A midpoint voltage control method, device and midpoint clamped three-level converter

Abstract

This invention discloses a midpoint voltage control method, device, and midpoint clamping three-level converter. The method includes: acquiring the current three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower bus voltages on the DC side of the midpoint clamping three-level converter; determining the midpoint voltage deviation based on the upper and lower bus voltage differences, and calculating the average midpoint current used to eliminate the midpoint voltage deviation; extracting the polarity sign of the three-phase modulation wave, and calculating the current flowing into the midpoint from the non-injected zero-sequence component and the current flowing into the midpoint corresponding to the injected zero-sequence component based on the three-phase modulation wave, polarity sign, and actual three-phase current; calculating the zero-sequence voltage injection amount based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint from the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence component; and superimposing the zero-sequence voltage injection amount onto the three-phase modulation wave. This method can provide midpoint voltage balance control with both high steady-state accuracy and fast dynamic response.

Description

Technical Field

[0001] This invention relates to the field of converter technology, and in particular to a midpoint voltage control method, device, and midpoint clamping three-level converter. Background Technology

[0002] Midpoint clamping three-level converters, with their advantages of low device voltage stress, low output voltage harmonics, and high equivalent switching frequency, have become core equipment in modern power electronics fields such as high-voltage high-power transmission, new energy grid-connected power generation, and active filtering. Their multi-level output characteristics significantly improve power quality and reduce the size and cost of filters, playing a vital role in promoting the development of high-performance power conversion technology. However, this converter inherently suffers from DC-side midpoint voltage fluctuations and imbalances during operation. If left uncontrolled, this can lead to capacitor voltage imbalance, overvoltage damage to power devices, and output waveform distortion, severely impacting the safe and reliable operation of the system and the quality of grid-connected power.

[0003] To address the midpoint voltage balance problem, existing control strategies mainly fall into two categories: redundant small vector control based on space vector pulse width modulation (SVPWM) and zero-sequence voltage injection based on carrier pulse width modulation (SPWM). However, both methods have significant limitations in practical applications: On the one hand, SVPWM-based methods require real-time complex coordinate transformations, sector judgments, vector action time calculations, and redundant small vector pairing and allocation. The computational load is extremely large, placing high demands on the microcontroller's computing power and storage resources. Especially in over-modulation regions or high-frequency operating conditions, the complex logic can easily lead to program execution timeouts or control lags. On the other hand, SPWM-based zero-sequence voltage injection methods often use approximate formulas or fixed lookup tables, failing to accurately establish the analytical relationship between the zero-sequence injection amount and the midpoint current. Essentially, this is a "blind injection" strategy, resulting in the calculated zero-sequence voltage injection amount deviating from the actual requirement. This leads to insufficient control accuracy and slow dynamic response when the midpoint voltage deviation is small.

[0004] In summary, existing technologies either rely on complex vector operations and sector pairing logic, resulting in huge computational overhead and easy control lag; or they use coarse approximation models, causing zero-sequence injection to deviate from the actual requirements and create control blind spots. Neither approach can achieve precise decoupling of the physical components of the midpoint current and accurate calculation of the zero-sequence voltage injection while maintaining lightweight algorithms and low computational consumption, thus failing to meet the high-precision, fast-response balance control requirements of the midpoint voltage. Summary of the Invention

[0005] The technical problem to be solved by this invention is to address the aforementioned shortcomings of the prior art by proposing a midpoint voltage control method, device, and midpoint clamping three-level converter. This method can provide midpoint voltage balance control for the midpoint clamping three-level converter, offering both high steady-state accuracy and fast dynamic response.

[0006] In a first aspect, the present invention provides a method for controlling the neutral point voltage of a neutral point clamping three-level converter, the method comprising:

[0007] Obtain the current three-phase modulation waveform, the actual three-phase current, and the voltage difference between the upper and lower busbars on the DC side of the neutral-point clamped three-level converter;

[0008] The midpoint voltage deviation is determined based on the voltage difference between the upper and lower busbars, and the average midpoint current used to eliminate the midpoint voltage deviation is calculated.

[0009] Extract the polarity sign of the three-phase modulated wave, and calculate the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component based on the three-phase modulated wave, polarity sign and actual three-phase current.

[0010] The zero-sequence voltage injection amount is calculated based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence component.

[0011] The zero-sequence voltage injection is superimposed on the three-phase modulation wave to achieve balanced control of the midpoint voltage.

