Improved variable virtual space vector modulation method for three-level inverter

Abstract

The invention discloses an improved variable virtual space vector modulation method for a three-level inverter, and relates to the technical field of inverter design. The expected voltage under the alpha-beta coordinate system input by the controller is converted into a 60-degree g-h coordinate system; six large sectors are divided according to the angle of the expected voltage, the large sectors are rotated to the position of 0-60 degrees, and space coordinates of all basic vectors under the neutral-point potential imbalance condition are given; selecting a k value to divide the large sector into five small sectors A1-A5, and obtaining space coordinates of each virtual vector; duty ratio calculation is carried out, and the duty ratio of the A5 small sector is calculated through an improved method; introducing a middle vector balance factor h, and calculating an h value to adjust a midpoint potential; and updating the duty ratio and outputting PWM. According to the invention, the balance speed of the neutral-point potential at a high modulation degree can be improved, output harmonics are reduced under the condition that the neutral-point potential is unbalanced, the quality of output electric energy is improved, and no fluctuation of the neutral-point potential is realized under a full power factor and a full modulation index.

Description

Technical Field

[0001] The invention relates to the technical field of inverter design, in particular to an improved variable virtual space vector modulation method of a three-level inverter. Background Art

[0002] Neutral point clamped (NPC) three-level inverter is widely used in high voltage and high power applications due to its advantages such as high voltage resistance and small output harmonics. However, the unbalanced midpoint potential will affect its output performance, and the difference in DC capacitance and switch tube pressure may also reduce the device life.

[0003] At present, the most commonly used methods for solving the midpoint potential problem include the redundant vector method and the virtual vector method. Among them: the redundant vector method can adjust the midpoint potential by adjusting the action time of the redundant small vector, introducing a balance factor, selecting different redundant small vectors according to the strength of the adjustment ability, and changing their action time to balance the midpoint potential. However, the balancing effect of this method is not ideal under high modulation index; the virtual vector method selects small vectors and medium vectors to synthesize a new virtual medium vector to ensure that the midpoint current generated within a cycle is zero, and discards the voltage vector that generates uncontrollable midpoint current within a cycle, which fundamentally solves the problem of midpoint potential fluctuation, but this method has limited ability to adjust the existing midpoint potential deviation.

[0004] In addition, the above methods are all aimed at the case of balanced midpoint potential, but in some specific applications, the midpoint potential needs to be controlled to an unbalanced state. For example, in a centralized photovoltaic inverter system, the voltages of the two DC side capacitors are usually controlled to an unbalanced state to achieve a single maximum power point tracking (MPPT) of the photovoltaic module. Considering that when the midpoint potential is unbalanced, the position of the basic vector will shift, thereby affecting the synthesis of the reference voltage, there are also some improved methods that consider the position changes of each basic vector when the midpoint potential is unbalanced to improve the accuracy of the reference vector synthesis. However, these methods involve using basic vectors from different large sectors for voltage synthesis, which may reduce the harmonic characteristics of voltage and current. There is also a variable virtual space vector modulation (VVSVM) method, which can have the ability to adjust the neutral point potential by changing the action time distribution of the three vectors in the synthetic virtual vector, but the midpoint potential cannot be precisely controlled. Summary of the invention

[0005] In order to solve the shortcomings of the background technology, the present invention provides an improved variable virtual space vector modulation method for a three-level inverter, which can improve the balancing speed of the mid-point potential at a high modulation index, reduce output harmonics when the mid-point potential is unbalanced, improve the output power quality, and achieve no fluctuation of the mid-point potential under full power factor and full modulation index.

[0006] To achieve the above object, the present invention adopts the following technical solution: an improved variable virtual space vector modulation method for a three-level inverter, comprising the following steps:

[0007] S1, expected voltage input

[0008] For the desired voltage in the α-β coordinate system input by the controller, transform it to the 60° gh coordinate system. The transformation formula is as follows:

[0009]

[0010] In the formula, v g and v h are the expected voltage in the gh coordinate system with an angle of 60°, V dc is the DC bus voltage, v α and v β are the expected voltages in the α-β coordinate system respectively;

[0011] S2. Large sector judgment

[0012] Divide the expected voltage into six large sectors, rotate the large sectors to the position of 0-60°, and give the spatial coordinates of each basic vector under the condition of unbalanced midpoint potential, which is expressed as follows:

[0013]

[0014] in,

[0015]

[0016] Where V OOO 、V PPP and V NNN is the basic zero vector, V POO 、V ONN 、V PPO and V OON is the basic small vector, V PON is the basic mean vector, V PNN and V PPN is the basic large vector, ζ is the imbalance coefficient, v P and v N are the voltages of the upper and lower capacitors on the DC side respectively;

