Continuous modulation fault-tolerant double-random low-vibration-noise control method for five-phase permanent magnet motor
Patent Information
- Application Number
- GB2024005341
- Authority / Receiving Office
- GB · GB
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-02-17
- Filing Date
- 2023-05-12
- Publication Date
- 2024-11-06
AI Technical Summary
The existing technology fails to effectively suppress high-frequency vibration and noise during fault-tolerant operation of multi-phase motors, and the low-vibration and noise modulation strategy is not suitable for fault-tolerant operation, resulting in the inability to establish a collaborative working mode of fault-tolerant control and low-vibration and noise modulation.
The continuous modulation fault-tolerant dual-random low vibration and noise control method of the five-phase permanent magnet motor is adopted. By establishing the reduced-order decoupling matrix and the basic fault-tolerant virtual voltage vector, the fault-tolerant control strategy of continuous modulation is realized, and combined with the random zero vector-variable delay SVPWM, Disperse the switching frequency and its harmonics at integer multiples into the specified frequency domain to reduce the high-frequency harmonic amplitude of the phase current.
Significantly improves the fault-tolerant operating performance, suppresses the sideband harmonics on both sides of the fundamental wave, overcomes the regulator parameter mismatch problem caused by changes in the traditional random switching frequency SVPWM sampling frequency, improves the stability of the system, and builds a fault-tolerant Control and low-noise modulation work together.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
A five-phase permanent magnet motor continuous modulation fault-tolerant double random low-noise control method Technical Field The present invention relates to the field of multi-phase motor fault-tolerant control, the field of multi-phase motor harmonic suppression, and the field of motor low-noise control, and in particular to a control method for realizing dual random SVPWM low-noise of a five-phase permanent magnet synchronous motor under continuous modulation fault-tolerant control. Background Art Compared with traditional three-phase motors, multi-phase motor drive systems have the advantages of greater power output, smaller torque pulsation, higher bus voltage utilization, more control freedom, and better fault tolerance. Therefore, they have good application prospects in electric aircraft, ship propulsion, submarines and other fields. The fault-tolerant performance of multi-phase motors improves the reliability of multi-phase motors, and fault-tolerant control strategies for multi-phase motors emerge in an endless stream. The principle of unchanged magnetic motive force before and after a fault provides a theoretical basis for the fault-tolerant control of multi-phase motors. This principle achieves fault-tolerant operation of motors with faults by adjusting the amplitude and phase of the normal phase current. Based on this principle, domestic and foreign scholars have derived an analytical expression for the fault-tolerant current in combination with different fault-tolerant control objectives, and achieved motor fault-tolerant operation by changing the current set value and designing the current controller. However, during fault-tolerant operation, the limited bandwidth of the PI controller limits the improvement of the fault-tolerant operation performance. To address this problem, a bidirectional PI controller suitable for fault-tolerant operation is proposed. Changing the harmonic space current set value can achieve multiple control objectives such as minimum copper loss, equal fault-tolerant current amplitude, and minimum torque pulsation. The bidirectional PI controller can effectively follow the alternating current set value during fault-tolerant operation, thereby significantly improving the fault-tolerant operation performance of the motor. The above research focused on improving operating performance through effective control of fault-tolerant current, but failed to analyze the vector distribution characteristics during fault-tolerant operation, and thus did not solve the problem of difficult sector allocation, resulting in the inability to combine low-noise modulation strategies to suppress high-frequency noise. In order to solve the problem of difficult sector allocation, a non-orthogonal reduced-order decoupling matrix was proposed, a mathematical model of the faulty motor was established, the basic vector distribution under fault-tolerant operation was analyzed, a fault-tolerant switch table was constructed, and the direct torque control method was used to realize the fault-tolerant operation of the motor. However, in this method, only a single basic vector acts on the inverter in each control cycle, and the modulatable range is limited to finite discrete points, resulting in poor fault-tolerant operation performance of the motor. Some scholars have combined the duty cycle modulation method with model predictive fault-tolerant control. This scheme allocates a certain