[0012] Furthermore, the midpoint voltage deviation is determined based on the voltage difference between the upper and lower busbars, specifically including:

[0013] The voltage difference between the upper and lower busbars is low-pass filtered to obtain the filtered voltage difference.

[0014] The midpoint voltage deviation is determined based on the filtered voltage difference: if the absolute value of the filtered voltage difference is less than or equal to the preset voltage dead zone threshold, the midpoint voltage deviation is set to zero; if the absolute value of the filtered voltage difference is greater than the voltage dead zone threshold, the filtered voltage difference is used as the midpoint voltage deviation.

[0015] Furthermore, the average neutral point current used to eliminate the neutral point voltage deviation is calculated, specifically including:

[0016] The average midpoint current is calculated based on the midpoint voltage deviation, the half-bus capacitor parameters of the midpoint clamping three-level converter, and the sampling period.

[0017] The target direction of the average midpoint current is opposite to the polarity of the midpoint voltage deviation.

[0018] Furthermore, the polarity sign of the three-phase modulated wave is extracted, specifically including:

[0019] The three-phase modulated waves are each subjected to amplitude limiting processing to ensure that each phase modulated wave is within a preset amplitude range, thus obtaining the amplitude-limited modulated wave;

[0020] Based on the sign of the modulation wave after amplitude limiting in each phase of the three-phase modulation wave, the polarity sign of each phase is determined. If the modulation wave after amplitude limiting of a certain phase is greater than or equal to zero, the polarity sign of that phase is determined to be positive; if the modulation wave after amplitude limiting of a certain phase is less than zero, the polarity sign of that phase is determined to be negative.

[0021] Furthermore, based on the three-phase modulation wave, polarity sign, and three-phase actual current, the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component are calculated. Specifically, this includes: calculating the current flowing into the midpoint of the non-injected zero-sequence component based on the three-phase modulation wave, polarity sign, and three-phase actual current; and calculating the current flowing into the midpoint of the injected zero-sequence component based on the polarity sign and three-phase actual current.

[0022] Based on the three-phase modulation wave, polarity sign, and actual three-phase current, calculate the current flowing into the midpoint from the non-injected zero-sequence component, specifically including:

[0023] The three-phase modulated wave is amplitude-limited to obtain the amplitude-limited three-phase modulated wave.

[0024] For each phase of the three phases of the midpoint clamped three-level converter, the actual current of the phase, the three-phase modulation wave after limiting of the phase, and the corresponding polarity sign are multiplied to obtain the non-injected phase product of the phase.

[0025] Multiply and accumulate the non-injected phases of the three phases to obtain the non-injected zero-sequence component of the current flowing into the midpoint;

[0026] Based on the polarity sign and the actual three-phase current, calculate the current flowing into the midpoint corresponding to the zero-sequence injection, specifically including:

[0027] For each phase of the three-phase neutral-clamped three-level converter, the actual current of that phase is multiplied by the corresponding polarity sign to obtain the injected phase product.

[0028] By multiplying and summing the injections of the three phases, we obtain the current flowing into the midpoint corresponding to the zero-sequence injection.

[0029] Furthermore, based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint from the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence, the zero-sequence voltage injection amount is calculated, specifically including:

[0030] The absolute value of the current flowing into the midpoint corresponding to the zero-sequence injection is compared with a preset current dead zone threshold, and the zero-sequence voltage injection amount is calculated based on the comparison result.

[0031] If the absolute value of the current flowing into the midpoint corresponding to the injected zero sequence is greater than the preset current dead zone threshold, the sum of the three-phase actual currents is calculated, and the preset midpoint current equation relationship satisfied by the average midpoint current, the sum of the three-phase actual currents, the current flowing into the midpoint of the non-injected zero sequence component, and the current flowing into the midpoint corresponding to the injected zero sequence is solved algebraically to obtain the original zero sequence voltage injection amount. The original zero sequence voltage injection amount is then limited to obtain the zero sequence voltage injection amount.

[0032] If the absolute value of the current flowing into the midpoint corresponding to the zero-sequence injection is less than or equal to the current dead zone threshold, then the zero-sequence voltage injection amount is determined to be zero.

