A five-phase inverter open-phase voltage vector modulation method based on pre-synthesized vectors

Abstract

The application discloses a kind of five-phase inverter open-phase voltage vector modulation methods based on pre-synthesis vector, it is related to motor control technical field.The application is to solve the problem that in the decoupling fault-tolerant algorithm of five-phase motor, due to the lack of spatial voltage dimension, the number, amplitude and phase angle of the double-plane space vector of five-phase motor change greatly, are dispersed and complex.The application contains single-phase open circuit open-phase modulation method, solves the problem that fault vector is dispersed and complex, space vector modulation is difficult, by simplifying vector distribution, reducing modulation harmonic and optimizing modulation range, the integrity of five-phase inverter open-phase output voltage is realized and guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of motor control technology. Background Technology

[0002] With the increasing electrification of high-power hydraulic mechanized equipment and the rise of new energy vehicles, electrification and electric drive equipment are placing greater demands on motor drive systems. Multiphase motors offer advantages such as lower torque ripple, stronger fault tolerance, and multi-dimensional spatial control, making them widely applicable in high-reliability fields such as aviation, marine, large industrial equipment, and new energy off-road vehicles. Inverter open-circuit or motor phase-loss fault-tolerant operation is crucial for improving the reliability of military and industrial equipment. Whether it's a motor winding meltdown, a broken connection line, or an inverter bridge arm fault, all can be equivalent to a phase-loss output problem in the inverter. Therefore, in the five-phase motor decoupling fault-tolerant algorithm, the inverter in a phase-loss state experiences significant changes in the number, amplitude, and phase angle of the two-plane spatial vectors of the five-phase motor due to the lack of a spatial voltage dimension, resulting in a dispersed and complex situation. Summary of the Invention

[0003] This invention addresses the problem in five-phase motor decoupling fault-tolerant algorithms where the absence of a spatial voltage dimension causes significant changes in the number, amplitude, and phase angle of the dual-plane spatial vectors of the five-phase motor, resulting in a dispersed and complex situation. The invention provides a five-phase inverter phase-loss voltage vector modulation method based on pre-synthesized vectors.

[0004] A method for phase-loss voltage vector modulation of a five-phase inverter based on pre-synthesized vectors, specifically:

[0005] Step 1: Under the phase loss fault state of the motor inverter, calculate the fault voltage vector in the fundamental space and harmonic space respectively;

[0006] Step 2: Combine two fault voltage vectors located in the same quadrant region or on the same quadrant half-axis in the same space into a pre-combined vector;

[0007] Step 3: Calculate the time percentage of each fault voltage vector in each space. The time percentage of the fault voltage vector is the ratio of the duration of the fault voltage vector to the duration of the synthesized pre-synthesized vector.

[0008] Step 4: Take all the fault voltage vectors located on the quadrant half-axis in each space that have not been synthesized as pre-synthesized vectors;

[0009] Step 5: Calculate the duration of action for each pre-synthesized vector;

[0010] Step 6: Calculate the duration of each fault voltage vector based on its time proportion and the duration of the corresponding pre-synthesized vector;

[0011] Step 7: Divide the region between two adjacent pre-synthesized vectors in the same space into a sector. Take the sector where the reference driving voltage vector of each space is located under the current pulse width modulation period as the reference sector. Calculate the conduction time of the non-fault phase under the fundamental plane and harmonic plane according to the effect time of the fault voltage vector in the reference sector under each space.

[0012] Step 8: Superimpose the conduction times of the non-faulty phases in the fundamental plane and harmonic plane, and use the superposition result as the current five-phase inverter phase loss voltage vector modulation signal input to the five-phase inverter to complete the modulation.

[0013] Furthermore, in step two above, the two fault voltage vectors are synthesized into a pre-synthesized vector according to the following formula:

[0014]

[0015] Among them, U xk1 and U xk2 These are two fault voltage vectors located in the same space, within the same quadrant region, or on the same quadrant half-axis, U Vk For U xk1 and U xk2 The pre-synthesized vector, T xk1 T xk2 T Vk U xk1 U xk2 U Vk The duration of action.

