Dual sliding mode observer active switching method under single-phase open-circuit fault of dual three-phase motor

Abstract

The application discloses a double sliding mode observer active switching method under single-phase open-circuit fault of a double three-phase motor, first, a first sliding mode observer and a second sliding mode observer are respectively constructed for two sets of windings of a double three-phase permanent magnet synchronous motor to estimate respective back electromotive forces; reference voltages of the second set of windings are subjected to coordinate rotation transformation to align the reference voltages of the two sets of windings in phase; amplitudes of the reference voltages of the two sets of windings are compared, when an amplitude difference exceeds a preset threshold, it is determined that a fault occurs, and according to which set of reference voltage amplitude increases, the fault is identified; according to the fault diagnosis result, the sliding mode observer corresponding to the healthy winding is autonomously switched to and the fault state is locked; through a phase-locked loop, a rotor position and a rotating speed are extracted from the back electromotive force output by the selected sliding mode observer. The application can realize real-time fault detection, observer autonomous switching, continuous and accurate estimation of the rotor position and the rotating speed.

Description

Technical Field

[0001] This invention relates to the field of motor control technology, specifically to rotor position estimation and fault diagnosis technology for dual three-phase permanent magnet synchronous motors, and particularly to an active switching method for dual sliding mode observers under single-phase open-circuit faults in dual three-phase motors. Background Technology

[0002] Dual three-phase permanent magnet synchronous motors (DTP-PMSMs) possess advantages such as high power density, low torque ripple, and good fault tolerance, demonstrating significant application potential in drive systems for low-altitude economically relevant fields such as electric vertical takeoff and landing (eVTOL) aircraft and unmanned aerial vehicles (UAVs). The core of achieving high-performance motor control lies in obtaining accurate rotor position information. However, in applications such as aerospace propulsion, the motor's operating environment is complex. Environmental factors such as high speed, extreme temperatures, and strong vibrations can cause position sensor failure, thereby affecting the safe and stable operation of the entire system. Therefore, sensorless control technology has become an important research direction for improving the reliability of motor drive systems.

[0003] In sensorless control strategies, the sliding mode observer (SMO) is particularly suitable for medium- and high-speed operation due to its strong robustness against parameter variations and external disturbances. Traditional sliding mode observers typically use a sign function as the switching control law, which, while exhibiting good robustness, introduces high-frequency chattering and phase lag. To mitigate this, researchers have employed continuous or smooth switching functions to improve the sliding mode observer, thereby accelerating convergence and suppressing chattering. For dual three-phase permanent magnet synchronous motors, this can be achieved based on their decoupled... A sliding mode observer is designed in subspace to obtain rotor position information by estimating the back electromotive force.

[0004] Although dual three-phase permanent magnet synchronous motors possess a certain degree of fault tolerance due to phase redundancy, the inverter remains a high-risk component in actual motor drive systems, with open-circuit faults in power switching devices being particularly typical. Once an open-circuit fault occurs in one phase of the inverter, the motor windings exhibit an asymmetrical operating state. Traditional sliding mode observers, designed based on symmetrical models, suffer from model mismatch under fault conditions, leading to distortion in back EMF estimation. This significantly reduces or even eliminates the accuracy of rotor position and speed estimations, severely impacting the system's continuous operation capability.

[0005] In existing technologies, some studies have proposed a scheme for designing dual sliding mode observers for dual three-phase motors. This scheme utilizes healthy windings to construct the observer under single-phase open-circuit fault conditions and considers the coupling effect between windings in the model, thereby avoiding back EMF distortion. However, these schemes typically assume that the fault type and location are known, do not address real-time fault diagnosis mechanisms, and do not provide autonomous switching strategies for the observer. In practical applications, they require manual judgment or support from external diagnostic systems, thus limiting their practicality.