[0033] Furthermore, the zero-sequence voltage injection is superimposed on the three-phase modulation wave to achieve balanced control of the midpoint voltage, specifically including:

[0034] Extract the minimum and maximum values ​​from the three-phase modulated wave;

[0035] Based on the minimum and maximum values, boundary constraints are applied to the zero-sequence voltage injection quantities to be superimposed, so that the superimposed phase results are all within the preset amplitude range.

[0036] The constrained zero-sequence voltage injection quantity is superimposed on the three-phase modulation wave to obtain the target three-phase modulation wave;

[0037] The neutral point voltage is balanced based on the target three-phase modulation wave.

[0038] Furthermore, based on the minimum and maximum values, the zero-sequence voltage injection amount to be superimposed is subject to boundary constraints to obtain the constrained zero-sequence voltage injection amount, specifically including:

[0039] Based on the polarity sign combination of the three-phase modulation wave, the target sector where the space voltage vector of the current midpoint clamping three-level converter is located is determined, and the reference boundary value corresponding to the target sector is obtained according to the preset target sector and zero-sequence voltage boundary correspondence table.

[0040] With the constraint that the superimposed three-phase modulated wave does not exceed the preset amplitude range, the reference boundary value is corrected using the maximum and minimum values ​​to obtain the physical allowable boundary, and the maximum allowable absolute value of the physical allowable boundary is extracted as the upper limit of the hardware topology constraint.

[0041] The dynamic limiting threshold is calculated based on the numerical range of the absolute value of the current midpoint voltage deviation. When the absolute value of the midpoint voltage deviation is in a preset first numerical range, the dynamic limiting threshold is determined as the initial limiting value. When the absolute value of the midpoint voltage deviation is in a preset second numerical range, the dynamic limiting threshold is reduced according to the magnitude of the absolute value of the midpoint voltage deviation and a preset decreasing rule. The upper limit of the second numerical range is less than the lower limit of the first numerical range.

[0042] The dynamic limiting threshold is clamped within the upper limit of the hardware topology constraint to obtain the final limiting threshold. The zero-sequence voltage injection amount to be superimposed is symmetrically truncated and limited using the final limiting threshold to obtain the constrained zero-sequence voltage injection amount.

[0043] In a second aspect, the present invention provides a midpoint voltage control device for a midpoint clamping three-level converter, the device comprising:

[0044] The acquisition unit is used to acquire the current three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower busbars on the DC side of the neutral-point clamped three-level converter.

[0045] The first calculation unit, connected to the acquisition unit, is used to determine the midpoint voltage deviation based on the voltage difference between the upper and lower busbars, and to calculate the average midpoint current used to eliminate the midpoint voltage deviation.

[0046] The second calculation unit, connected to the first calculation unit, is used to extract the polarity sign of the three-phase modulated wave, and calculate the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component based on the three-phase modulated wave, polarity sign and three-phase actual current.

[0047] The third calculation unit, connected to the second calculation unit, is used to calculate the zero-sequence voltage injection amount based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint of the injected zero-sequence component.

[0048] The control unit, connected to the third calculation unit and the acquisition unit respectively, is used to superimpose the zero-sequence voltage injection amount onto the three-phase modulation wave to achieve balanced control of the midpoint voltage.

[0049] Thirdly, the present invention provides a midpoint clamping three-level converter, which includes: a converter main circuit, a sampling and detection circuit, and a midpoint voltage control device for the midpoint clamping three-level converter described in the second aspect;

[0050] The DC side of the converter's main circuit includes an upper bus capacitor and a lower bus capacitor connected in series. The connection point of the upper bus capacitor and the lower bus capacitor constitutes the neutral point.

[0051] The sampling and detection circuit is connected to the main circuit of the converter and is used to collect the actual three-phase current of the main circuit of the converter and the voltage difference between the upper and lower bus on the DC side.

[0052] The midpoint voltage control device is connected to the converter main circuit and the sampling and detection circuit respectively. It is used to generate a target three-phase modulation wave with zero-sequence voltage injection superimposed on it based on the voltage difference between the upper and lower bus and the actual three-phase current collected by the sampling and detection circuit. Based on the target three-phase modulation wave, a drive signal is generated to modulate the converter main circuit, thereby realizing the balance control of the midpoint voltage.