[0016] Furthermore, in step three above, the β-axis voltage in the harmonic space is made zero to obtain the time proportion of the fault voltage vector:

[0017]

[0018] Among them, U 31 U 32 U xk1 U xk2 The harmonic associated vector, θ 31 θ 32 U 31 U 32 The vector angle.

[0019] Furthermore, in step five above, the duration of action of each pre-synthesized vector is calculated using the following formula:

[0020]

[0021] Among them, U r Let θ be the reference driving voltage vector, and Ur The angle after turning, T s θ is the pulse width modulation period time. k U is the angle covered by the k-th sector. Vk1 and U Vk2 These are two pre-composite vectors used to divide the k-th sector, T. Vk1 and T Vk2 U Vk1 and U Vk2 The duration of action.

[0022] Furthermore, in step six above, the calculation of the duration of the fault voltage vector includes:

[0023] When the pre-synthesized vector U Vk From the fault voltage vector U xk1 and U xk2 During synthesis, U xk1 and U xk2 The duration of action is obtained from the following formula:

[0024]

[0025] Among them, T xk1 T xk2 T Vk U xk1 U xk2 U Vk The duration of action, j xk1 and j xk2 U xk1 and U xk2 The percentage of time spent on this topic;

[0026] When the pre-synthesized vector U Vk The fault voltage vector U on the quadrant half-axis xk When, then U xk The duration of action is equal to U Vk The duration of action.

[0027] This invention proposes a vector modulation method for the phase-loss output voltage of a five-phase voltage source inverter based on pre-synthesized vectors. It includes a single-phase open-circuit phase-loss modulation method, solving the problems of dispersed and complex phase-loss fault vectors and the difficulty of spatial vector modulation. By simplifying the vector distribution, reducing modulation harmonics, and optimizing the modulation range, the integrity of the phase-loss output voltage of the five-phase inverter is achieved and guaranteed. The advantages of this invention are as follows:

[0028] The distributed dual-space fault vector distribution is simplified, and a new vector distribution is established based on the pre-synthesized vector, which regularizes the complex vector distribution. Moreover, the pre-synthesized vector has zero associated harmonics on the β-axis in the harmonic space, thus solving the harmonic association problem of fault vectors. Attached Figure Description

[0029] Figure 1 This is a diagram of the drive structure of a five-phase motor voltage source inverter.

[0030] Figure 2 The voltage vector distribution diagram of a five-phase motor is shown in (a) and (b) represents the fundamental frequency space and the harmonic frequency space.

[0031] Figure 3 The diagram shows the vector distribution of a phase loss fault in an inverter. (a) represents the fundamental frequency space, and (b) represents the harmonic frequency space.

[0032] Figure 4 The distribution diagram of the pre-synthesized vectors is shown in (a) for the fundamental wave space and (b) for the harmonic wave space.

[0033] Figure 5 A schematic diagram of the switching signal for space vector modulation of a fundamental plane phase loss fault;

[0034] Figure 6 This is a block diagram of a method for implementing phase loss fault space vector modulation based on pre-synthesized vectors. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0036] In practical applications, it is necessary to determine the missing phase based on open-circuit faults obtained through relevant diagnostic techniques. For different phase-loss conditions, the quantity, amplitude, and phase angle of the voltage space vector under different phase-loss conditions need to be analyzed. Then, optimization and synthesis methods are used to simplify modulation and reduce harmonics. Finally, the superposition theorem is used to improve the modulation ratio, thus realizing the inverter phase-loss space vector modulation method under this type of open-circuit fault. The specific implementation method is as follows.

[0037] Specific Implementation Method 1: Five-phase motor voltage source inverter drive structure as follows Figure 1 As shown, the calculation method for the dual-plane space voltage vector of a five-phase inverter under normal conditions is as follows:

[0038]

[0039] In the formula, U jn For a five-phase voltage j∈{a,b,c,d,e}, U dc U is the bus voltage.α1β1 and U α3β3 These are the voltage vectors in the fundamental space and harmonic space, respectively.