[0006] Therefore, a technical solution is needed that can detect faults in real time, automatically switch observers, and ensure the continuity and accuracy of rotor position and speed estimation when a single-phase open-circuit fault occurs, so as to improve the fault tolerance and operational reliability of the dual three-phase permanent magnet synchronous motor drive system under fault conditions. Summary of the Invention

[0007] The purpose of this invention is to provide an active switching method for dual sliding mode observers under single-phase open-circuit faults in dual three-phase motors, in order to solve the problems in the prior art where the performance of traditional sliding mode observers in dual three-phase permanent magnet synchronous motors deteriorates significantly or even fails under single-phase open-circuit faults, and the existing dual sliding mode observer methods lack real-time fault diagnosis and autonomous switching mechanisms, resulting in limited continuous operation capability of the system.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A dual-sliding-mode observer active switching method for a dual three-phase permanent magnet synchronous motor under a single-phase open-circuit fault includes:

[0010] For the two sets of windings of the dual three-phase permanent magnet synchronous motor, a first sliding mode observer and a second sliding mode observer are constructed respectively to estimate the back electromotive force of the two sets of windings.

[0011] The reference voltage of the second winding is subjected to coordinate rotation transformation so that the reference voltages of the two windings are aligned in phase.

[0012] The amplitudes of the reference voltages of the two windings are compared. When the amplitude difference exceeds a preset threshold, a fault is determined to have occurred, and the winding where the fault is located is identified based on the amplitude comparison results.

[0013] Based on the fault diagnosis results, the system automatically switches to the sliding mode observer corresponding to the healthy winding and locks the fault status to avoid accidental switching.

[0014] The rotor position and speed are extracted from the back electromotive force output by the selected sliding mode observer using a phase-locked loop.

[0015] In a further preferred embodiment, both the first and second sliding mode observers employ a hyperbolic tangent function. As a switching function, it suppresses high-frequency chattering:

[0016]

[0017] in It is the sliding surface of the winding.

[0018] In a further preferred embodiment, the sliding surface of the first and second sliding mode observers is defined as the difference between the actual current value and the estimated current value.

[0019] In a further preferred embodiment, the rotation angle of the coordinate rotation transformation is 30 degrees to eliminate the inherent phase difference between the two sets of windings of the dual three-phase motor.

[0020] A further preferred embodiment is that the fault diagnosis criteria are: both sets of windings are in... The difference in reference voltage amplitude on the shaft exceeds the preset threshold, or the two sets of windings are... If the difference in the amplitude of the reference voltage on the shaft exceeds the preset threshold, a fault is identified.

[0021] A further preferred embodiment is that the fault diagnosis is based on the following expression:

[0022]

[0023] in To diagnose the trigger threshold, The reference voltage for the first winding is at shaft and Components on the axis, The reference voltage of the second winding after rotational transformation is in shaft and Components on the axis, This is a fault indicator.

[0024] Further optimized solutions, based on fault indicators Perform autonomous switching. A value of 0 indicates no fault. A value of 1 indicates a fault in the first winding. A value of 2 indicates a fault in the second winding; the rule for autonomous switching is: when... When the back electromotive force of the first sliding mode observer is equal to 0, When the back electromotive force of the second sliding mode observer is equal to 1, when When the value is equal to 2, the back electromotive force of the first sliding mode observer is used.

[0025] In a further preferred embodiment, locking the fault state means: after detecting a fault and completing the switch, keeping the fault flag in the fault state and not performing fault diagnosis operations for a preset period of time to avoid erroneous switching caused by signal fluctuations.

[0026] The present invention also provides a sensorless control system for a dual three-phase permanent magnet synchronous motor, comprising:

[0027] The first sliding mode observer module and the second sliding mode observer module are respectively configured to estimate the back electromotive force of the two sets of windings of the dual three-phase permanent magnet synchronous motor;

[0028] The coordinate rotation transformation module is configured to perform coordinate rotation transformation on the reference voltage of the second set of windings so that the reference voltages of the two sets of windings are aligned in phase.

[0029] The fault diagnosis module is configured to compare the amplitudes of two sets of winding reference voltages. When the amplitude difference exceeds a preset threshold, a fault is determined to have occurred, and the winding where the fault is located is identified.

[0030] The active switching module is configured to automatically switch to the sliding mode observer corresponding to the healthy winding based on the fault diagnosis results and lock the fault status.

[0031] The phase-locked loop module is configured to extract rotor position and speed from the back electromotive force output by the selected sliding mode observer.