action time for the zero vector to optimize the effective vector amplitude, thereby expanding the modulation range to improve the motor operation performance. However, due to the failure to effectively utilize the basic vector, the modulation range at this time has been expanded but is still discontinuous. The above method analyzes the vector distribution during the fault-tolerant operation of the motor, but ignores the advantages of the multi-phase motor with many vectors, and thus fails to propose an effective multi-vector fault-tolerant control strategy, resulting in rich sideband harmonic components on both sides of the phase current fundamental wave, which is easy to excite low-frequency radial forces. The motor is generally driven by a voltage source inverter using pulse width modulation technology. The opening and closing of the power tube will generate concentrated high-frequency harmonics at the switching frequency and its integer multiples, causing high-frequency vibration and noise. Domestic and foreign scholars have carried out a lot of research on this issue, mainly suppressing concentrated high-frequency harmonics from the modulation strategy. Random PWM technology is a control strategy for suppressing high-frequency harmonics that is widely studied at this stage. This method does not require changing the hardware topology structure and has great research value and broad application prospects. At present, the commonly used random modulation strategies mainly include random zero vector SVPWM and random switching frequency SVPWM. In contrast, random switching frequency SVPWM has better harmonic dispersion effect, but the sampling frequency of this method will change randomly with the switching frequency, and the controller parameters cannot change accordingly, thus affecting the motor operation performance. However, the existing low-noise modulation strategy is not suitable for fault-tolerant operation. This is because the fault-tolerant vector is irregularly distributed and the sector allocation is difficult, resulting in the failure to establish a collaborative working mode between the low-noise modulation strategy and fault-tolerant control. Summary of the invention The present invention replaces the traditional single basic vector with a fault-tolerant virtual voltage vector, and uses two adjacent groups of fault-tolerant vectors to synthesize any reference voltage at the current moment, realizing a continuously modulated fault-tolerant control strategy, which not only significantly improves the fault-tolerant operation performance, but also effectively suppresses the sideband harmonics on both sides of the fundamental wave. Combined with the random zero vector-variable delay SVPWM, the harmonics at the switching frequency and its integer multiples are dispersed into the specified frequency domain without affecting the vector control operation performance, significantly reducing the high-frequency harmonic amplitude of the phase current, and realizing low-noise control after fault tolerance. In order to achieve the above-mentioned invention object, the present invention adopts the following technical scheme: The five-phase permanent magnet motor continuous modulation fault-tolerant double random low-noise control method includes the following steps: Step 1: Establish a reduced-order decoupling matrix, and use the constructed reduced-order decoupling matrix to establish a mathematical model of a five-phase permanent magnet motor with a single-phase winding open circuit; Step 2: Establish the basic voltage vector distribution after the fault, and synthesize multiple basic fault-tolerant virtual voltage vectors that make the harmonic space component zero; Step 3: synthesize the basic fault-tolerant virtual voltage vectors to obtain a vector set that can generate a standard PWM waveform; Step 4: synthesize adjacent vectors of the vectors that can generate the standard PWM waveform so that the set of fault-tolerant vectors is evenly distributed in space; Step 5: Use the zero vector to reduce the amplitude of some vectors so that the vector sets have the same amplitude; Step 6: Replace the traditional single vector control set with the fault-tolerant virtual voltage vector control set, and perform sector judgment according to the spatial distribution of the virtual voltage vector control set; Step 7: Use two adjacent vectors to synthesize the reference voltage vector, calculate the vector action time, and obtain the duty cycle of each phase; Step 8: Based on vector control, the two zero vector action times in one switching cycle are randomly allocated. And to a certain extent, the high-frequency PWM harmonics caused by the concentrated switching action are dispersed; Step 9: Use variable delay SVPWM to make the switching period T s Random changes constitute a random zero vector-variable delay SVPWM algorithm, which disperses the harmonics at the switching frequency and its integer multiples into the specified frequency domain, greatly reducing the amplitude of high-frequency harmonics. Furthermore, the reduced-order decoupling matrix makes the motor mathematical model of the fundamental wave space