[0053] This invention constructs a refined analytical model of the midpoint current based on the polarity sign of the modulation wave. The current flowing into the midpoint is decoupled into two parts: the "non-injected zero-sequence component current" and the "injected zero-sequence corresponding current." This model is then combined with the average midpoint current required to eliminate deviations for inverse precise solution. This method completely eliminates the complex vector logic of SVPWM and overcomes the blindness and approximation errors of the traditional SPWM zero-sequence injection method, achieving high-precision, fast-response balanced control of the midpoint voltage with extremely low computational power. Specific beneficial effects are as follows:

[0054] (1) Finely decouple the midpoint current to eliminate approximation error.

[0055] To address the control dead zone and approximation error inherent in existing SPWM zero-sequence injection methods due to their reliance on approximate formulas or fixed lookup tables, this invention finely decouples the midpoint current. By extracting the polarity symbol of the modulation wave as a key decoupling variable, it accurately separates and quantifies the physical contributions of the "non-injected zero-sequence component" and the "injected zero-sequence component," thereby clearly characterizing the actual impact of load current and zero-sequence voltage on the midpoint potential. Based on this, a strict analytical mapping relationship is established between the zero-sequence injection quantity and the midpoint current, completely eliminating the calculation errors and control blind zones caused by traditional approximation models, and ultimately achieving on-demand accurate calculation of the zero-sequence voltage.

[0056] (2) It balances lightweight algorithm with high control performance, effectively reducing hardware costs.

[0057] This invention eliminates the cumbersome coordinate transformation, sector judgment, and redundant small vector pairing logic in SVPWM, and directly performs simple algebraic and logical operations based on conventional three-phase modulation waves, transforming complex vector geometric calculations into lightweight algebraic analysis. Thanks to this simplified architecture, the computing power and storage overhead of the microcontroller are greatly reduced, thereby effectively eliminating the control lag caused by program execution timeouts in over-modulation regions or high-frequency operating conditions.

[0058] (3) Significantly improves dynamic response speed and enhances robustness under all operating conditions.

[0059] This invention dynamically calculates the average midpoint current value used to eliminate deviations by real-time acquisition of the actual three-phase current and the voltage difference between the upper and lower busbars, and relies on a real-time physical quantity dynamic mapping mechanism to enable the zero-sequence voltage injection to instantly follow changes in operating conditions. This control strategy, which combines dynamic feedforward and real-time feedback, ensures extremely strong rapid response and recovery capabilities when facing load changes or grid disturbances.

[0060] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0061] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. The above and other features and advantages will become more apparent to those skilled in the art from the detailed description of exemplary embodiments with reference to the accompanying drawings, in which:

[0062] Figure 1 This is a schematic diagram of the midpoint voltage control method for a midpoint clamping three-level converter provided in an embodiment of the present invention;

[0063] Figure 2 This is a diagram illustrating a method for controlling the midpoint voltage balance of a three-level NPC converter with segmented zero-sequence injection, provided in an embodiment of the present invention.

[0064] Figure 3 This is a schematic diagram of the limiting adjustment logic provided in an embodiment of the present invention;

[0065] Figure 4 This is a schematic diagram of the midpoint voltage control device for a midpoint clamping three-level converter provided in an embodiment of the present invention;

[0066] Figure 5 This is a schematic diagram of a midpoint clamping three-level converter provided in an embodiment of the present invention.

[0067] Reference numerals: 10, acquisition unit; 20, first calculation unit; 30, second calculation unit; 40, third calculation unit; 50, control unit. Detailed Implementation

[0068] It is understood that the specific embodiments and accompanying drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0069] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0070] It is understood that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, while the parts unrelated to the present invention are not shown in the drawings.

[0071] It is understood that each unit or module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units or modules may be integrated into one entity structure.

[0072] It is understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of this invention may occur in a different order than that marked in the accompanying drawings.

[0073] It is understood that the flowcharts and block diagrams of this invention illustrate the possible architecture, functions, and operations of systems, apparatuses, devices, and methods according to various embodiments of this invention. Each block in the flowchart or block diagram may represent a unit, module, program segment, or code, containing executable instructions for implementing the specified function. Furthermore, each block or combination of blocks in the block diagram and flowchart can be implemented using a hardware-based system to achieve the specified function, or using a combination of hardware and computer instructions.

[0074] It is understood that the units and modules involved in the embodiments of the present invention can be implemented by software or by hardware. For example, the units and modules can be located in a processor.

[0075] Example 1:

[0076] This embodiment provides a method for controlling the neutral point voltage of a neutral point clamping three-level converter, and its overall control flow is as follows: Figure 1 As shown. This method is mainly applied to the neutral point voltage control device or controller of a neutral point clamping three-level converter. By adjusting the neutral point potential in real time, it ensures the voltage balance of the upper and lower bus capacitors on the DC side of the converter, thereby improving the system's operational stability and power quality.