[0032] Beneficial effects

[0033] The beneficial effects of this invention are:

[0034] (1) By using the fault diagnosis method based on the comparison of the reference voltage amplitude of the controller output, it is possible to detect open circuit faults in real time and accurately identify faulty windings. The diagnosis response time is less than 50 milliseconds, the accuracy reaches 100%, and the diagnosis process has strong anti-interference ability and does not require additional sensors or complex signal processing.

[0035] (2) By using a dual sliding mode observer architecture and an active switching strategy, the observer can quickly switch to the healthy winding after a fault occurs, which significantly improves the fault tolerance of the system under fault conditions. Experimental results show that compared with the traditional sliding mode observer method, the peak value of the speed estimation error after the fault is reduced from nearly 60 rpm to about 40 rpm, an improvement of 33%; the peak value of the rotor angle estimation error is reduced from about 0.5 radians to about 0.4 radians, an improvement of 20%; and the recovery time after the fault is shortened from more than 3 seconds to less than 1 second, an improvement of more than 66%.

[0036] (3) By eliminating the 30-degree phase difference between the two sets of windings through coordinate rotation, the consistency and accuracy of rotor position estimation are ensured, and the continuity of rotor position and speed estimation before and after the fault is guaranteed. After switching, the position estimation is smooth and continuous without jumps, which significantly enhances the overall robustness and reliability of the motor drive system under the condition of no position sensor. It is particularly suitable for application scenarios with extremely high reliability requirements, such as electric vertical take-off and landing aircraft and UAVs.

[0037] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0038] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0039] Figure 1 This is an overall block diagram of a sensorless control system based on a dual sliding mode observer.

[0040] Figure 2 This is a schematic diagram of the proactive strategy switching process.

[0041] Figure 3 The diagram shows a comparison of the reference voltage after rotation transformation, where (a) is the reference voltage waveform before rotation and (b) is the reference voltage waveform after rotation.

[0042] Figure 4 The diagram shows the results of the fault diagnosis experiment, where (a) is the fault diagnosis result of phase C open circuit and (b) is the fault diagnosis result of phase U open circuit.

[0043] Figure 5 The diagram shows a comparison of single-phase open-circuit faults under sensorless closed-loop control, where (a) represents the control result using the traditional sliding mode observer method and (b) represents the control result using the dual sliding mode observer active switching strategy. Detailed Implementation

[0044] The specific embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of this application can be combined with each other.

[0045] like Figure 1 As shown, this invention proposes an overall architecture for a sensorless control system based on dual sliding mode observers. Addressing the fault-tolerant control requirements of dual three-phase permanent magnet synchronous motors under single-phase open-circuit faults, this system constructs two independent sliding mode observers to monitor the state of the two three-phase windings respectively. Real-time fault diagnosis is achieved based on the amplitude comparison of the controller's output reference voltage. Upon detecting a fault, the system autonomously switches to the observer corresponding to the healthy winding, thereby ensuring the continuity and accuracy of rotor position and speed estimation, significantly improving the system's fault tolerance and operational reliability.

[0046] Example 1:

[0047] This embodiment provides a dual sliding mode observer active switching method for a dual three-phase permanent magnet synchronous motor under a single-phase open-circuit fault, applicable to sensorless operation of the dual three-phase permanent magnet synchronous motor under a single-phase open-circuit fault. The method includes the following steps:

[0048] Step 1: Design a dual sliding mode observer to estimate the back electromotive force of the two sets of windings respectively:

[0049] For a dual three-phase permanent magnet synchronous motor with two sets of three-phase windings, the first set of windings includes phases A, B, and C, and the second set of windings includes phases U, V, and W. The two sets of windings are spatially separated by 30 electrical degrees. Independent sliding mode observers SMO1 and SMO2 are designed for the two sets of windings respectively to estimate their back electromotive force.

[0050] For the first set of windings, The sliding surface is defined in the coordinate system as follows:

[0051]

[0052]

[0053] For the second set of windings, The sliding surface is defined in the coordinate system as follows:

[0054]

[0055] in, For the first set of windings in The actual current in the coordinate system This corresponds to the estimated current; For the second set of windings in The actual current in the coordinate system This is the corresponding estimated current.

