A zero sequence rejection type minimum copper loss model-free fault-tolerant control method and system

Abstract

The application discloses a zero-sequence suppression type minimum copper loss model-free fault-tolerant control method and system. After phase failure occurs, the method can effectively suppress zero-sequence current (ZSC) under the premise of minimum copper loss, expand the safety belt load range, and maintain strong robustness under serious parameter mismatch. The method first proposes a phase domain ZSS-MCL reference current analytical generation algorithm. The algorithm can directly output asymmetric reference current without complex offline optimization, suppress ZSC from the source, and maximize torque under the safety thermal limit. Secondly, the MFPC based on the super-local model is constructed, the state observer is used to compensate the lumped disturbance in real time, the high dynamic tracking of the asymmetric current is realized, and the dependence of the traditional predictive control on accurate internal parameters is eliminated. Compared with the traditional fault-tolerant control method, the method discards the complex offline optimization and multiple subspace redundant calculation, directly provides the analytical solution of the reference current, suppresses the ZSC from the source, and maximizes the load capacity of the motor.

Description

Technical Field

[0001] This invention belongs to the field of motor control, specifically relating to the drive and fault-tolerant control of multiphase open-winding permanent magnet synchronous motors, and particularly to a model-free predictive current control method for a common DC bus type six-phase open-winding permanent magnet synchronous motor under phase loss fault. Background Technology

[0002] With the advancement of power electronics technology and its diverse applications in motor drive systems, multiphase motors have been extensively studied. In comparison, multiphase open-winding permanent magnet synchronous motors (PMSMs) have rapidly become a research hotspot in recent years due to their advantages such as higher output power, lower power per phase, and stronger fault tolerance. Among these advantages, the fault-tolerant operation capability of multiphase open-winding PMSMs plays a crucial role in system reliability evaluation. After a phase loss fault occurs in the motor drive system, existing fault-tolerant control research often requires complex subspace reconstruction after the phase loss, increasing the complexity of the control system. Furthermore, there is limited existing research on fault-tolerant control of six-phase open-winding PMSMs, and traditional fault-tolerant control methods are not ideal in phase loss control of six-phase open-winding PMSMs. Summary of the Invention

[0003] This invention provides a zero-sequence suppression type minimum copper loss model-free fault-tolerant control method, which can effectively suppress zero-sequence current under the premise of minimum copper loss; it can directly output asymmetric reference current without complex offline optimization, suppressing ZSC from the source and maximizing torque under safe thermal limits.

[0004] This method provides model-free fault-tolerant control for a six-phase open-winding permanent magnet synchronous motor under a phase-loss fault, and includes the following steps:

[0005] Step S1: After a phase loss fault occurs in a six-phase open-winding permanent magnet synchronous motor, an independent orthogonal subspace is reconstructed. The three reconstructed axes are orthogonal to each other and are completely perpendicular to the α-β plane under normal operating conditions.

[0006] Step S2: Construct a vector space decoupling matrix as a dimension reduction matrix based on the constructed orthogonal subspace. The first two rows of this matrix are exactly the same as the first two rows of the 6x6 vector space decoupling matrix in the normal operating state of the six-phase open-winding permanent magnet synchronous motor before phase loss.

[0007] Step S3: Determine the objective function for stator copper loss of the six-phase open-winding permanent magnet synchronous motor under phase loss condition; Step S4: Construct constraint equations, including:

[0008] 1. The constraints are: 1) The reconstructed α-β axis current is completely consistent with the reference value before the phase loss; 2) The zero-sequence current of both windings is zero.

[0009] 2. Substitute the constraints into the dimension reduction matrix to obtain the constraint equations;

[0010] Step S5: Analyze the constraint equations to obtain the phase current and variables after phase loss. The function;

[0011] Step S6: Substitute the function obtained in step S5 into the stator copper loss objective function and solve it. Substitute the solution into the function obtained in S5 to obtain the fault-tolerant phase current analytical formula that can maintain healthy electromagnetic torque, suppress zero-sequence circulating current in the winding, and ensure the lowest heat generation of the whole machine.

[0012] Step S7: Based on the analytical formula of the fault-tolerant phase current, the control current under the phase loss state can be obtained to achieve model-free fault-tolerant control.