of the faulty motor consistent with that of the normal motor. The reduced-order decoupling matrix is constructed as follows: Where, δ = 2π / 5; The mathematical model of the five-phase permanent magnet motor with single-phase winding open circuit is constructed as follows: Among them, u d and u q are dq axis voltage, u y is the y-axis voltage; i d and i q is d q Shaft current, i y is the y-axis current; r s is the winding resistance; L d and L q are the dq axis inductance respectively; L ls is the harmonic plane leakage inductance; f Represents the flux linkage of a permanent magnet. Furthermore, the specific process of step 2 is: adopting the method of pre-synthesis of basic fault-tolerant virtual voltage vector, according to the corresponding distribution of voltage vectors in the fundamental plane and harmonic plane, using some voltage vectors that are similar in the fundamental plane direction and opposite in the harmonic plane direction, thereby synthesizing a fault-tolerant virtual voltage vector with zero harmonic plane. Further, the specific process of step 3 is: since the basic fault-tolerant virtual voltage vector with a zero harmonic plane cannot generate a standard PWM waveform when synthesizing the reference voltage vector, which creates difficulty in control, the basic fault-tolerant virtual voltage vector is used to synthesize the transition vector, so that a standard PWM waveform can be output when synthesizing the reference voltage vector. Further, the specific process of step 4 is: synthesize adjacent vectors of the vectors that can generate the standard PWM waveform so that the fault-tolerant virtual vector set is evenly distributed in space: since the vector set that can generate the standard PWM waveform is irregularly distributed in space, although the vector set presents a left-right symmetrical distribution, the phase angle of the vectors differs greatly. Therefore, at the position where the angle area between adjacent voltage vectors is too large, a new fault-tolerant virtual voltage vector is constructed so that the distribution of the voltage space vector is as even as possible. Furthermore, the specific process of step 5 is: using the zero vector to reduce the amplitude of some vectors, synthesizing the zero vector and the space voltage vector whose amplitude needs to be reduced in a certain ratio, so that the amplitudes of all vectors in the vector set are the same. Further, the specific process of step 6 is: according to the obtained fault-tolerant voltage vector control set, the sector It is judged that the entire fundamental wave space is divided into 14 sectors according to the distribution of the spatial voltage vector, and the sector selection is carried out according to the reference current angle in the stationary coordinate system, which provides a premise for selecting the action vector. θ ref = atan2(u α ,u β ) Among them, θ ref is the angle of the reference voltage vector, atan2 is the inverse tangent function, u α 、u β are the reference voltages for the α and β axes respectively. Furthermore, the specific process of step 7 is: replace the traditional single voltage space vector with a fault-tolerant virtual voltage vector, calculate the action time of each fault-tolerant virtual vector using the volt-second balance principle, synthesize the reference voltage vector, and use two adjacent fault-tolerant virtual voltage vectors and the zero vector to synthesize any reference voltage at the current moment, thereby realizing a continuously modulated fault-tolerant control strategy. Further, the specific process of step 8 is: using random zero vector SVPWM, randomly assigning the action time of two zero vectors in one switching cycle, randomizing the switching action, and reducing a part of the high-frequency PWM harmonics. The principle of random zero vector SVPWM is as follows: random zero vector SVPWM introduces random variables to randomly assign the action time of two zero vectors in one switching cycle, while not affecting the synthesized desired voltage vector, expanding the harmonic frequency band and reducing the harmonic peak. The expression of the action time of two zero vectors is: Where T 00 , T 15 They represent the action time of two zero vectors respectively, T0 represents the total action time of the zero vector, and R is a random number uniformly distributed in the range [0, 1]. Further, the specific process of step 9 is: using variable delay SVPWM so that the switching period T s Random changes; at the same time, the two zero vector action times in one switching cycle are randomly allocated to form a random zero vector-variable delay SVPWM algorithm, which disperses the harmonics at the switching frequency and its integer multiples into the specified frequency domain, greatly reducing the harmonic amplitude. The principle of variable delay SVPWM is as follows: The variable delay SVPWM randomly changes the switching frequency while keeping the sampling frequency unchanged, and ensures that the average value of the switching frequency is equal to the sampling frequency; this not only retains the advantage of the random switching frequency SVPWM in terms