[0056] To suppress the high-frequency chattering introduced by the traditional sign function switching law, this embodiment employs a continuous and smooth hyperbolic tangent function. As a switching function:

[0057]

[0058] in For sliding surface .

[0059] Based on the hyperbolic tangent switching function, the back electromotive force estimated by the dual sliding mode observer is expressed as:

[0060]

[0061] Where k is the gain of the sliding mode observer. Too large a gain will cause chattering, while too small a gain will result in slow convergence. k should be greater than the back EMF amplitude. In this embodiment, k is set to 500 based on the motor parameters and the desired convergence speed. The first set of winding back electromotive force estimated for the first sliding mode observer SMO1 The second set of winding back electromotive force is estimated for the second sliding mode observer SMO2.

[0062] Because the two windings of the dual three-phase motor are spatially 30 electrical degrees out of phase, there is a 30-degree phase difference between their back electromotive forces (EMFs). To eliminate this phase difference and align the two back EMF signals in phase, a coordinate rotation transformation is required for the back EMF of the second winding. Rotate counterclockwise by 30 degrees to obtain the back electromotive force after rotation.

[0063]

[0064] In the rotation matrix After rotation transformation, Phase alignment is performed to prepare for subsequent rotor position estimation. The back electromotive force obtained in this step will be used in step 4 for the phase-locked loop to extract position information.

[0065] Step 2: Propose a fault diagnosis method based on the comparison of controller output voltage amplitude:

[0066] The fault diagnosis method proposed in this embodiment is based on the reference voltage output by the controller. In the vector control system of a dual three-phase permanent magnet synchronous motor, the reference voltages of the two windings are obtained through the speed controller and the current controller. (First winding) and (Second winding) is used to drive the six-phase inverter to generate PWM signals.

[0067] Similar to the back electromotive force treatment, since the two windings are spatially separated by 30 degrees, there is also a 30-degree phase difference between the two reference voltages. To accurately compare the amplitudes of the two reference voltages, a coordinate rotation transformation is needed for the reference voltage of the second winding. Rotate counterclockwise by 30 degrees to obtain the rotated reference voltage. :

[0068]

[0069] like Figure 3 As shown, Figure 3 (a) shows the reference voltage waveforms of the two windings before the rotational transformation. between, There is a significant phase difference between them. Figure 3 (b) shows the reference voltage waveform after rotation transformation. The waveforms completely overlap. The waveforms also completely overlap, verifying that the coordinate rotation transformation can effectively eliminate the phase difference between the reference voltages of the two windings, making them consistent in both amplitude and phase.

[0070] Under normal operating conditions, the reference voltage amplitudes of the two windings are similar. When a single-phase open-circuit fault occurs in one of the windings, the current in that winding cannot change according to the command value. To reduce error, the current controller will continuously increase the output voltage, resulting in a significant increase in the reference voltage amplitude corresponding to the faulty winding. Based on this phenomenon, the fault diagnosis criteria are set as follows:

[0071]

[0072] in This is the diagnostic trigger threshold. When the above conditions are met, a single-phase open-circuit fault is determined in the system, and the winding where the fault occurs can be identified based on which set of reference voltage amplitude increases: If or If the first winding (phases A, B, and C) is faulty, the fault flag F is set to 1; if or If a fault occurs, the second winding (U, V, W phases) is determined to be faulty, and the fault flag F is set to 2. If there is no fault, the fault flag F is set to 0. Furthermore, since directly comparing instantaneous absolute values ​​can cause frequent jumps due to the zero-crossing points of the AC waveform, the amplitude of the reference voltage used here is the amplitude after low-pass filtering or its effective value.

[0073] Diagnostic threshold The value of the threshold needs to balance diagnostic sensitivity and anti-interference capability. If the threshold is too small, false diagnoses may occur due to voltage fluctuations; if the threshold is too large, the diagnostic response speed will decrease. In this embodiment, the threshold... The voltage setting is 15% to 25% of the rated voltage amplitude, and the specific value can be adjusted according to the actual working conditions.