[0040] Equation (1) yields 32 spatial voltage vectors formed by the sequential combination of the five-phase bridge arms under normal conditions. These 32 voltage vectors exist in both the fundamental plane and the third harmonic plane, with large, medium, and small vectors that coexist. Their dual spatial voltage vector distribution is as follows: Figure 2 As shown. However, when a single-phase fault occurs (taking phase A fault as an example in this embodiment), the faulty phase does not output voltage. At this time, according to formula (1), the spatial vector distribution under a single-phase open-circuit fault can still be obtained as follows. Figure 3 As shown.

[0041] and Figure 2 In comparison, the voltage vectors are reduced, the spatial voltage vector arrangement is more scattered, and the modulation strategy is more difficult to implement. As shown in the fault vector distribution diagram, any voltage vector in the fundamental subplane satisfies the condition that the amplitude of the harmonic-associated vector along the α-axis is equal to and opposite in direction to the amplitude of the fundamental vector along the α-axis. In the case of a phase loss and open circuit, the voltage is subject to absolute constraints; the voltage and current only need to control three dimensions: the α and β axes of the fundamental plane and the β axis of the harmonic plane. This also places demands on the independence of the modulation method.

[0042] To address the issues of fault vector dispersion, complex modulation, and associated harmonics, this implementation proposes a pre-synthesized vector approach. This approach uses a new synthesized vector to represent the equivalent effects of two vectors. The synthesized vector follows a fixed pattern, simplifying modulation and resolving the associated harmonics of the fundamental and harmonic subplane vector voltages.

[0043] A method for phase-loss voltage vector modulation of a five-phase inverter based on pre-synthesized vectors, specifically:

[0044] Under the phase loss fault condition of the motor inverter, calculate the fault voltage vector in the fundamental space and harmonic space respectively.

[0045] Based on the vector composition relationship, two fault voltage vectors located in the same quadrant region or on the same quadrant half-axis in the same space are combined into a pre-composite vector:

[0046]

[0047] In the formula, U xk1 and U xk2 These are two fault voltage vectors located in the same quadrant region or on the same quadrant half-axis within the same space, such as... Figure 3 U8 and U in (a) 13 , or U 12 and U 10 k represents the sector number, U Vk For U xk1and U xk2 The pre-synthesized vector, T xk1 T xk2 T Vk U xk1 U xk2 U Vk The duration of action.

[0048] Let U xk1 and U xk2 The harmonic-associated vector is represented as U 31 U 32 To eliminate harmonic voltage interference and make the harmonic β-axis voltage zero, we have:

[0049]

[0050] In the formula, θ 31 θ 32 U 31 U 32 The vector angle.

[0051] From vector Figure 3 It can be seen that there are two types of pre-synthesized vectors: one is synthesized from different phase vectors, such as: the fundamental wave plane (U8, U...). 13 (U4,U) 14 ), (U2,U7), (U1,U 11 ); Harmonic plane (U2, U 14 (U8,U) 11 (U1,U) 13 (U4, U7). Another method is synthesis by in-phase vectors, such as: fundamental plane (U 10 U 12 ), (U3,U5); harmonic plane (U3,U5) 10 (U5,U) 12 The two pre-composite vectors have the same composition rules and amplitudes, but different phases and exhibit symmetry. Figure 3 (a) different phase vectors U8 and U 13 For example, let's consider U8 (0.4412∠59.5529°) and U... 13 Substituting (0.3245∠46.4362°) into equations (2) and (3), we get:

[0052]

[0053] In the formula, U V1 For U8 and U 13 The synthesized pre-synthesized vectors, T8 and T 13 They are U8 and U respectively 13 Duration of action, T V1 For UV1 The duration of action. Indicates T8 at T V1 The proportion in T represents 13 In T V1 The proportion in. It can be seen that the pre-synthesized vector U... V amplitude V S =0.3944, and the angle between it and the α axis is θ1 = 55.4642°. The time percentage of U8 is 0.618, U 13 The time percentage was 0.382.