of obvious harmonic suppression effect, but also does not affect the control performance; the variable delay SVPWM calculates the delay time Δt of the next switching cycle in each sampling cycle, and the expression of the nth switching cycle is: T s (n) = T samp +Δt n -Δt n-1 Δt=R vtd T samp Where, T s (n) represents the generated n switching cycles, Δt n and Δt n-1 Indicates the delay time of the nth and n-1th switching cycles; T samp is the sampling period; R vtd is a random number uniformly distributed between [0,1]; the random zero vector-variable delay SVPWM can disperse the high-frequency harmonics at the switching frequency and its integer multiples into the specified frequency domain. The beneficial effects of the present invention are: 1. The present invention constructs a fault-tolerant virtual voltage vector composed of multiple basic vectors, which not only eliminates the need to control the harmonic spatial components, but also solves the problem of fault-tolerant operation sector allocation; combined with the fault-tolerant multi-vector control set, a continuous modulation fault-tolerant control strategy can be proposed to ensure fault-tolerant operation performance while suppressing sideband harmonics. 2. The present invention uses a random zero vector-variable delay SVPWM with a constant sampling frequency, which overcomes the disadvantage of the conventional random switching frequency SVPWM sampling frequency also randomly changing, resulting in the incompatibility of regulator parameters, and improves the stability of the system. 3. The present invention constructs a collaborative working mode of fault-tolerant control and low-noise modulation. BRIEF DESCRIPTION OF THE DRAWINGS FIG1 is a control block diagram of a five-phase permanent magnet motor continuous modulation fault-tolerant dual random low-noise control strategy; Figure 2 is a fundamental wave distribution diagram of the open-circuit vector of a single-phase winding of a five-phase motor; FIG3 is a distribution diagram of the y-axis component of the open-circuit vector of a single-phase winding of a five-phase motor; FIG4 is a vector distribution diagram of a synthetic y-axis component of zero; FIG5 is a distribution diagram of a vector set that can generate a standard PMW waveform; FIG6 is a distribution diagram of an extended vector set; FIG7 is a distribution diagram of vector sets optimized for phase angle and amplitude; Figure 8 is a schematic diagram of fault-tolerant space vector modulation Figure 9 shows the switch sequence; FIG10 is a schematic diagram of the principle of fault-tolerant random zero vector SVPWM; FIG11 is a schematic diagram of the fault-tolerant dual random vector SVPWM principle; FIG12 is a diagram showing the speed simulation results; FIG13 is a diagram showing the current simulation results; FIG14 is a diagram showing the effect of the Fourier analysis of the fault-tolerant fixed switching frequency B-phase current. Figure 15 is a diagram showing the effect of the fault-tolerant dual random B-phase current Fourier analysis. FIG. 16 is a flow chart of the method of the present invention. DETAILED DESCRIPTION The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention. As shown in the structural block diagram of FIG1 , the present invention is a five-phase permanent magnet motor continuous modulation fault-tolerant dual random low-noise control strategy, which mainly includes continuous modulation fault-tolerant control and random zero vector-variable delay SVPWM. The present invention takes a five-phase PMSM as the control object, performs fault-tolerant control after one phase is open-circuited, and suppresses harmonics of the phase current. The specific measures are as follows (as shown in FIG. 16 ): Step 1: The five-phase permanent magnet motor is driven by a five-phase two-level voltage source inverter. Under normal circumstances, there are 30 active vectors and 2 zero vectors. Assuming an open circuit fault occurs, the available vectors are reduced from 32 to 16. The present invention uses a simplified decoupling matrix to ensure that the discrete equations before and after the fault are the same. In the formula, δ=2π / 5. The mathematical model of the five-phase permanent magnet motor under single-phase open circuit is constructed as follows: Among them, u d and u q are dq axis voltage, u y is the y-axis voltage; i d and i q is d q Shaft current, i y is the y-axis current; r s is the winding resistance; L d and L q are the dq axis inductance respectively; L ls is the harmonic plane leakage inductance; f Represents the flux linkage of a permanent magnet. Step 2: Establish the basic voltage vector distribution after the fault and synthesize the basic voltage vector that makes the harmonic spatial component zero. Taking the single-phase winding of a five-phase motor as an example, Figures 2 and 3 are traditional single vector control sets, including the fundamental plane and the y-axis component. Obviously, the vector distribution at this time is irregular, which is not conducive to improving the fault-tolerant operation performance and suppressing the sideband harmonics. To this end, the basic