[0013] Furthermore, under phase loss conditions, the maximum phase current is less than or equal to the upper limit of the rated current. upper limit of rated current The following conditions must be met:

[0014]

[0015] Among the conditions: This is the reference current amplitude.

[0016] Furthermore, under phase-loss conditions, this method determines the next switching state combination of inverter unit 1 and inverter unit 2 at both ends of the six-phase open-winding permanent magnet synchronous motor through the following steps:

[0017] State acquisition and modeling steps: measure the current current, speed and rotor position in real time, use the voltage equation of permanent magnet synchronous motor to discretize and construct a prediction model, and establish the mapping relationship between voltage and current at the next moment;

[0018] Vector enumeration and prediction steps: The basic voltage vectors generated by inverter unit 1 and inverter unit 2 are substituted into the prediction model to obtain the predicted value at time k+1 as the candidate vector.

[0019] Cost function evaluation steps: Construct a cost function that minimizes the current tracking error, which tracks the sum of squares of the differences between the predicted and reference values ​​obtained from the vector enumeration and prediction steps; the cost function that minimizes the cost function is as follows:

[0020]

[0021] in: This is the predicted value at time k+1 of the healthy phase of inverter unit 1. This is the predicted value at time k+1 of the healthy phase of inverter unit 2. , subscript Health-related factors;

[0022] Optimal decision and execution steps: Compare the cost function values ​​of all candidate vectors, select the basic voltage vector corresponding to the minimum value, and directly use it as a switching signal to drive the inverter without the need for PWM modulation.

[0023] Furthermore, the phase current discrete linear extended state observer in the cost function of the effective voltage vector set is obtained, and the discrete linear extended state observer uses lumped perturbation. Independent state-space variables The state equation of the discrete linear extended state observer is:

[0024]

[0025] in: This is the measured current value. The error between the current measurement values ​​at time k+1 and time k; For aggregated disturbances, Let be the lumped disturbance error between time k+1 and time k; Ts is the control period. It is the gain constant. The effective voltage vector applied to inverter unit 1 and inverter unit 2;

[0026] The eigenvalues ​​of the discrete linear extended state observer are configured in At this point, the gain is and ; is the closed-loop bandwidth of the discrete linear extended state observer.

[0027] This invention also provides a zero-sequence suppression type minimum copper loss model-free fault-tolerant control system, which performs model-free fault-tolerant control under a phase loss fault in a six-phase open-winding permanent magnet synchronous motor, and is configured to include:

[0028] The first module is used to acquire the operating status of the six-phase open-winding permanent magnet synchronous motor in real time and determine whether a phase loss fault has occurred.

[0029] Second module: After a phase loss fault occurs, a computer program is run to implement the control method described in any one of claims 1-4.

[0030] The beneficial effects of this invention are: after a phase loss fault, it can effectively suppress zero-sequence current (ZSC) with minimal copper loss, expand the safe load range, and maintain strong robustness under severe parameter mismatch. First, this method proposes a phase-domain ZSS-MCL reference current analytical generation algorithm. This algorithm can directly output the asymmetric reference current without complex offline optimization, suppressing ZSC at its source and maximizing torque under safe thermal limits. Second, it constructs an MFPC based on a hyperlocal model, using a state observer to compensate for lumped disturbances in real time, achieving high dynamic tracking of the asymmetric current and eliminating the dependence of traditional predictive control on precise internal parameters. Compared to traditional fault-tolerant control methods, this method abandons complex offline optimization and multi-subspace redundant calculations, directly providing an analytical solution for the reference current, suppressing ZSC at its source while maximizing the motor's load capacity. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is the topology diagram of a six-phase open-winding permanent magnet synchronous motor after phase F is disconnected.

[0033] Figure 2 The parabola representing the objective function of copper loss and its global minimum heating point;

[0034] Figure 3 The current waveforms of the method proposed in this invention are shown under normal operation and after a phase loss fault. Figure 3 (a) The waveforms of the entire process were recorded. Figure 3 (b)-(e) provide transient magnified views of each indicator. Detailed Implementation

[0035] This section describes the invention more fully with reference to the accompanying drawings, in which illustrative embodiments of the invention are shown. However, the invention is also embodied in many different forms and should not be construed as limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure is thorough and complete, and to fully convey the scope of the invention to those skilled in the art.