[0074] like Figure 2 The diagram illustrates the active switching strategy. The system first checks if the fault flag F is 0. If F equals 0, it indicates the system is fault-free or has not yet detected a fault. In this case, fault diagnosis logic is executed, monitoring the amplitude difference between the two sets of winding reference voltages in real time. If the diagnostic conditions are met, the fault flag F is set to 1 or 2 based on the voltage amplitude comparison result, and the observer switching operation is performed. If F is not equal to 0, it indicates the system has detected a fault and completed the switching. In this case, the fault diagnosis step is skipped, and the output of the currently selected observer is used directly to avoid erroneous switching due to signal fluctuations.

[0075] like Figure 4 The results of the fault diagnosis experiment are shown. Figure 4 (a) illustrates the diagnostic process for an open-circuit fault in phase C. Before the fault occurred, the phase C current... It is a normal sine wave with an amplitude of approximately 2 amperes, and the reference voltage for both windings. , , , The amplitudes are similar, and the fault flag is 0. At approximately 90 milliseconds, an open-circuit fault occurs in phase C. It rapidly decays to near zero. Almost simultaneously, the reference voltage of the first winding... The amplitude increased significantly, while the reference voltage of the second winding... and The overall performance remains largely unchanged. When the fault diagnosis logic detects a voltage amplitude difference exceeding the threshold, it immediately changes the fault flag from 0 to 1, accurately identifying a fault in the first winding. The diagnostic response time is less than 50 milliseconds.

[0076] Figure 4 (b) illustrates the diagnostic process for an open-circuit fault in phase U. Before the fault occurred, the phase U current... The waveform is a normal sine wave with an amplitude of approximately 2 amperes. The reference voltage amplitudes of the two windings are similar, and the fault flag is 0. At approximately 110 milliseconds, an open-circuit fault occurs in phase U. It rapidly decays to near zero. Almost simultaneously, the reference voltage of the second winding... The amplitude increased significantly, while the reference voltage of the first winding... and The overall performance remains largely unchanged. When the fault diagnosis logic detects that the voltage amplitude difference exceeds the threshold, it immediately changes the fault flag from 0 to 2, accurately identifying a fault in the second winding, with a diagnostic response time of less than 50 milliseconds.

[0077] The above experimental results verify that the fault diagnosis method based on reference voltage amplitude comparison proposed in this invention can detect single-phase open circuit faults in real time and accurately identify the winding where the fault is located, with a diagnosis accuracy of 100% and strong anti-interference ability in the diagnosis process.

[0078] Step 3: Implement active switching control of the sliding mode observer:

[0079] Based on the fault flag F output by the fault diagnosis module in step 2, the active switching module autonomously selects which set of sliding mode observer's back EMF output to use. The switching rules are as follows:

[0080] When F equals 0, it indicates that the system is fault-free, and the back electromotive force of the first sliding mode observer SMO1 is used by default. , .

[0081] When F equals 1, it indicates a fault in the first winding, and the system immediately switches to the second sliding mode observer SMO2, using the rotated back electromotive force. , .

[0082] When F equals 2, it indicates a fault in the second winding, and the back electromotive force of the first sliding mode observer SMO1 is used. , .

[0083] To avoid erroneous switching due to signal fluctuations or interference, after detecting a fault and completing the observer switch, the system locks the state of the fault flag F and refrains from performing fault diagnosis for a preset period of time. This state-locking mechanism ensures that the system operates stably in fault-tolerant mode after a fault, avoiding position estimation jumps and system instability that may be caused by repeated switching.

[0084] like Figure 2 The active switching strategy flow shown indicates that when the fault flag F is not 0, the system skips the fault diagnosis step and enters the observer output selection logic. Based on the value of F, the corresponding back EMF signal is selected, thereby achieving fast and stable fault-tolerant switching.

[0085] Step 4: Extract rotor position and speed using a phase-locked loop:

[0086] The back EMF signal selected by the active switching module in step 3 is input into the phase-locked loop (PLL) module, and the rotor electrical angle and electrical angular velocity are extracted through the PLL algorithm. This embodiment uses a PLL algorithm based on proportional-integral control; the specific calculation steps are as follows:

[0087] set up and Let be the back EMF signal selected at time k, where the subscript x represents the winding number (1 or 2), and the superscript y indicates whether a rotational transformation has occurred (no superscript or R). Based on the relationship between back EMF and rotor position, the position estimation error is defined as:

[0088] (13)

[0089] in The rotor electrical angle estimated at time k is the rotor position.