[0054] Then use the in-phase vector U 10 and U 12 For example, U 10 (0.1453∠90°) and U 12 Substituting (0.6155∠90°) into equations (2) and (3) respectively, we get:

[0055]

[0056] In the formula, U V2 For U 10 and U 12 The synthesized pre-synthesized vector, T 10 and T 12 U 10 and U 12 Duration of action, T V2 For U V2 The duration of action. T represents 10 In T V2 The proportion in T represents 12 In T V2 The proportion in the second type of pre-synthesized vector magnitude V can be determined. L =0.5257, with an angle of 0° with the β-axis. U 12 The time percentage was 0.80902, U 10 The time percentage was 0.19098.

[0057] The above synthesis rules also apply to the pre-synthesis vector operations of the harmonic plane.

[0058] like Figure 4 As shown, two distributions of pre-synthesized vectors in the spatial plane are presented. All remaining fault vectors located on the quadrant half-axis within the spatial plane are used as pre-synthesized vectors, and the region between two adjacent pre-synthesized vectors is divided into a sector. This embodiment has a total of 8 sectors, and the angles and pre-synthesized vector amplitudes of these 8 sectors are shown in the following formulas:

[0059]

[0060] In the formula, θ1=55.4642°, θ2=90o-θ1=34.5358°, V S =0.3944 ( Figure 4 (a) U in V1 (amplitude), V M =0.4472 ( Figure 4 (a) Amplitude of U9), V L =0.5257( Figure 4 (a) U in V2 (amplitude). Since a sector includes two pre-synthesized vectors, and a pre-synthesized vector belongs to two adjacent sectors at the same time, the duplicates are removed in equation (6).

[0061] Based on the distribution, modulation is performed in each sector using two vectors of unequal amplitude. The duration of the two pre-synthesized vectors in each sector is calculated. For the k-th sector:

[0062]

[0063] In the formula, U r The reference driving voltage vector (obtained manually in practical applications) is θ, where θ is U. r The angle of the turn, T s θ is the pulse width modulation period. k U is the angle covered by the k-th sector. Vk1 and U Vk2 These are two pre-composite vectors used to divide the k-th sector, T. Vk1 and T Vk2 U Vk1 and U Vk2 The duration of action.

[0064] The duration of the fault vector is calculated based on its time proportion and the duration of action of the pre-synthesized vector. Specifically, when the pre-synthesized vector U... Vk By fault vector U xk1 and U xk2 If obtained through synthesis, then U xk1 and U xk2 The duration of action is obtained from the following formula:

[0065]

[0066] Among them, T xk1 T xk2 T Vk U xk1 U xk2 U Vk Duration of action

[0067] j xk1 For T xk1 In T Vk The proportion in and has j xk2 For T xk2 In T Vk The proportion of and has

[0068] When the pre-synthesized vector U Vk The fault vector U on the quadrant half-axis xk , then U xk The duration of action is equal to U Vk The duration of action.

[0069] Taking the first sector as an example, the pre-synthesized vector U9 in the first sector is Figure 3 (a) The fault vector on the half-axis of the middle quadrant, therefore the fault vector action time of U9 is the first sector pre-synthesized vector U. V1 The duration of action, i.e., T9 = T V1 .

[0070] Pre-synthesized vector U V1 By fault vectors U8 and U 13 If obtained through synthesis, then U8 and U 13 The duration of action T8 and T 13 They are respectively:

[0071] T8 = 0.618T V1 T 13 =0.382T V1 .

[0072] If the reference driving voltage vector in the fundamental space falls within the first sector during the current pulse width modulation period, then the switching signals and their durations in the first sector are as follows: Figure 5 As shown in the figure, the conduction time of each effective bridge arm switch is:

[0073]

[0074] In the formula, T0 = T s -T V1 -T V2 T0 is the zero vector action time. b T c T d and T e These are the conduction times of the non-faulty phases in the fundamental plane of phases B, C, D, and E, respectively.