[0036] Unless defined to the contrary, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It will be further understood that terms such as those defined in common dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant field, and will not be interpreted in an idealized or highly formal sense unless specifically defined herein.

[0037] This invention focuses on a six-phase open-winding permanent magnet synchronous motor (PMSM) and proposes a zero-sequence suppression-type minimum copper loss model-free predictive control method for a six-phase open-winding PMSM under phase-loss fault conditions. First, it proposes an analytical algorithm for generating a phase-domain zero-sequence suppression-type minimum copper loss reference current. This algorithm can directly output the asymmetric reference current without complex offline optimization, suppressing zero-sequence current at its source and maximizing torque under safe thermal limits. Second, it constructs a model-free predictive control framework based on a hyperlocal model, utilizing a state observer to compensate for lumped disturbances in real time, achieving high dynamic tracking of the asymmetric current and eliminating the dependence of traditional predictive control on precise internal parameters (referring to electrical parameters such as motor resistance and inductance). Compared to traditional fault-tolerant control methods, this method abandons complex offline optimization and multi-subspace redundant calculations, directly providing an analytical solution for the reference current, maximizing the motor's load-carrying capacity while suppressing zero-sequence current at its source.

[0038] Terminology Definition

[0039] A healthy phase refers to a phase that has not been interrupted and continues to provide power to a six-phase open-winding permanent magnet synchronous motor.

[0040] The health phase factor refers to a specific unbroken phase sequence. For example, if the six phases are A to F and phase E is broken, then the health phase factor refers to phase sequences A, B, C, D, and F.

[0041] The system described in this article refers to a six-phase open-winding permanent magnet synchronous motor system, which includes at least a six-phase open-winding permanent magnet synchronous motor, inverter unit 1, and inverter unit 2.

[0042] Two sets of windings refer to the coil windings of an inverter unit under normal operating conditions, such as... Figure 1 As shown, when there is no phase loss, the two sets of windings of inverter unit 1 and inverter unit 2 are ABC and DEF respectively. When a phase is lost, the two sets of windings are the remaining phases. For example, after phase F is lost, the two sets of windings are ABC and DE.

[0043] This embodiment describes in detail the control method provided for the F-phase disconnection of a six-phase open-winding permanent magnet synchronous motor. However, those skilled in the art should understand that this control method is not limited to the F-phase disconnection situation, but can control the disconnection of any phase.

[0044] Figure 1 This is the topology diagram of a six-phase open-winding permanent magnet synchronous motor after phase F is lost. Inverter unit 1 and inverter unit 2 are both two-level inverters, U... dc This is the DC bus voltage. A1 ~S F1 and S A2 ~S F2 These are the upper transistors of each bridge arm of inverter unit 1 and inverter unit 2, respectively. ’ A1 ~S ’ F1 and S ’ A2 ~S ’ F2 These are the lower transistors of each bridge arm of inverter unit 1 and inverter unit 2, i sA ~ i sF These are the phase currents of SOEW-PMSM, respectively.

[0045] Figure 2 The parabola of the copper loss objective function and its global minimum heating point, corresponding to process S2, demonstrate the analytical solution process under the constraints of minimum copper loss and zero-sequence current suppression.

[0046] When phase F is lost, the following controls are performed:

[0047] S1: Mathematical modeling of a six-phase open-winding permanent magnet synchronous motor after a phase loss fault:

[0048] S11: Under healthy conditions, the phase variables of a six-phase open-winding permanent magnet synchronous motor can be mapped to three mutually orthogonal two-dimensional subspaces using the classic 6×6 vector space decoupling matrix T6. Employing amplitude invariance constraints, the standard T6 is as follows:

[0049] (1)

[0050] In the formula, rows 1 and 2 are the fundamental subspace (α-β), rows 3 and 4 are the harmonic subspace (z1-z2), and rows 5 and 6 are the zero-order subspace (o1-o2). Each column corresponds to the phase sequence A, B, C, D, E, F, respectively.