[0017] S3. k value calculation and small sector judgment

[0018] According to the midpoint current i np and the polarity of the midpoint potential deviation ΔV, select a k value from 1 / 3, 2 / 3 or 5 / 6, and divide the large sector into five small sectors A1 to A5. The function expression of the boundary line between different small sectors is as follows:

[0019]

[0020] Where l1 is the boundary between small sectors A1 and A2, l2 is the boundary between small sectors A2 and A3, l3 is the boundary between small sectors A2 and A4, l4 is the boundary between small sectors A3 and A5, l5 is the boundary between small sectors A4 and A5, X VM and Y VM They are variable virtual middle vector V VM The horizontal and vertical coordinates in the gh coordinate system, the variable virtual center vector V VM The coordinate expression of is as follows:

[0021]

[0022] Then the spatial coordinates of each virtual vector are obtained, which are expressed as follows:

[0023]

[0024] Where V V0 is the virtual zero vector, V VS1 and V VS2 is a virtual small vector, V VL1 and V VL2 is a virtual large vector;

[0025] S4. Duty cycle calculation

[0026] Each small sector needs to use three vectors to synthesize the desired voltage vector. The duty cycle of each vector is calculated using the volt-second balance principle. The duty cycle calculation method for small sectors A1 to A4 uses the variable virtual vector method. The duty cycle calculation method for small sector A5 is improved:

[0027] Using V VL1 、V VL2 and V VM The duty cycles of these three vectors are calculated based on the volt-second balance principle:

[0028]

[0029] Where, d VL1 d VL2 and d VM V VL1 、V VL2 and V VM The duty cycle of V VM It is synthesized from three vectors and is expressed as follows:

[0030]

[0031] The calculated d VM According to formula (8), it is assigned to V ONN 、V PPO and V PON , find the actual duty cycle of each vector and the control cycle time T S Multiply them to get the action time T of each vector;

[0032] S5. Midpoint potential adjustment and calculation of midpoint vector balance factor h value

[0033] By V PNN and V PPN Each acts half the time to completely replace V PON , which is expressed as follows:

[0034]

[0035] Then, the mid-vector balance factor h is introduced and formula (8) is rewritten as:

[0036]

[0037] The amount of charge flowing through the midpoint of the DC side by the variable virtual mid vector generated by substitution is:

[0038]

[0039] The relationship between charge and voltage is:

[0040] Q=C·ΔV (12)

[0041] The combination of formula (11) and formula (12) can calculate the h value to adjust the midpoint potential, which is expressed as follows:

[0042]

[0043] In the formula, i a 、i b and i c is the three-phase current, C is the capacitance of a single capacitor on the DC side;

[0044] S6, duty cycle update and output PWM

[0045] According to the calculated h value, the vector action time that needs to be adjusted is calculated, which is expressed as follows:

[0046] T'=(1-h)·d VM ·T S / 3 (14)

[0047] According to T', the action time T calculated in S4 is adjusted to obtain the desired output time. Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention uses two large vectors to synthesize a new virtual vector, which is redundant with the basic middle vector, and introduces the middle vector balance factor to combine it into a new variable virtual middle vector. By considering the two balance factors of the middle vector balance factor h and the k value of the small sector division, the dynamic balance process of the midpoint potential can be accelerated and the steady-state midpoint potential fluctuation can be reduced. In addition, the simplified coordinates of each vector are determined by the gh coordinate system, which avoids complex trigonometric operations. At the same time, the unbalance coefficient ζ is introduced, so that the proposed algorithm can still ensure good output power quality when the midpoint potential is unbalanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is a control flow chart of the method of the present invention;

[0049] Figure 2 It is a schematic diagram of small sector division at three k values ​​in the method of the present invention;

[0050] Figure 3 : is a comparison diagram of output waveforms of the three methods in the embodiment when the load is 10Ω, 5mH and the unbalance coefficient ζ=0.25;

[0051] Figure 4 1 is a comparison diagram of output waveforms of the three methods in the embodiment when the load is 10Ω, 5mH and the unbalance coefficient ζ=0;

[0052] Figure 5 1 is an output waveform diagram of the method of the present invention in the embodiment when the load is 10Ω, 5mH and the unbalance coefficient ζ=±1;

[0053] Figure 6 It is a dynamic process diagram of adjusting from ζ=0.5 back to midpoint potential balance when the load is 8Ω, 16mH and the modulation index is 0.95 in the three methods in the embodiment. DETAILED DESCRIPTION

[0054] The technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0055] like Figure 1-2 As shown, an improved variable virtual space vector modulation method for a three-level inverter is provided. Figure 1 As shown, the following steps are included:

[0056] S1, expected voltage input

[0057] For the desired voltage in the α-β coordinate system input by the controller, transform it to the 60° gh coordinate system. The transformation formula is as follows:

[0058]

[0059] In the formula, v g and v h are the expected voltage in the gh coordinate system with an angle of 60°, V dc is the DC bus voltage, v α and v β are the expected voltages in the α-β coordinate system respectively.