fault-tolerant virtual voltage vector pre-synthesis method is adopted. According to the corresponding distribution of the voltage vectors in the fundamental plane and the harmonic plane, some voltage vectors are similar in the fundamental plane direction and opposite in the harmonic plane direction, so as to synthesize a fault-tolerant virtual voltage vector with zero harmonic plane. For example, the voltage vector v8 and v 13 The directions are similar on the αβ plane, but opposite on the harmonic plane. Therefore, by synthesizing the two vectors in a ratio opposite to the vector length, the amplitude of the synthesized fault-tolerant virtual voltage vector in the harmonic space can be made zero. U3, U5 and U6 are synthesized by a similar method. U1 and U4 are the basic voltage vectors v9 and v6, as shown in Figure 4. Establishing a fault-tolerant virtual voltage vector and then proposing a fault-tolerant control strategy with continuous modulation can not only suppress sideband harmonics, but also provide a way to suppress high-frequency noise in combination with a low-noise modulation strategy. The traditional single vector distribution includes the fundamental plane and the y-axis component. Obviously, the vector distribution at this time is irregular, which is not conducive to improving the fault-tolerant operation performance, and there are many low-frequency harmonic components in the phase current. Step 3: Synthesize the basic voltage vectors, consider the generation of standard PWM waves, and construct a new fault-tolerant virtual vector control set. Since the basic fault-tolerant virtual voltage vector with zero harmonic plane cannot generate a standard PWM waveform when synthesizing the reference voltage vector, which creates difficulty in control, the basic fault-tolerant virtual voltage vector is used to synthesize the transition vector, so that a standard PWM waveform can be output when synthesizing the reference voltage vector. By synthesizing the transition vector to generate a standard PWM wave, the vectors between U2 and U3, U5 and U6 are synthesized respectively, because the adjacent U2 (v8
[1000] 、v 13
[1101] ) and U3(v4
[0100] ,v 14
[1110] ) cannot generate a standard PWM waveform. To solve this problem, we use v8, v 12 and v 14 、v 12 Two vectors between U2 and U3 are synthesized. U5 (v2, v7) and U6 (v1, v 11 ) are used to synthesize two vectors by using v7, v3 and v1, v3 in the same way, as shown in Figure 5. Step 4: Considering the uniform angle distribution of the fault-tolerant voltage vector control set, the voltage vector that can generate the standard PWM waveform is resynthesized. The vectors that can generate standard PWM waveforms are synthesized by adjacent vectors to make the fault-tolerant virtual vector set uniformly distributed in space: Since the vector set that can generate standard PWM waveforms is irregularly distributed in space, although the vector set is in the form of bilateral symmetric distribution, the phase angle of the vectors differs greatly. Therefore, at the position where the angle area between adjacent voltage vectors is too large, a new fault-tolerant virtual voltage vector is constructed to make the distribution of the voltage space vector as uniform as possible. Synthesize the vectors between U1 and U2, U3 and U4, U4 and U5, and U6 and U1 respectively. Let U1(v9) and U2(v8,v 13 ) is fixed at 0.5 to synthesize a new vector. The other quadrants are synthesized in the same way, and the constructed voltage vector control set is shown in Figure 6. Step 5: Use the zero vector to reduce the amplitude of some vectors so that the vector set approaches the rule in terms of spatial phase angle and amplitude: Since the vector set has a large difference in amplitude after phase angle optimization, the zero vector is used to reduce the amplitude of some vectors. The zero vector and the spatial voltage vector whose amplitude needs to be reduced are synthesized in a certain ratio so that the amplitudes of all vectors in the vector set are the same. The amplitude of some vectors is reduced by using the zero vector, so that the vector set has the same amplitude: the duty cycle of the four vectors synthesized in the first step is fixed to 0.75, and the duty cycle of the vector synthesized in the second step is fixed to 0.9, so that the amplitude of the virtual voltage vector is evenly distributed, as shown in Figure 7. At this time, the fault-tolerant virtual voltage vector is synthesized from multiple basic vectors. This measure not only solves the problem of difficult sector allocation during fault-tolerant operation, but also eliminates the need to control the harmonic spatial components, simplifying the design steps of multi-vector fault-tolerant control. Step 6: Replace the traditional single vector control set with the fault-tolerant virtual voltage vector control set, and perform sector judgment according to the spatial distribution of the fault-tolerant virtual voltage vector control set. By using the αβ axis reference voltage in the stationary coordinate system, the inverse tangent function is used to obtain The spatial angle of the reference voltage vector is