[0090] The position error is adjusted using a proportional-integral controller, and the estimated electrical angular velocity is:

[0091] (14)

[0092] (15)

[0093] in and These are the proportional coefficient and the integral coefficient, respectively, and s is the intermediate accumulation term. This is the estimated electric angular velocity. In this embodiment, The value is 100. The value is 50.

[0094] Based on the estimated electrical angular velocity, the rotor electrical angle is calculated by integration:

[0095] (16)

[0096] in The sampling period is 10 kHz in this embodiment. It equals 100 microseconds.

[0097] Final estimated rotor electrical angle and electric angular velocity The output is fed to the speed controller and current controller for closed-loop vector control of the motor, enabling sensorless operation.

[0098] like Figure 5 The results of a comparative experiment on a single-phase open-circuit fault under sensorless closed-loop control are shown. The experimental conditions were: a dual three-phase permanent magnet synchronous motor with a rated power of 5 kW, a rated speed of 3000 rpm, 3 pole pairs, a stator resistance of 0.5 Ω, a dq-axis inductance of 5 mH, and a permanent magnet flux linkage of 0.15 Wb. The motor operated at a given speed of 1000 rpm and a load torque of 5 N·m. An open-circuit fault was injected into phase C after approximately 3 seconds of operation.

[0099] Figure 5 (a) shows the control results using the traditional single sliding mode observer method. The C-phase current was [data missing] before the fault occurred. The actual rotational speed is a normal sine wave. (Blue curve) and estimated rotational speed (Purple curve) remains stable at 1000 rpm, actual rotor electrical angle (Green curve) and estimated electrical angle (Orange curves) basically overlap, speed estimation error and angle estimation error All fluctuations were within a relatively small range. At approximately 3 seconds, an open-circuit fault occurred in phase C. It rapidly decays to 0. Because traditional sliding mode observers are based on symmetrical model designs, the asymmetrical operation of the motor windings after a fault leads to model mismatch, resulting in severe distortion in the back EMF estimation. Estimated speed. A sharp drop occurred, from 1000 rpm to approximately 850 rpm, and continued to oscillate, deviating significantly from the actual speed. Speed ​​estimation error. Peak speed is nearly 60 revolutions per minute, angle estimation error The peak value reached approximately 0.5 radians. The recovery time after the fault exceeded 3 seconds, and the system continued to oscillate, failing to fully recover to a stable state, resulting in a severe deterioration in system control performance.

[0100] Figure 5 (b) Demonstrates the control results using the dual sliding mode observer active switching strategy proposed in this invention. Experimental conditions and Figure 5 (a) Completely identical. The system was operating normally before the fault occurred. At approximately 3 seconds, an open-circuit fault occurred in phase C. The back EMF rapidly decays to 0. The fault diagnosis module detects the first winding fault within approximately 50 milliseconds of the fault occurring, and the active switching module immediately switches the back EMF signal source from SMO1 to SMO2. Since SMO2 only uses the healthy second winding (U, V, W phases) for observation, the influence of the faulty winding on the model is avoided, and the back EMF estimation remains accurate. Estimated rotational speed. There was only a brief dip, from 1000 rpm to about 880 rpm, before it quickly and smoothly recovered and closely tracked the actual speed. After recovery, there was no oscillation, and the speed estimation error was reduced. The peak speed is approximately 40 revolutions per minute, and the angle estimation error is... The peak value is approximately 0.4 radians. The recovery time after a fault is less than 1 second, the system quickly returns to a stable state, the position estimation is continuous and smooth, without jumps or oscillations, and the control performance is significantly better than traditional methods.

[0101] contrast Figure 5 (a) and Figure 5 (b) It can be seen that the dual sliding mode observer active switching method proposed in this invention significantly improves the fault-tolerant control performance under single-phase open-circuit faults compared with the traditional single sliding mode observer method: the peak value of the speed estimation error is reduced from nearly 60 revolutions per minute to about 40 revolutions per minute, an improvement of 33%; the peak value of the angle estimation error is reduced from about 0.5 radians to about 0.4 radians, an improvement of 20%; most importantly, the recovery time after the fault is shortened from more than 3 seconds to less than 1 second, an improvement of more than 66%; the position estimation continuity is good, with no jumps or oscillations, and the system stability is significantly enhanced.