[0060] S2. Large sector judgment

[0061] The large sectors are judged according to the angle of the expected voltage. After the judgment, the other large sectors are rotated to the position of 0-60°, and the spatial coordinates of each basic vector under the condition of unbalanced midpoint potential are given, which are expressed as follows:

[0062]

[0063] in,

[0064]

[0065] Where V OOO 、V PPP and V NNN is the basic zero vector, V POO 、V ONN 、V PPO and V OON is the basic small vector, V PON is the basic mean vector, V PNN and V PPN is the basic large vector, ζ is the imbalance coefficient, v P and v N are the voltages of the upper and lower capacitors on the DC side respectively.

[0066] According to the angular position of the desired voltage vector, it can be divided into six large sectors of N=1 to 6, that is, a large sector is divided every 60°.

[0067] S3. k value calculation and small sector judgment

[0068] According to the midpoint current i np and the polarity of the midpoint potential deviation ΔV, select a k value from 1 / 3, 2 / 3 or 5 / 6, divide the large sector N = 1 into small sectors, and combine the division of small sectors with different k values Figure 2As shown, the large sector N = 1 is divided into five small sectors n = A1 to A5. Figure 2 A1 to A5 are the specific locations of the five small sectors, and the function expression of the boundary line between different small sectors is as follows:

[0069]

[0070] Where l1 is the boundary between small sectors A1 and A2, l2 is the boundary between small sectors A2 and A3, l3 is the boundary between small sectors A2 and A4, l4 is the boundary between small sectors A3 and A5, l5 is the boundary between small sectors A4 and A5, X VM and Y VM They are variable virtual middle vector V VM The horizontal and vertical coordinates in the gh coordinate system, the variable virtual center vector V VM The coordinate expression of is as follows:

[0071]

[0072] Then the spatial coordinates of each virtual vector can be obtained, which is expressed as follows:

[0073]

[0074] Where V V0 is the virtual zero vector, V VS1 and V VS2 is a virtual small vector, V VL1 and V VL2 It is a virtual large vector.

[0075] S4. Duty cycle calculation

[0076] Each small sector needs to use three vectors to synthesize the desired voltage vector. The duty cycle of each vector is calculated using the volt-second balance principle. The duty cycle calculation method for small sectors A1 to A4 adopts the traditional variable virtual vector method. The duty cycle calculation method for small sector A5 is improved as follows:

[0077] In the A5 small sector, V VL1 、V VL2 and V VM The duty ratios of these three vectors are calculated based on the volt-second balance principle:

[0078]

[0079] Where, d VL1 d VL2 and d VM V VL1 、VVL2 and V VM The duty cycle of V VM It is synthesized from three vectors and is expressed as follows:

[0080]

[0081] The calculated d VM According to formula (8), it is assigned to V ONN 、V PPO and V PON , the actual duty cycle of each vector can be obtained, and then, with the control cycle time T S Multiply them to get the action time T of each vector.

[0082] S5. Midpoint potential adjustment and calculation of midpoint vector balance factor h value

[0083] By V PNN and V PPN Each acts half the time to completely replace V PON , which is expressed as follows:

[0084]

[0085] Then, the mid-vector balance factor h is introduced and formula (8) is rewritten as:

[0086]

[0087] The amount of charge flowing through the midpoint of the DC side by the variable virtual mid vector generated by substitution is:

[0088]

[0089] In the formula, i a 、i b and i c is the three-phase current.

[0090] The relationship between charge and voltage is:

[0091] Q=C·ΔV (12)

[0092] The combination of formula (11) and formula (12) can calculate the h value to adjust the midpoint potential, which is expressed as follows:

[0093]

[0094] Where C is the capacitance of a single capacitor on the DC side.

[0095] S6, duty cycle update and output PWM

[0096] According to the calculated h value, the vector action time that needs to be adjusted is calculated, which is expressed as follows:

[0097] T'=(1-h)·d VM ·T S / 3 (14)

[0098] By adjusting the action time T calculated in S4 according to T', the desired output time can be obtained.