[0051] S12: When a phase-loss fault occurs in phase F, the physical constraints of the system force the current in phase F to be zero. This means that the system's degrees of freedom are reduced from 6 to 5. At this point, if the 6th column of T6 is directly removed, a 6×5 truncation matrix is ​​obtained, and the inner product of its row vectors will no longer be zero. Specifically, the inner product of the β-axis vector and the o2-axis vector is:

[0052] (2)

[0053] Equation (2) shows that after the phase is broken, the originally decoupled fundamental axis and the zero-sequence component undergo severe deep cross-coupling. If the coordinate system is not reconstructed, the torque output will be directly and severely disturbed by the zero-sequence current.

[0054] S13: To re-establish independent orthogonal subspaces after phase loss, this control method reconstructs a new 5×5 dimensionality-reduced matrix T5. The reconstruction of T5 must adhere to the following two rigorous mathematical principles:

[0055] (1) Principle of fundamental invariance: In order to ensure that the circular stator rotating magnetomotive force can continue to be generated after fault tolerance, the first two rows of the new matrix must be completely consistent with the first two rows of the original T6.

[0056] (2) Subspace Orthogonality Principle: The remaining three reconstruction axes must be mutually orthogonal and completely perpendicular to the α-β plane. Under normal operating conditions, the six-phase motor will decouple into three mutually orthogonal planes, namely the fundamental plane (α-β axis), the harmonic plane (z1-z2 axis), and the zero-sequence plane (o1-o2 axis). After a phase loss fault occurs, the original decoupling matrix becomes invalid, and the original coordinate axes, except for the α-β axis, are no longer valid. Therefore, it is necessary to recalculate the coordinate matrix that satisfies decoupling. Since the dimension-reduced matrix is ​​5×5, the three axes described here are the coordinate axes corresponding to the new decoupling matrix, except for the α-β axis. Traditional methods would plan the subspace based on these coordinate axes and then calculate the voltage vector, but this method does not require this and only uses it as a dimension reduction tool.

[0057] The analytical expression for the dimension reduction matrix T5 is as follows:

[0058] (3)

[0059] Through the aforementioned T5 matrix, the 5D system, after phase loss, regains physical decoupling. However, in traditional fault-tolerant strategies, the high-dimensional and asymmetric coefficients of T5 and its inverse matrix are directly embedded in the high-frequency control closed loop of the DSP, which imposes a disastrous computational burden on the multivariable microprocessor. This control method abandons this traditional approach. In this method, T5 serves only as an offline mathematical tool. In the subsequent S2, the first two rows (fundamental frequency torque constraint) and the last row (zero-sequence current suppression constraint) provided by T5 are directly used to explicitly derive the optimal current analytical solution in the global phase domain using the Lagrange extremum method. Once the analytical solution is established, T5 is completely removed from the underlying online controller code, freeing up a significant amount of computational space.

[0060] S2: Analytical solution for the zero-sequence suppression minimum copper loss reference current:

[0061] S21: Under the F-phase loss condition, the objective function of the system is to minimize the stator copper loss of the remaining five phases (equivalent to minimizing the sum of squares of currents):

[0062] (4)

[0063] In the formula: i A~E These are the currents of each phase after a phase loss fault.

[0064] S22: To maintain torque stability after fault tolerance and effectively suppress ZSC, the following two constraints must be met:

[0065] (1) Ensure that the fundamental magnetomotive force of the stator remains unchanged, that is, the reconstructed α-β axis current is completely consistent with the reference value before the phase loss.

[0066] (2) Force the zero-sequence current of the two windings to be zero.

[0067] Combining the above conditions with T5, we can obtain four equality constraint equations:

[0068] (5)

[0069] The above equation contains 5 unknown variables and 4 equality constraints. From the perspective of system degrees of freedom, only 1 control degree of freedom remains for copper loss optimization. This control method transforms the complex multi-constraint optimization into a single-degree-of-freedom unconstrained convex optimization through algebraic dimensionality reduction mapping.

[0070] S23: Analytical equation (5) can be used to determine the remaining 4-phase current i D~E Mapped to a unique independent variable i A The function is shown below.