[0102] Example 2:

[0103] This embodiment describes the technical solution of the present invention from the perspective of system architecture. For example... Figure 1 As shown, this embodiment provides a sensorless control system for a dual three-phase permanent magnet synchronous motor, including the following functional modules:

[0104] First sliding mode observer module SMO1: configured to receive current signals from the first set of three-phase windings (phase A, phase B, and phase C). , and reference voltage signal , Estimate the back electromotive force of the first winding based on sliding mode control theory , Internally, this module defines the sliding mode surface as the difference between the actual and estimated current values, uses the hyperbolic tangent function as the switching function to suppress high-frequency chattering, and sets the sliding mode gain k according to the motor parameters and the desired convergence speed.

[0105] Second sliding mode observer module SMO2: configured to receive current signals from the second set of three-phase windings (U phase, V phase, W phase). , and reference voltage signal , Estimate the back electromotive force of the second winding based on sliding mode control theory , The design of this module is similar to that of SMO1, also using the hyperbolic tangent function as the switching function, and the sliding mode gain k is consistent with that of SMO1.

[0106] Coordinate rotation transformation module: configured as a reference voltage for the second set of windings. , and back electromotive force , A coordinate rotation transformation is performed. Since the two windings of the dual three-phase motor are 30 electrical degrees out of phase, this module uses a rotation matrix to rotate the signal of the second winding counterclockwise by 30 degrees, obtaining the rotated reference voltage. , and back electromotive force , This aligns the signal of the first winding with the signal of the second winding in phase.

[0107] Fault diagnosis module: configured to receive reference voltages from both sets of windings. , and , The system compares the amplitude differences in real time. When the amplitude difference of the reference voltages on the α-axis of the two windings exceeds a preset threshold Th, or the amplitude difference of the reference voltages on the β-axis exceeds a preset threshold Th, a single-phase open-circuit fault is determined to have occurred in the system. Further, the system identifies the winding where the fault occurs based on which set of reference voltages shows an increased amplitude, and outputs a fault flag F. F equals 0 when there is no fault, F equals 1 when the first winding is faulty, and F equals 2 when the second winding is faulty. The specific value of the preset threshold Th can be adjusted according to actual operating conditions to balance diagnostic sensitivity and anti-interference capability.

[0108] Active switching module: Configured to receive the fault flag F and the back EMF outputs of two sets of sliding mode observer modules, and autonomously select which observer's back EMF signal to use based on the value of F. When F equals 0, the back EMF of the first sliding mode observer SMO1 is used. , When F equals 1, switch to the back electromotive force of the second sliding mode observer SMO2. , When F equals 2, the back electromotive force of the first sliding mode observer SMO1 is used. , The module is also configured to lock the state of the fault flag F after detecting a fault and completing the switchover, and not perform fault diagnosis operations for a preset period of time to avoid erroneous switching caused by signal fluctuations and ensure stable system operation.

[0109] Phase-locked loop (PLL) module: Configured to receive the back EMF signal output by the active switching module and extract the rotor electrical angle using a PLL algorithm. and electric angular velocity This module employs a phase-locked loop (PLL) algorithm based on proportional-integral (PI) control. It calculates the position estimation error based on the relationship between the back electromotive force and the rotor position, adjusts the error using the PLI controller, estimates the electrical angular velocity, and calculates the rotor electrical angle through integration. The estimated rotor position and speed information are output to the speed controller and current controller for closed-loop vector control of the motor, enabling sensorless operation.

[0110] The aforementioned functional modules interact and work collaboratively via a data bus or shared memory to form a complete sensorless control system for dual three-phase permanent magnet synchronous motors. This system can be implemented on embedded control platforms such as digital signal processors (DSPs) or field-programmable gate arrays (FPGAs). Each functional module is implemented through software programming, and the modular design facilitates system development, debugging, and maintenance.

