual embodiments can be combined with other described embodiments.​

Claims

1. A method for diagnosing open-circuit fault of switch tubes of an inverter for a dual three-phase permanent magnet synchronous motor, characterized by, The method comprises the following steps: Collecting six-phase stator currents of the motor; Mapping the six-phase stator currents to a torque plane, a harmonic plane and a zero sequence plane through a vector space decoupling transformation; Decomposing the current in the torque plane into two independent torque sub-plane currents, each of which corresponds to a set of three-phase windings; Monitoring the harmonic current in the harmonic plane, and starting a fault diagnosis process when a fault diagnosis starting condition is met; the process of diagnosing an inverter open-circuit fault is as follows: According to the characteristics of the two torque sub-plane currents in a fundamental period when the open-circuit fault occurs, calculating a first-level fault flag and a supplementary fault flag, and preliminarily judging the type of the inverter open-circuit fault; Calculating a second-level fault flag to locate the fault bridge arm; Calculating a third-level fault flag to locate a specific switch tube in the fault bridge arm; In the diagnosis process, the amplitude and phase of the harmonic plane current are extracted synchronously by using a phase-locked loop based on a second-order generalized integrator, and the harmonic current reference value is dynamically updated after time delay compensation to suppress motor torque fluctuation.

2. The method according to claim 1, wherein The vector space decoupling transformation is completed by using the following formula: wherein, and IABCand IUVWrepresent the three-phase currents of the first set of windings ABC and the second set of windings UVW of the dual three-phase motor, respectively; for the torque plane current; harmonic plane harmonic current; of the zero sequence plane of the current; The torque plane is split into two torque sub-planes by a Clark transformation , , the currents of the torque sub-planes are , m = 1,2, there is the relation , .

3. The method according to claim 2, wherein, The fault diagnosis starting condition is as follows: 。 4. The method according to claim 1, wherein, The process of updating the harmonic current reference value by using the phase-locked loop based on the second-order generalized integrator is as follows: Harmonic current m=1,2 input second-order generalized integrator, generating two orthogonal signals and ; performing a Park transformation on the quadrature signal to obtain a direct axis current and a quadrature axis current ; said direct axis current as an updated harmonic current reference value; The cross-axis current is compared with a reference current , and the error is used to generate a reference frequency adjustment amount by a PI regulator The motor speed is compared with the reference frequency adjustment amount , and the error is used to generate phase information of the harmonic current reference value by integration ; combining the amplitude and the phase information to synthesize a new harmonic current reference value .

5. The method according to claim 1, wherein, The process of obtaining the first-level fault flag is as follows: computing initial primary fault flag bits : is the starting time of fault diagnosis, is the fundamental period, m = 1, 2; are the initial first and second set of winding corresponding to the first level fault flag bit, respectively; To normalize the acquisition of primary fault flag bits : , First and second set of winding corresponding to the first level of fault flag bit, respectively.

6. The method according to claim 5, wherein supplemental fault flag is obtained as follows: wherein M m is a supplementary fault flag, is the time when the first level fault flag is 0 after the start of the fault diagnosis.

7. The method according to claim 6, wherein The initial determination of the inverter open-circuit fault type is based on the primary fault flag. With supplementary fault flag bits The fault types, determined by the combination of numerical values, include: When , , and , it is a first set of inverter single tube open circuit fault; When , , and , a single-pipe open-circuit fault occurs in the two-phase bridge arms of the first set of inverters with a phase angle difference of 120°. When , , and , it is a second set of inverter single tube open circuit failure; When , , and , a single-pipe open-circuit fault occurs in two-phase bridge arms with a phase angle difference of 120° in the second set of inverters. When , , the parallel two-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 30°, the cross two-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 150°, or the single-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 90° occurs respectively. When , the cross two-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 30°, the parallel two-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 150°, or the single-tube open circuit fault in the two-phase bridge arm with a phase angle difference of 90° occurs, respectively.

8. The method according to claim 7, wherein, The process of calculating the second-level fault flag to locate the fault bridge arm is as follows: First, the initial secondary fault flag bit is calculated by the following equation , : Then, the numerical processing: Time, ; else ; Next, the secondary failure flag bit is calculated by the following equation : Finally, according to the fault type determined by the primary fault flag bit, select The phase or two phases with the smallest value in the middle are the fault phase bridge arms.

9. The method according to claim 8, wherein, The process of calculating the third-level fault flag to locate the specific switch tube in the fault bridge arm is as follows: First, the current of the determined faulty phase is calculated to obtain an initial tertiary fault flag bit according to the following formula : In the formula, is the phase current of the faulty phase; To perform normalization to obtain a three-level fault flag bit : Then, according to the three-level fault flag bit Determine the specific switch position in the fault phase: 。

Images