[0077] From the perspective of topology applicability, this method focuses on three-level NPC (Neutral Point Clamped) converters. As a classic neutral point clamped topology, the NPC converter achieves three-level output through diode clamping, offering advantages such as low voltage stress on switching devices and high output waveform quality. Furthermore, the applicability of this control method is not limited to traditional NPC topologies; it can be further extended to other neutral point clamped topologies such as T-type three-level converters (TNPC) and active neutral point clamped (ANPC) converters. Specifically, the T-type three-level converter uses bidirectional switches to replace some clamping diodes, optimizing conduction losses while maintaining three-level output capability; the active neutral point clamped converter introduces active switching devices to replace passive clamping diodes, effectively solving the problem of uneven device loss distribution in NPC topologies and improving DC voltage utilization and fault handling capabilities. Although the above topologies differ in their specific circuit structures, they all possess the core feature of midpoint clamping, namely, the existence of a neutral point on the DC side composed of series capacitors, and the potential of this neutral point needs to be actively controlled to maintain balance. Therefore, the midpoint voltage control method proposed in this embodiment has good compatibility and applicability to the above topologies.

[0078] This method specifically encompasses steps S1 to S5, each executed sequentially and interconnected, constructing a complete closed-loop control system from data acquisition, deviation calculation, current analysis to zero-sequence injection and modulation correction. Through the coordinated operation of these five steps, rapid detection and accurate compensation of the neutral point voltage deviation are achieved, ensuring that the converter maintains dynamic equilibrium of the neutral point potential under different operating conditions. The specific contents of steps S1 to S5 are as follows:

[0079] Step S1: Obtain the current three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower busbars on the DC side of the neutral-point clamped three-level converter.

[0080] Step S2: Determine the midpoint voltage deviation based on the voltage difference between the upper and lower busbars, and calculate the average midpoint current used to eliminate the midpoint voltage deviation.

[0081] As a specific implementation method, when determining the midpoint voltage deviation based on the voltage difference between the upper and lower busbars, the acquired voltage difference between the upper and lower busbars first needs to be low-pass filtered to effectively filter out high-frequency noise interference and extract a smoother and more reliable filtered voltage difference. Then, the absolute value of this filtered voltage difference is compared with a preset voltage dead zone threshold. If the absolute value is less than or equal to the preset voltage dead zone threshold, it indicates that the voltage fluctuation is within the allowable small tolerance range. To avoid unnecessary frequent adjustments, the midpoint voltage deviation is directly set to zero. Conversely, if the absolute value is greater than the preset voltage dead zone threshold, it is determined that there is a substantial voltage shift at the midpoint, and the filtered voltage difference is directly output as the final midpoint voltage deviation.

[0082] In practical implementation, various filtering methods can be used to filter the bus voltage difference, such as first-order low-pass filters, Butterworth filters, moving average filters, or Kalman filters. The filter cutoff frequency is not a fixed value and can be flexibly adjusted within the range of 1kHz to 5kHz according to the noise characteristics of the actual system.

[0083] As one specific implementation method, the average neutral point current I, which is intended to eliminate the neutral point voltage deviation, is calculated. mid-avg At that time, it is necessary to comprehensively consider and base the current midpoint voltage deviation (denoted as Δu) on the following: np The half-bus capacitor parameter (denoted as C) of the midpoint clamped three-level converter and the set sampling period (denoted as Ts) are used to accurately solve the problem. The specific calculation formula can be expressed as I. mid-avg =-C*Δu np / Ts. Based on this calculation mechanism, with the sampling period Ts remaining fixed, the calculated average midpoint current I can be intuitively verified using this formula. mid-avg The absolute value of the average midpoint current exhibits a characteristic of changing in the same direction as the absolute value of the midpoint voltage deviation and the half-bus capacitance parameter, meaning it increases accordingly as both increase. Furthermore, the negative sign in the formula clearly indicates that, in order to effectively achieve the cancellation and correction effects, the target direction of the calculated average midpoint current is set to be opposite to the polarity of the midpoint voltage deviation.

[0084] Step S3: Extract the polarity sign of the three-phase modulation wave, and calculate the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component based on the three-phase modulation wave, polarity sign and actual three-phase current.