[0071] (6)

[0072] Substituting equation (6) into the copper loss objective function (4), we can obtain

[0073] (7)

[0074] After simplification, we get:

[0075] (8)

[0076] The objective function degenerates into a single function about i. A To find the equation of the quadratic parabola, and in order to locate the globally lowest hot spot, we need to find the i-th equation of the parabola function. A By taking the first derivative of the first derivative and setting it to zero, we can capture the unique global minimum hot spot.

[0077] (9)

[0078] Solution:

[0079] (10)

[0080] S24: Substituting the obtained extreme points back into the phase mapping relationships, we finally obtain the analytical formula for the fault-tolerant phase current under the F-phase phase loss fault, which can maintain healthy electromagnetic torque, suppress zero-sequence circulating current in the winding, and ensure the lowest overall heat generation:

[0081] (11)

[0082] Under the derived zero-sequence suppression minimum copper loss architecture, the microprocessor only needs to obtain the reference current of the system control target α-β axis, and can directly generate the 5-phase optimal reference current with zero-sequence current suppression in the phase domain with extremely low computational cost.

[0083] S3: Thermal limit boundary calculation under phase loss fault;

[0084] S31: In practical engineering applications of fault-tolerant control, both the power semiconductor devices of the frequency converter and the stator windings of the motor are subject to strict physical thermal limits. That is, after an OPF (Optical Power Forward Failure) occurs, the peak phase current of the remaining healthy phases must never exceed the upper limit of the system's rated current. ,and , where k = (A ~ E). This hardware limitation directly determines the maximum average torque that the system can output under fault-tolerant conditions.

[0085] To accurately evaluate the load capacity of the proposed zero-sequence suppression minimum copper loss strategy, the analytical solution derived in Section S2 needs to be transformed into a sinusoidal amplitude-phase form. This control method uses i d Unity power factor control with a power factor of 0 = 0, i.e.

[0086] (12)

[0087] In the formula, I m Let θ be the amplitude of the fundamental current. e It is an electrical angle.

[0088] S32: Substituting equation (12) into equation (11) and extracting the fundamental amplitude of each phase current, we can obtain:

[0089] (13)

[0090] As can be seen from the above amplitude distribution pattern, after the forced implementation of zero-sequence suppression constraint, the current distribution of the two windings exhibits a significant imbalance. Specifically, the current amplitude of phases D and E decreases to 86.6% of the rated value, while phases B and C bear the greatest load pressure, with their peak current soaring to 1.803 times the reference given value.

[0091] S33: Therefore, phases B and C constitute the "thermal bottleneck" for the fault-tolerant operation of this system. To ensure that phases B and C do not burn out due to overcurrent, the analytical expression in equation (13) is transformed into a sinusoidal amplitude phase form, resulting in the following boundary constraints:

[0092] (14)

[0093] Because the electromagnetic torque of a surface-mounted six-phase open-winding permanent magnet synchronous motor is related to the fundamental current amplitude I... m The above inequality shows a strictly linear proportional relationship, indicating that under extremely harsh operating conditions with a phase-loss fault in phase F and a zero-sequence suppression-type minimum copper loss strategy, the maximum continuous available torque of the system will be rated to 55.4% of the rated torque in healthy condition to ensure absolute electrical safety. In traditional fault-tolerant strategies, equal-amplitude symmetrical constraints are often forcibly applied to suppress zero-sequence current, resulting in a limit torque that is often only 40%~45% and a low voltage utilization rate.

[0094] S4: Establish a hyperlocal model for the healthy phase subsystem and optimize the output:

[0095] S41: When a phase loss fault occurs, the internal electromagnetic dynamics of the six-phase open-winding permanent magnet synchronous motor become severely asymmetrical. To eliminate the complex dependence on the fault-tolerant physical parameters, this control method establishes a data-driven hyperlocal model for the healthy phase subsystem.

[0096] In the continuous time domain, the current dynamics of the 5-phase healthy subsystem are lumped and abstracted into the following first-order differential equation:

[0097] (15)

[0098] Among them, i h =[i A i B i C i D i E ] T The vector of the healthy phase current; u h The effective voltage vector applied to the two-terminal inverter; α is the proportional gain; F = [F] A ,F B ,F C ,F D ,F E ] T Defined as a generalized lumped disturbance, it encompasses all unknown nonlinear dynamics of the system.