[0099] Example

[0100] In order to verify the superiority of this method, the virtual vector method, the traditional variable virtual vector method and the improved variable virtual vector method of the present invention are used for experimental comparison. The experimental load is a resistive inductive load, and two power factor conditions are verified, which are 10Ω, 5mH and 8Ω, 16mH respectively. ab represents the line voltage between phases ab, i a Represents the a-phase current. Figure 3 It can be seen that when the midpoint potential is deviated, the output current THD of the traditional variable virtual vector method is larger than that of the other two methods; Figure 4 It can be seen that when the midpoint potential is balanced, the current harmonics output by the three methods are not much different. Therefore, the virtual vector method and the variable virtual vector method improved by the present invention can ensure the output power quality when the midpoint potential is unbalanced. Figure 5 It can be seen that the improved variable virtual vector method of the present invention can still ensure that the output current harmonic THD only increases by 0.34% when the midpoint potential is unbalanced to the limit; Figure 6 It can be seen that the dynamic balancing process of the improved variable virtual vector method of the present invention is faster than that of the other two methods, and the balancing time is reduced by 61.8 ms and 12.8 ms respectively.

[0101] In summary, compared with the virtual vector method and the traditional variable virtual vector method, the improved variable virtual vector method of the present invention has a smaller output current THD and a faster speed of dynamically balancing the midpoint potential.

[0102] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other forms of assembly without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-restrictive, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations within the meaning and range of equivalents of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

[0103] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. An improved variable virtual space vector modulation method for a three-level inverter, characterized in that: The following steps are involved: S1, expected voltage input For the desired voltage in the α-β coordinate system input by the controller, transform it to the 60° gh coordinate system. The transformation formula is as follows: In the formula, v g and v h are the expected voltage in the gh coordinate system with an angle of 60°, V dc is the DC bus voltage, v α and v β are the expected voltages in the α-β coordinate system respectively; S2. Large sector judgment Divide the expected voltage into six large sectors, rotate the large sectors to the position of 0-60°, and give the spatial coordinates of each basic vector under the condition of unbalanced midpoint potential, which are expressed as follows: in, Where V OOO 、V PPP and V NNN is the basic zero vector, V POO 、V ONN 、V PPO and V OON is the basic small vector, V PON is the basic mean vector, V PNN and V PPN is the basic large vector, ζ is the imbalance coefficient, v P and v N are the voltages of the upper and lower capacitors on the DC side respectively; S3. k value calculation and small sector judgment According to the midpoint current i np and the polarity of the midpoint potential deviation ΔV, select a k value from 1 / 3, 2 / 3 or 5 / 6, and divide the large sector into five small sectors A1 to A5. The function expression of the boundary line between different small sectors is as follows: Where l1 is the boundary between small sectors A1 and A2, l2 is the boundary between small sectors A2 and A3, l3 is the boundary between small sectors A2 and A4, l4 is the boundary between small sectors A3 and A5, l5 is the boundary between small sectors A4 and A5, X VM and Y VM They are variable virtual middle vector V VM The horizontal and vertical coordinates in the gh coordinate system, the variable virtual center vector V VM The coordinate expression is as follows: Then the spatial coordinates of each virtual vector are obtained, which are expressed as follows: Where V V0 is the virtual zero vector, V VS1 and V VS2 is a virtual small vector, V VL1 and V VL2 is a virtual large vector; S4. Duty cycle calculation Each small sector needs to use three vectors to synthesize the desired voltage vector. The duty cycle of each vector is calculated using the volt-second balance principle. The duty cycle calculation method for small sectors A1 to A4 uses the variable virtual vector method. The duty cycle calculation method for small sector A5 is improved: Using V VL1 、V VL2 and V VM The duty cycles of these three vectors are calculated based on the volt-second balance principle: Where, d VL1 ,d VL2 and d VM V VL1 、V VL2 and V VM The duty cycle of V VM It is synthesized from three vectors and is expressed as follows: The calculated d VM According to formula (8), it is assigned to V ONN 、V PPO and V PON , find the actual duty cycle of each vector and the control cycle time T S Multiply them to get the action time T of each vector; S5. Midpoint potential adjustment and calculation of midpoint vector balance factor h value By V PNN and V PPN Each acts half the time to completely replace V PON , which is expressed as follows: Then, the mid-vector balance factor h is introduced and formula (8) is rewritten as: The amount of charge flowing through the midpoint of the DC side by the variable virtual mid vector generated by substitution is: The relationship between charge and voltage is: Q=C·ΔV (12) The combination of formula (11) and formula (12) can calculate the h value to adjust the midpoint potential, which is expressed as follows: In the formula, i a 、i b and i c is the three-phase current, C is the capacitance of a single capacitor on the DC side; S6, duty cycle update and output PWM According to the calculated h value, the vector action time that needs to be adjusted is calculated, which is expressed as follows: T'=(1-h)·d VM ·T S / 3 (14) By adjusting the action time T calculated in S4 according to T', the desired output time can be obtained.

Images