used to determine the sector in which it is located. It should be noted that the sector angle under the multi-vector fault-tolerant control set is not equally divided. The spatial angle of the reference voltage vector can be calculated by the following formula: θ ref = atan2(u α ,u β ) Among them, θ ref is the angle of the reference voltage vector, atan2 is the inverse tangent function, u α 、u β are the reference voltages for the α and β axes respectively. Step 7: Replace the traditional single vector control set with the fault-tolerant virtual vector control set, calculate the action time of each basic vector by the volt-second balance principle, and synthesize the reference voltage vector, as shown in Figure 8. Assuming that the reference voltage is located in the third sector in Figure 7, the PWM wave at this time is shown in Figure 9 (assuming that phase a is open). Obviously, each normal phase can generate a standard PWM wave that is symmetrical about the center. At the same time, there are 3 effective vectors and 2 zero vectors acting successively, making full use of the advantages of the multi-phase motor vector. By using two adjacent fault-tolerant virtual voltage vectors and zero vectors to synthesize any reference voltage at the current moment, a continuously modulated fault-tolerant control strategy is realized, which not only significantly improves the fault-tolerant operation performance, but also effectively suppresses the sideband harmonics on both sides of the fundamental wave. Step 8: Use the random zero vector SVPWM algorithm to randomly distribute the action time of two zero vectors in a switching cycle, randomize the switching action, and thus disperse the high-frequency PWM harmonics caused by the concentrated switching action to a certain extent. The principle of random zero vector SVPWM is as follows: The random zero vector SVPWM randomly distributes the action time of two zero vectors in one switching cycle by introducing random variables, as shown in Figure 10. It expands the harmonic frequency band and reduces the harmonic peak without affecting the synthetic desired voltage vector. The expression of the action time of the two zero vectors is: Where T 00 , T 15 They represent the action time of two zero vectors respectively, T0 represents the total action time of the zero vector, and R is a random number uniformly distributed in the range [0, 1]. Step 9: Further adopt variable delay SVPWM to make the switching period T s Random changes disperse the harmonics at the switching frequency and its integer multiples into the specified frequency domain, greatly reducing the amplitude of high-frequency harmonics. The variable delay SVPWM randomly changes the switching frequency while keeping the sampling frequency unchanged, and ensures that the average value of the switching frequency is equal to the sampling frequency. This not only retains the advantage of the random switching frequency SVPWM in terms of obvious harmonic suppression effect, but also does not affect the control performance. The variable delay SVPWM calculates the delay time Δt of the next switching cycle in each sampling cycle, and the expression of the nth switching cycle is: T s (n) = T samp +Δt n -Δt n-1 Δt=R vtd T samp Where, T s (n) represents the generated n switching cycles, Δt n and Δt n-1 Indicates the delay time of the nth and n-1th switching cycles; T samp is the sampling period; R vtd is a random number uniformly distributed between [0,1]. The PWM schematic diagram of the random zero vector-variable delay SVPWM is shown in Figure 11. It can disperse the high-frequency harmonics at the switching frequency and its integer multiples into the specified frequency domain, significantly reduce the harmonic amplitude, and does not affect the control performance of the system, and has excellent harmonic suppression effect. Figure 12 is a speed simulation result diagram when using the continuous modulation fault-tolerant control low noise strategy, and Figure 13 is a phase current simulation result diagram. It can be seen from the figure that the motor can quickly reach 400r / min and remain stable, and its phase current is a sine wave with little difference in amplitude, indicating that the continuous modulation fault-tolerant control strategy can improve the fault-tolerant control performance. Figures 14 and 15 are Fourier analysis diagrams of the B-phase current. It can be seen from Figure 14 that the current harmonics at a fixed switching frequency have obvious peaks at the switching frequency and its integer multiples, with a harmonic content of 0.93% at 10kHz and 1.18% at 20kHz; when the random zero vector-variable delay SVPWM is used, as shown in Figure 15, the harmonic spectrum tends to be continuous, and the maximum harmonic content is 0.29%. According to the above analysis, the random zero vector-variable delay SVPWM proposed in the present invention can effectively disperse the harmonics at the switching frequency and its integer multiples into the specified frequency domain, significantly reduce the harmonic amplitude, and thus reduce vibration noise. In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.