[0099] S42: Accurate and high-bandwidth acquisition of the lumped disturbance F is a prerequisite for achieving model-free predictive tracking. To avoid the high-frequency noise amplification problem caused by traditional direct difference, this paper designs a Discrete Linear Extended State Observer (DLESO) with a "prediction-correction" closed-loop structure. By expanding F into independent state variables, DLESO provides excellent noise resistance and robust estimation.

[0100] make The generalized perturbation is extended into a new state-space variable. Then the continuous state-space model of equation (15) is reconstructed as:

[0101] (16)

[0102] Where h(t) represents the rate of change of the disturbance over time. In high-speed microcontrollers, due to the control period T s Since it is extremely short, it is reasonable to assume that the dynamics of the disturbance tend to be stable within one period, i.e., h(t)≈0.

[0103] S43: Discretize equation (16) to zero order to construct DLESO with a "prediction-correction" closed-loop structure. Using the optimal control input of the previous step and the current measurement value of the current step, the iterative equation of the observer is derived as follows:

[0104] Step 1: Error Calculation

[0105] (17)

[0106] Step 2: Status Correction

[0107] (18)

[0108] Step 3: State Prediction

[0109] (19)

[0110] Based on the discrete pole placement theory, to ensure the convergence of the error dynamic matrix, the eigenvalues ​​of DLESO are uniformly placed at 1-ω0T. s The observer gain is thus analytically determined as: and Where ω0 is the observer's closed-loop bandwidth. This configuration provides a rigorous mathematical guarantee of convergence for lumped perturbation estimation.

[0111] S44: High-fidelity perturbation estimation using DLESO output The control system is no longer dependent on the motor inductance matrix and flux linkage model. At this point, the future state evolution of the healthy phase current under any switching combination of the dual-ended inverter can be predicted.

[0112] For a dual inverter, the j-th alternative voltage vector can be expressed as:

[0113] (20)

[0114] Among them, U dc S1 is the DC bus voltage, and S2 is the difference between the switching states of inverter unit 1 and inverter unit 2 at both ends of the six-phase open-winding permanent magnet synchronous motor.

[0115] The phase current prediction model for the next control cycle is as follows:

[0116] (twenty one)

[0117] S45: To achieve accurate and highly dynamic tracking of the asymmetric optimal reference current under phase loss fault conditions, this patent constructs an extremely concise unweighted cost function, which only contains the sum of squared tracking errors of the healthy phase current:

[0118] (twenty two)

[0119] The controller directly selects the optimal switching state combination for the next cycle by minimizing the cost function (22) in the effective voltage vector set, thereby achieving optimal control of the phase loss system without the participation of motor model parameters.

[0120] Steps S44 and S45 can also be described as follows:

[0121] State acquisition and modeling: Real-time measurement of current current, speed and rotor position, and construction of prediction model (Equation 21) using the voltage equation of permanent magnet synchronous motor (PMSM) discretized by Euler method, to establish the mapping relationship between voltage and current at the next moment.

[0122] Vector enumeration and prediction: For the basic voltage vector generated by the inverter, substitute it into the prediction model to calculate the candidate current value at time k+1.

[0123] Cost function evaluation steps: Construct a cost function that minimizes the current tracking error, which tracks the sum of squares of the differences between the predicted and reference values ​​obtained from the vector enumeration and prediction steps. The cost function is minimized as follows:

[0124]

[0125] in: This is the predicted value at time k+1 of the healthy phase of inverter unit 1. This is the predicted value at time k+1 of the healthy phase of inverter unit 2. , subscript For health-related factors; as mentioned above, phase F is disconnected. The subscripts A, B, C, D, and E represent respectively Figure 1 Phases A, B, C, D, and E of inverter unit 1; The subscripts A, B, C, D, and E represent respectively Figure 1 Phases A, B, C, D, and E of inverter unit 2.