[0085] As a specific implementation method, when extracting the polarity sign of the three-phase modulated wave, it is first necessary to perform amplitude limiting processing on each of the three-phase modulated waves to strictly constrain each phase modulated wave within a preset amplitude range to obtain the amplitude-limited modulated wave. Subsequently, the corresponding polarity sign is determined according to the positive or negative state of the amplitude-limited modulated wave of each phase. The specific determination rule is as follows: if the amplitude-limited modulated wave of a certain phase is greater than or equal to zero, then the polarity sign of that phase is defined as positive; if it is less than zero, then it is defined as negative. Based on the obtained polarity sign and three-phase actual current, two types of accurate calculations of the midpoint current are further carried out: First, the current flowing into the midpoint from the non-injected zero-sequence component is calculated according to the three-phase modulation wave, polarity sign, and three-phase actual current. Specifically, the three-phase modulation wave after limiting is first obtained. Then, for each phase of the three-phase midpoint clamped three-level converter, the actual current of the phase, the three-phase modulation wave after limiting, and the corresponding polarity sign are multiplied to obtain the non-injected phase product of the phase. Finally, the non-injected phase products of the three phases are accumulated to obtain the result. Second, the current flowing into the midpoint corresponding to the injected zero-sequence is calculated according to the polarity sign and three-phase actual current. Specifically, for each phase of the three-phase midpoint clamped three-level converter, the actual current of the phase is multiplied by the corresponding polarity sign to obtain the injected phase product of the phase. Then, the injected phase products of the three phases are accumulated to obtain the result.

[0086] Step S4: Calculate the zero-sequence voltage injection amount based on the average neutral point current, the actual three-phase current, the current flowing into the neutral point from the non-injected zero-sequence component, and the current flowing into the neutral point from the injected zero-sequence component.

[0087] As a specific implementation method, when calculating the zero-sequence voltage injection amount, it is necessary to first compare the absolute value of the current flowing into the midpoint corresponding to the injected zero-sequence with a preset current dead zone threshold. If the absolute value is greater than the preset current dead zone threshold, it indicates that there is effective adjustment space. In this case, the sum of the three-phase actual currents needs to be calculated first. Then, combined with the preset midpoint current equation satisfied by the average midpoint current, the sum of the three-phase actual currents, the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence, the original zero-sequence voltage injection amount is derived by algebraic solution. Subsequently, the original zero-sequence voltage injection amount is subjected to amplitude limiting processing to ensure control safety, and finally the required zero-sequence voltage injection amount is obtained. Conversely, if the absolute value of the current flowing into the midpoint corresponding to the injected zero-sequence is less than or equal to the current dead zone threshold, it indicates that the adjustment effect is too small. To avoid introducing unnecessary control interference, the zero-sequence voltage injection amount is directly determined to be zero.

[0088] Specifically, when determining the dead zone, the current dead zone can be either a hard dead zone or a soft dead zone (gradual transition). Regarding the specific threshold, the current dead zone threshold can be set within the range of 1A to 10A, depending on the actual accuracy of the current sensor; meanwhile, the voltage dead zone threshold can be adjusted within the range of 2V to 20V.

[0089] Step S5: The zero-sequence voltage injection is superimposed on the three-phase modulation wave to achieve balanced control of the midpoint voltage.

[0090] As a specific implementation method, when superimposing the zero-sequence voltage injection amount onto the three-phase modulation wave to achieve neutral point voltage balance control, the minimum and maximum values ​​of the three-phase modulation wave are first extracted; then, the zero-sequence voltage injection amount to be superimposed is constrained based on these two extreme values ​​to obtain the constrained zero-sequence voltage injection amount, thereby ensuring that the result of each phase after superposition is within the preset amplitude range; next, the constrained zero-sequence voltage injection amount is superimposed onto the original three-phase modulation wave to generate the target three-phase modulation wave; finally, the precise balance control of the neutral point potential is performed based on the target three-phase modulation wave.

[0091] This embodiment employs a segmented zero-sequence injection method as the basic strategy for midpoint balance control. In practice, this method is not used independently but can be combined with other midpoint balance methods such as the active power direction method and the predicted current position method. By automatically switching between appropriate methods according to different operating conditions, or by weighted mixing of the results from each method, the advantages of different methods can be fully utilized, compensating for the limitations of a single method, thereby achieving a better overall control effect across the entire operating range.