Claims
1. A fault-tolerant dual random low vibration control method for continuous modulation of a five-phase permanent magnet motor, characterized by comprising the following steps:Step 1: establishing a reduced-order decoupling matrix and utilizing the reduced-order decoupling matrix to formulate a mathematical model of the five-phase permanent magnet motor with a single-phase winding open circuit;Step 2: establishing a basic voltage vector distribution after fault, and synthesizing multiple basic fault-tolerant virtual voltage vectors to ensure harmonic spatial components being zero;Step 3: synthesizing the basic fault-tolerant virtual voltage vectors to obtain a vector set that generates standard pulse width modulation (PWM) waveforms;Step 4: synthesizing the vectors that generate standard PWM waveforms by adjacent vectors to evenly distribute the fault-tolerant vector set in space;Step 5: utilizing zero vectors to reduce amplitudes of certain vectors and making the vector set equal in the amplitudes;Step 6: replacing a traditional single vector control set with a fault-tolerant virtual voltage vector control set, and performing sector identification based on a spatial distribution of the virtual voltage vector control set;Step 7: synthesizing a reference voltage vector by two adj acent vectors and calculating a duration of vector action to determine a duty cycle of each phase;Step 8: distributing the action time of two zero vectors within a switching cycle in a random manner based on vector control to partially disperse high-frequency PWM harmonics caused by concentrated switching actions; andStep 9: implementing variable delay Space Vector PWM (SVPWM) to randomly adjust a switching period T& to establish a random zero vector variable delay SVPWM algorithm, wherein the algorithm disperses harmonics at a switching frequency and integer multiples of the switching frequency into specified frequency domain, resulting in a significant reduction in amplitudes of high-frequency harmonics.
2. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a reduced decoupling matrix is used to ensure consistency between the mathematical model of the motor in the fundamentalspace of the faulty motor and that of a normal motor; construction of the reduced decoupling matrix is as follows:COS (5 — 1 cos 2J-1 cos 3(5-1 cos 4(5-1 2 sin (5 SU12(5 sin 35 sin 4(5 T = — dark sin 35 sin 65 sin 95 sin 12(5 1 1 1 1wherein (5=2^ / 5;the mathematical model of the five-phase permanent magnet motor under single-phase opencircuit is constructed as follows:ud k 4 0 o’ d dt k ud = k + 0 Lq 0 k + 0 + Wf u i 0 0 L, i 0 L rj L y J L Is J L y Jwherein Ud and uq represent the d-q axis voltage, while uy represents the j-axis voltage; id and iq are the d-q axis current, and iy is the j-axis current; rs is the winding resistance; Ld and Lq are the d-q axis inductors, respectively; Lis is the harmonic plane leakage inductance; represents the magnetic flux of the permanent magnet.
3. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 2 is as follows: adopting the basic fault-tolerant virtual voltage vector pre-synthesis method, certain voltage vectors are utilized in similar directions in the fundamental plane and in opposite directions in the harmonic plane according to the corresponding distribution of voltage vectors in the fundamental plane and in the harmonic plane, to synthesize the fault-tolerant virtual voltage vectors whose harmonic plane is zero.
4. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 3 is as follows: due to the difficulty in generating standard PWM waveforms when synthesizing reference voltage vectors using the basic fault-tolerant virtual voltage vectors with zero harmonic plane component, and resulting in difficulty in control; a transition vector is synthesized by the basic fault-tolerant virtual voltage vectors to enable the output of standard PWM waveforms when synthesizing reference voltage vectors.
5. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific processof step 4 is as follows: combining vectors that can generate standard PWM waveforms by adjacent vectors to ensure the fault-tolerant virtual vector set is uniformly distributed in space, because the vector set that can generate standard PWM waveform is irregular in space distribution, although the vector set presents the form of left and right symmetric distribution, the phase angle of the vector is different; therefore, constructing new fault-tolerant virtual voltage vectors in the position wherein the angle region between adjacent voltage vectors is too large to ensure a more uniform distribution of voltage space vectors.
6. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 5 is as follows: utilizing the zero vector to reduce the amplitude of certain vectors, combining the zero vector with the space voltage vector that requires a slight reduction based on a certain proportion, thereby ensuring uniform amplitudes for all vectors within the vector set.
7. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 6 is as follows: performing the sector judgment according to the amplitude and angle distribution of the obtained fault-tolerant voltage vector control set; subsequently, dividing the entire fundamental space into 14 sectors based on the distribution of spatial voltage vectors, selecting sectors based on the reference current angle in the static coordinate system, which serves as a prerequisite for selecting the action vector;9=atm2(ua,u^wherein 0ref is the angle of the reference voltage vector, atan2 is the arctangent function, and ua and up are the a / ?-axis reference voltages, respectively.
8. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 7 is as follows: replacing the traditional single voltage space vector with a fault-tolerant virtual voltage vector, using the principle of volt-second balance to calculate the action time of each fault-tolerant vector, and synthesizing the reference voltage vector; by utilizing two adjacent fault-tolerant vectors and zero vectors to synthesize any reference voltage at the current moment; consequently, achieving the fault-tolerant control strategy with continuous modulation.
9. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific processof step 8 is as follows: random zero vector SVPWM is adopted to randomly allocate the action time of two zero vectors within a switching cycle, randomize the switching action, and reduce part of the high-frequency PWM harmonics; the principle of random zero vector SVWPM is as follows: random zero vector SVPWM introduces random variables and randomly allocates the action time of two zero vectors within a switching cycle, while this process does not impact the synthesis of the desired voltage vector, and effectively expands the harmonic frequency band and reduces harmonic spikes; the expression for the action time of the two zero vectors is:Xo = ^0wherein Too and Tn represent the action time of the two zero vectors, To represents the total time of zero vector action, and R is a uniformly distributed random number in the range of [0, 1],10. The fault-tolerant dual random low vibration control method for continuous modulation of the five-phase permanent magnet motor according to claim 1, characterized in that a specific process of step 9 is as follows: variable delay SVPWM is adopted to make the switching period Ts change randomly; while the operation time of two zero vectors in a switching cycle is randomly allocated to form a random zero vector-variable delay SVPWM algorithm, which disperses the harmonic at the switching frequency and integer multiples of the switching frequency into specified frequency domain, greatly reducing the harmonic amplitude; the principle of variable delay SVPWM is as follows:the variable delay SVPWM randomly adjusts the switching frequency while maintaining the sampling frequency unchanged; It ensures that the average switching frequency equals the sampling frequency. This approach retains the significant advantage of harmonic suppression effect of SVPWM with random switching frequency, while also preserving control performance, variable delay SVPWM calculates the delay time for the next switching cycle at each sampling cycle At, and the expression for the nth switching cycle is:At = R^mpwherein Tsamp is the sampling period; Aftd is a random number with a uniform distribution of [0, 1]; andthe dual random SVPWM is capable of dispersing high-frequency harmonics at theswitching frequency and integer multiples of the switching frequency into specified frequency domain.
Citation Information
Patent Citations
Fault-tolerant control method based on virtual voltage vector model prediction
CN108667389A
Nine-phase open winding permanent magnet synchronous motor model prediction control method based on virtual vectors
CN113315436A
Dead-beat-based five-phase permanent magnet synchronous motor model prediction fault-tolerant control method
CN115021636A
Method and apparatus for determining the flux angle of rotating field machine or for position-oriented operation of the machine
US4763058A