[0126] Optimal Decision and Execution: Compare the cost function values ​​of all candidate vectors, select the voltage vector corresponding to the minimum value, and directly generate a switching signal to drive the inverter without the need for traditional PWM modulation.

[0127] Figure 3 The diagram shows the current waveforms of the proposed method under normal operation and after a phase loss fault. The system initially operates at a steady state at 300 r / min under rated conditions (1600 N·m) to calibrate the continuous thermal limit current boundary (i.e., the peak value of the rated phase current). At t=1s, the system actively and smoothly reduces the load torque to 850 N·m (approximately 53.1% of full load, strictly within the theoretical derating boundary of 55.4% derived in this invention) and waits for the system to establish a new steady state. Subsequently, at t=1.5s, a phase loss fault in phase F is artificially triggered, and the fault-tolerant control method proposed in this invention is simultaneously activated.

[0128] At the moment of a phase loss fault, the control strategy proposed in this invention achieves an extremely fast feedforward response, with no significant drop in motor speed. Although the torque experiences a slight disturbance at the moment of phase loss, it immediately returns to stability within milliseconds, and steady-state torque pulsation is greatly suppressed.

[0129] Regarding current reconfiguration, faulty phase i f Upon resetting to zero, the healthy phase current is reconstructed into a smooth waveform with varying amplitudes but strong sinusoidal characteristics. It is particularly noteworthy that, to compensate for the torque loss caused by the phase loss, the amplitudes of the B and C phase currents expand significantly according to theoretical principles. However, since the 850 N·m load point strictly adheres to the 55.4% derating threshold established in this paper, the peak values ​​of the expanded B and C phase currents just touch but do not exceed the continuous thermal limit boundary defined under the initial rated operating conditions. Figure 3 As shown in (c). Meanwhile, thanks to the hard constraint of the first priority of this method, the zero-sequence current is clamped to near 0A; the small zero-sequence fluctuations after the phase loss fault are mainly due to the unavoidable inverter dead zone effect and high-frequency switching ripple, and their macroscopic destructiveness has been completely eliminated.

[0130] This experimental setup not only verifies that the proposed strategy possesses excellent fault-tolerant and smooth switching capabilities, but also rigorously demonstrates at the physical level that under the zero-sequence suppression-type minimum copper loss strategy, the system can achieve long-term safe steady-state operation with the theoretically derived optimal derating coefficient.

[0131] The above description uses the F-phase failure as an example, but it is not a limiting description. It should be understood that this method is applicable to... Figure 1 The diagram shows any one of phases A, B, C, D, E, or F that is disconnected. The variable subscripts are adjusted according to the disconnection situation. The above description is based on phase F being disconnected, so the variable subscripts are A, B, C, D, and E, representing phases A, B, C, D, and E, respectively. If phase D is disconnected, the variable subscripts are A, B, C, E, and F, representing phases A, B, C, E, and F, respectively.

[0132] This control method can be implemented through the following system:

[0133] A zero-sequence suppression type minimum copper loss model-free fault-tolerant control system is configured to include:

[0134] The first module is used to acquire the operating status of the six-phase open-winding permanent magnet synchronous motor in real time and determine whether a phase loss fault has occurred.

[0135] The second module: After a phase loss fault occurs, a computer program is run to implement this control method.

[0136] The computer program executed by the second module is stored internally or in a storage unit that is communicatively connected to the second module. The second module directly reads and executes the computer program stored in the storage unit.

[0137] The second module uses existing computer-readable media with storage and computer program execution capabilities, such as microcomputer systems including single-chip microcomputers.

[0138] The storage unit can be any existing hardware memory, software memory, or virtual memory capable of storing computer programs and having a communication interface.

[0139] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

[0140] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprising” and “including,” as used herein, specify the presence of the illustrated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof.