[0092] Specifically, the basic measure for constraint handling is to limit the amplitude of the modulated wave within the interval [-1, 1] to ensure it is in the linear modulation region. Based on this, an additional constraint can be introduced: a limit on the rate of change of the modulated wave, i.e., setting an upper limit on the amplitude of the change between adjacent moments of the modulated wave. Using this method, even if the command value undergoes a step change, the actual output modulated wave will not change abruptly, thereby significantly reducing current surges and improving the dynamic response characteristics of the system.

[0093] As a specific implementation method, the zero-sequence voltage injection amount to be superimposed is subject to boundary constraints based on the minimum and maximum values ​​to obtain the constrained zero-sequence voltage injection amount. This specifically includes the following steps:

[0094] First, based on the polarity sign combination of the three-phase modulated wave, the target sector where the space voltage vector of the current midpoint clamped three-level converter is located is determined. Space vector modulation maps the three-phase modulated wave into a voltage vector on a plane, and different polarity sign combinations correspond to different sector positions. Based on this, according to a preset correspondence table between the target sector and the zero-sequence voltage boundary, the reference boundary value corresponding to the target sector is directly obtained. This reference boundary value reflects the theoretical adjustment range of the zero-sequence voltage in the current sector.

[0095] Secondly, with the constraint that the superimposed three-phase modulated waves do not exceed the preset amplitude range, the reference boundary values ​​are corrected using the aforementioned maximum and minimum values ​​to obtain the physical allowable boundary. This correction process ensures that after the zero-sequence voltage is injected, the modulated waves of each phase remain within the linear modulation region that can be realized by the hardware, avoiding waveform distortion and output performance degradation caused by overmodulation. Subsequently, the maximum allowable absolute value of the physical allowable boundary is extracted as the upper limit of the hardware topology constraint. This upper limit value is jointly determined by the topology of the converter's main circuit and the modulation strategy, and is the physical boundary that the zero-sequence voltage injection cannot incur.

[0096] Next, the dynamic limiting threshold is calculated based on the numerical range of the current absolute value of the midpoint voltage deviation. This dynamic limiting mechanism is designed to balance rapid recovery of the midpoint voltage with a smooth transition of the output waveform. Specifically, when the absolute value of the midpoint voltage deviation is within a preset first numerical range, it indicates a significant deviation in the midpoint potential. In this case, the dynamic limiting threshold is set as the initial limiting value, allowing for a larger zero-sequence voltage injection to accelerate the recovery of the midpoint voltage. When the absolute value of the midpoint voltage deviation is within a preset second numerical range, it indicates that the midpoint potential is close to equilibrium. At this point, the dynamic limiting threshold is reduced according to a preset decreasing rule based on the magnitude of the absolute value of the midpoint voltage deviation, gradually reducing the amount of zero-sequence voltage injection to avoid drastic fluctuations when the midpoint voltage is close to equilibrium. It should be noted that the upper limit of the second numerical range is lower than the lower limit of the first numerical range; the two ranges do not overlap and form a complete numerical coverage, ensuring a clear limiting strategy at any level of deviation.

[0097] Finally, the dynamic limiting threshold is clamped within the upper limit of the hardware topology constraints to obtain the final limiting threshold. This clamping operation ensures that the dynamic adjustment process is always limited to the physically realizable range, preventing over-limit injection due to algorithm calculation anomalies. Subsequently, the zero-sequence voltage injection amount to be superimposed is symmetrically truncated and limited using the final limiting threshold. That is, when the zero-sequence voltage injection amount exceeds the positive and negative final limiting thresholds, it is limited to the boundary value, thus obtaining the constrained zero-sequence voltage injection amount. This constrained zero-sequence voltage injection amount satisfies both the physical constraints of the hardware topology and adapts to the dynamic adjustment requirements of the midpoint voltage deviation, achieving coordinated optimization of control performance and output quality.

[0098] Regarding the specific implementation methods for sector determination, this implementation adopts the following three approaches: First, the lookup table method, which quickly matches sectors using a pre-set table; second, the conditional judgment method, which uses logical conditions to filter and determine sectors sequentially; and third, the sector determination method based on the αβ coordinate system, which directly uses the α and β components of the voltage vector in a two-phase stationary coordinate system for mathematical judgment, without the need for angle calculations or coordinate transformations, and has high execution efficiency in digital control systems. In practical applications, the appropriate method can be flexibly selected based on system computing resources and real-time requirements.