[0111] This system is particularly suitable for drive systems of electric vertical takeoff and landing (eVTOL) aircraft or unmanned aerial vehicles (UAVs). In applications such as aviation propulsion, the motor operates in complex environments. Environmental factors such as high speed, extreme temperatures, and strong vibrations can lead to position sensor failure or open-circuit faults in inverter power devices. This system, through a dual sliding mode observer architecture and active switching strategy, ensures continuous and stable motor operation even after position sensor failure or single-phase open-circuit faults occur. This significantly improves the safety and reliability of the flight system, meeting the demand for highly reliable drive systems in low-altitude economic sectors.

[0112] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A dual sliding mode observer active switching method for a dual three-phase permanent magnet synchronous motor under a single-phase open-circuit fault, characterized in that: include: For the two sets of windings of the dual three-phase permanent magnet synchronous motor, a first sliding mode observer and a second sliding mode observer are constructed respectively to estimate the back electromotive force of the two sets of windings. The reference voltage of the second winding is subjected to coordinate rotation transformation so that the reference voltages of the two windings are aligned in phase. The amplitudes of the phase-aligned reference voltages output by the two winding controllers are compared. When the amplitude difference exceeds a preset threshold, a fault is determined to have occurred, and the winding where the fault is located is identified based on the amplitude comparison results. The fault diagnosis is based on the following expression: in To diagnose the trigger threshold, and The reference voltage for the first winding is at shaft and Components on the axis, and The reference voltage of the second winding after rotational transformation is in shaft and Components on the axis, For fault identification; Based on the fault diagnosis results, the system automatically switches to the sliding mode observer corresponding to the healthy winding and locks the fault status to avoid accidental switching. The rotor position and speed are extracted from the back electromotive force output by the selected sliding mode observer using a phase-locked loop.

2. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 1, characterized in that: Both the first and second sliding mode observers employ the hyperbolic tangent function. As a switching function, it suppresses high-frequency chattering: in It is the sliding surface of the winding.

3. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 2, characterized in that: The sliding surface of the first and second sliding mode observers is defined as the difference between the actual current value and the estimated current value.

4. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 1, characterized in that: The rotation angle of the coordinate rotation transformation is 30 degrees, which eliminates the inherent phase difference between the two sets of windings of the dual three-phase motor.

5. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 1, characterized in that: The fault diagnosis criteria are as follows: both sets of windings are in The difference in reference voltage amplitude on the shaft exceeds the preset threshold, or the two sets of windings are... If the difference in the amplitude of the reference voltage on the shaft exceeds the preset threshold, a fault is identified.

6. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 1, characterized in that: According to the fault label Perform autonomous switching. A value of 0 indicates no fault. A value of 1 indicates a fault in the first winding. A value of 2 indicates a fault in the second winding; the rule for autonomous switching is: when... When the back electromotive force of the first sliding mode observer is equal to 0, When the back electromotive force of the second sliding mode observer is equal to 1, when When the value is equal to 2, the back electromotive force of the first sliding mode observer is used.

7. The active switching method of dual sliding mode observers under single-phase open-circuit fault of a dual three-phase permanent magnet synchronous motor according to claim 1, characterized in that: The term "lock fault state" refers to maintaining the fault flag in the fault state after a fault is detected and a switch is completed, and not performing fault diagnosis operations for a preset period of time to avoid erroneous switching caused by signal fluctuations.

8. A sensorless control system for a dual three-phase permanent magnet synchronous motor based on the method of any one of claims 1 to 7, characterized in that: include: The first sliding mode observer module and the second sliding mode observer module are respectively configured to estimate the back electromotive force of the two sets of windings of the dual three-phase permanent magnet synchronous motor; The coordinate rotation transformation module is configured to perform coordinate rotation transformation on the reference voltage of the second set of windings so that the reference voltages of the two sets of windings are aligned in phase. The fault diagnosis module is configured to compare the amplitude of the phase-aligned reference voltage output from the two sets of winding controllers. When the amplitude difference exceeds a preset threshold, a fault is determined to have occurred, and the winding where the fault is located is identified. The active switching module is configured to automatically switch to the sliding mode observer corresponding to the healthy winding based on the fault diagnosis results and lock the fault status. The phase-locked loop module is configured to extract rotor position and speed from the back electromotive force output by the selected sliding mode observer.

Images