Claims

1. A zero-sequence suppression type minimum copper loss model-free fault-tolerant control method, characterized in that, This method provides model-free fault-tolerant control for a six-phase open-winding permanent magnet synchronous motor under a phase-loss fault, and includes the following steps: Step S1: After a phase loss fault occurs in a six-phase open-winding permanent magnet synchronous motor, an independent orthogonal subspace is reconstructed. The three reconstructed axes are orthogonal to each other and are completely perpendicular to the α-β plane under normal operating conditions. Step S2: Construct a vector space decoupling matrix as a dimension reduction matrix based on the constructed orthogonal subspace. The first two rows of this matrix are exactly the same as the first two rows of the 6x6 vector space decoupling matrix in the normal operating state of the six-phase open-winding permanent magnet synchronous motor before phase loss. Step S3: Determine the objective function for stator copper loss of a six-phase open-winding permanent magnet synchronous motor under phase loss conditions; Step S4: Construct constraint equations, including: (1). The constraints are: 1) The reconstructed α-β axis current is completely consistent with the reference value before the phase loss; 2) The zero-sequence current of both windings is zero. (2). Substitute the constraints into the dimension reduction matrix to obtain the constraint equations; Step S5: Analyze the constraint equations to obtain the phase current and variables after phase loss. The function; Step S6: Substitute the function obtained in step S5 into the stator copper loss objective function and solve it. Substitute the solution into the function obtained in S5 to obtain the fault-tolerant phase current analytical formula that can maintain healthy electromagnetic torque, suppress zero-sequence circulating current in the winding, and ensure the lowest heat generation of the whole machine. Step S7: Based on the analytical formula of the fault-tolerant phase current, the control current under the phase loss state can be obtained to achieve model-free fault-tolerant control.

2. The control method according to claim 1, characterized in that: Under phase loss conditions, the maximum phase current is less than or equal to the upper limit of the rated current. upper limit of rated current The following conditions must be met: ， Among the conditions: This is the reference current amplitude.

3. The control method according to claim 1, characterized in that: This method, under phase-loss conditions, determines the next switching state combination of inverter unit 1 and inverter unit 2 at both ends of a six-phase open-winding permanent magnet synchronous motor through the following steps: State acquisition and modeling steps: measure the current current, speed and rotor position in real time, use the voltage equation of permanent magnet synchronous motor to discretize and construct a prediction model, and establish the mapping relationship between voltage and current at the next moment; Vector enumeration and prediction steps: The basic voltage vectors generated by inverter unit 1 and inverter unit 2 are substituted into the prediction model to obtain the predicted value at time k+1 as the candidate vector. Cost function evaluation steps: Construct a cost function that minimizes the current tracking error, which tracks the sum of squares of the differences between the predicted and reference values ​​obtained from the vector enumeration and prediction steps; the cost function that minimizes the cost function is as follows: ， in: This is the predicted value at time k+1 of the healthy phase of inverter unit 1. This is the predicted value at time k+1 of the healthy phase of inverter unit 2. , subscript Health-related factors; Optimal decision and execution steps: Compare the cost function values ​​of all candidate vectors, select the basic voltage vector corresponding to the minimum value, and directly use it as a switching signal to drive the inverter without the need for PWM modulation.

4. The control method according to claim 3, characterized in that: The effective voltage vector is obtained by minimizing the phase current discrete linear extended state observer in the cost function, and the discrete linear extended state observer is obtained by lumped perturbation. Independent state-space variables The state equation of the discrete linear extended state observer is: ， in: This is the measured current value. The error between the current measurement values ​​at time k+1 and time k; For aggregated disturbances, Let be the lumped disturbance error between time k+1 and time k; Ts is the control period. It is the gain constant. The effective voltage vector applied to inverter unit 1 and inverter unit 2; The eigenvalues ​​of the discrete linear extended state observer are configured in At this point, the gain is and ; is the closed-loop bandwidth of the discrete linear extended state observer.

5. A zero-sequence suppression type minimum copper loss model-free fault-tolerant control system, characterized in that, This system performs model-free fault-tolerant control under a phase-loss fault in a six-phase open-winding permanent magnet synchronous motor, and is configured to include: The first module is used to acquire the operating status of the six-phase open-winding permanent magnet synchronous motor in real time and determine whether a phase loss fault has occurred. Second module: After a phase loss fault occurs, a computer program is run to implement the control method described in any one of claims 1-4.

6. The zero-sequence suppression type minimum copper loss model-free fault-tolerant control system according to claim 5, characterized in that, The computer program running on the second module is stored internally or in a storage unit that is communicatively connected to the second module.

Images