[0075] Since only the voltage along the β axis needs to be considered in the harmonic space, only the voltage along the harmonic plane needs to be considered. Figure 4 (b) Pre-synthesized vector U VV2 and UVV5 When the reference driving voltage vector U in harmonic space r3 (In practical applications, this is achieved through manual setting) If the value is greater than zero, then U is selected. VV2 Modulation is performed; when U r3 If it is less than zero, then select U. VV6 Modulation is performed. Using U r3 Selecting U greater than zero VV2 For example:

[0076]

[0077] The vectors in the harmonic plane are not centrosymmetric, so the high-level signals can be directly added together. The equivalent centering method yields the operating times of each phase switch as follows:

[0078]

[0079] In the formula, T B3 T C3 T D3 and T E3 These are the conduction times of the non-faulty phases (phases B, C, D, and E) in the harmonic plane, respectively. According to the principle of dual-plane voltage superposition, the modulation method adds the modulated waves of the two planes, as shown in equation (12):

[0080] T X =T X1 +T X3 -0.5T s ,X∈{B,C,D,E} (12),

[0081] In the formula, T X This is the vector modulation signal for the missing phase voltage of a five-phase inverter obtained by superimposing the four-phase pulse width modulation signal waves of the fundamental plane and the harmonic plane.

[0082] This implementation method relates to a space vector modulation method for phase loss in a five-phase motor or a five-phase inverter bridge arm fault. First, it modulates the fundamental plane using a centralized equivalent approach to address the waveform's non-central symmetry. Then, it uses the superposition theorem to synthesize the PWM signals from the two planes, resolving the dual-plane synthesis problem. This implementation method simplifies the fault vector distribution and modulation process, reduces the need for modulation harmonic optimization, and can output a complete space voltage vector even under inverter phase loss conditions.

[0083] While the invention has been described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the invention. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of the invention as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for phase-loss voltage vector modulation of a five-phase inverter based on pre-synthesized vectors, characterized in that, Specifically: Step 1: Under the phase loss fault state of the motor inverter, calculate the fault voltage vector in the fundamental space and harmonic space respectively; Step 2: Combine two fault voltage vectors located in the same quadrant region or on the same quadrant half-axis in the same space into a pre-combined vector; Step 3: Calculate the time percentage of each fault voltage vector in each space. The time percentage of the fault voltage vector is the ratio of the duration of the fault voltage vector to the duration of the synthesized pre-synthesized vector. Step 4: Take all the fault voltage vectors located on the quadrant half-axis in each space that have not been synthesized as pre-synthesized vectors; Step 5: Calculate the duration of action for each pre-synthesized vector; Step 6: Calculate the duration of each fault voltage vector based on its time proportion and the duration of the corresponding pre-synthesized vector; Step 7: Divide the region between two adjacent pre-synthesized vectors in the same space into a sector. Take the sector where the reference driving voltage vector of each space is located under the current pulse width modulation period as the reference sector. Calculate the conduction time of the non-fault phase under the fundamental plane and harmonic plane according to the effect time of the fault voltage vector in the reference sector under each space. Step 8: Superimpose the conduction times of the non-faulty phases in the fundamental plane and harmonic plane, and use the superposition result as the current five-phase inverter phase loss voltage vector modulation signal input to the five-phase inverter to complete the modulation; In step two, the two fault voltage vectors are combined into a pre-combined vector according to the following formula: ， in, and These are two fault voltage vectors located in the same quadrant region or on the same quadrant half-axis within the same space. for and Synthesized pre-synthesized vector, , , They are respectively , , Duration of action; In step three, the harmonic space is lowered. When the shaft voltage is zero, the percentage of time for obtaining the fault voltage vector is: ， in, , They are respectively , Harmonic-associated vectors, , They are respectively , The vector angle; In step five, the duration of action of each pre-synthesized vector is calculated using the following formula: ， in, As a reference driving voltage vector, for The angle after turning The pulse width modulation period time. For the first The angle covered by each sector and They are respectively divided into the first Two pre-composite vectors for each sector, and They are respectively and The duration of action.

2. The method for phase-loss voltage vector modulation of a five-phase inverter based on pre-synthesized vectors according to claim 1, characterized in that, Step six involves calculating the duration of the fault voltage vector, including: When the pre-synthesized vector From the fault voltage vector and During synthesis, and The duration of action is obtained from the following formula: ， in, , , They are respectively , , Duration of action and They are respectively and The percentage of time spent on this topic; When the pre-synthesized vector Fault voltage vector on the quadrant half-axis At that time, The duration of action is equal to The duration of action.

Images