[0099] This embodiment provides a method for controlling the neutral point voltage of a neutral point clamping three-level converter. The core idea is as follows: First, obtain the three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower bus. Based on this, calculate the average neutral point current used to eliminate the neutral point voltage deviation. Then, according to the polarity sign of the modulation wave, decompose the neutral point current components corresponding to non-injected and injected zero-sequence voltages. Next, calculate the zero-sequence voltage injection amount, and finally superimpose it onto the modulation wave to achieve neutral point voltage balance. To more completely and accurately illustrate the implementation process of this control method, the following will combine... Figure 2 The method for segmented zero-sequence injection midpoint voltage balance control of a three-level NPC converter is specifically broken down into multiple steps, including signal acquisition, modulation wave limiting processing, target neutral current calculation, sector determination, midpoint current decomposition, zero-sequence voltage injection calculation, zero-sequence voltage constraint, and zero-sequence voltage superposition output. This comprehensively describes the entire closed-loop control process from signal processing to drive output. The specific steps are as follows:

[0100] (1) Signal acquisition:

[0101] The controller acquires the basic signals required for operation. These basic signals include a three-phase modulated wave m. abc ,m abc Including m a m b and m c , i.e. m abc =[m a ,m b ,m c The basic signal also includes the three-phase actual current i. abc i abc Including i a i b andi c , i.e. i abc =[i a i b i c The basic signal also includes the voltage difference ΔU between the upper and lower buses. dc Half-bus capacitance C and sampling period T s .

[0102] (2) Modulation wave limiting processing:

[0103] To ensure the modulated wave remains within a legal control range and prevent overmodulation, the acquired original three-phase modulated wave is amplitude-limited. The three-phase modulated wave is restricted to the normalized range [-1, 1], meaning that for k ∈ {a, b, c}, the amplitude-limited modulated wave m... limited[k] =max(-1,min(1,mabc[k])). This step ensures that the input signals for sector determination and current calculation are valid.

[0104] (3) Calculation of target centerline current:

[0105] Based on the collected voltage difference Δu between the upper and lower busbars dc This allows us to deduce the neutral current required to balance the midpoint potential. Using the volt-ampere characteristic of a capacitor, we can calculate the required average neutral current i. mid_avg =-Δu dc ×C / Ts. This current value represents the average current flowing out of or into the midpoint in the next sampling period to eliminate the current bus voltage deviation. The physical meaning of this current is: within a sampling period, the average current flowing out of or into the midpoint is required to compensate for the current midpoint voltage deviation; its polarity is opposite to the voltage deviation. If the upper bus voltage is too high, current needs to be drawn from the midpoint; if the lower bus voltage is too high, current needs to be injected into the midpoint.

[0106] (4) Sector determination:

[0107] Based on the amplitude-limited three-phase modulated wave, through the sign function s[k]=sign(m limited [k]) Determine the polarity sign of each phase modulation wave, and according to (s a ,s b ,s c The combination of these factors determines the current working sector N∈{1,2,3,4,5,6}. Specifically, the correspondences are as follows: sector 1 corresponds to (+,-,-), sector 2 to (+,+,-), sector 3 to (-,+,-), sector 4 to (-,+,+), sector 5 to (-,-,+), and sector 6 to (+,-,+). Sector determination is used to determine the connection relationship between each phase modulation wave and the DC midpoint, and is a prerequisite for midpoint current decomposition.

[0108] (5) Decomposition of midpoint current:

[0109] The midpoint current is decomposed into two independent components. One is the non-injected zero-sequence component, i, which flows into the midpoint. in This reflects the current component that naturally flows through the midpoint when no zero-sequence voltage is injected, due to the non-zero nature of the three-phase modulation wave itself. The calculation formula is i.in =∑ k∈{a,b,c} (i abc [k]×(-m limited [k]×s[k])); the second is the current i injected into the midpoint corresponding to the zero sequence. in0 This reflects the ability to regulate the midpoint current when a unit zero-sequence voltage is injected; the calculation formula is i. in0 =∑ k∈{a,b,c} (i abc [k]×s[k]) .

[0110] (6) Calculation of zero-sequence voltage injection:

[0111] Determine the absolute value of the injected zero-sequence current |i in0 | Is it greater than the current dead zone threshold i? threshold . If i in0 |>i threshold This indicates that the zero-sequence control capability is sufficient. Therefore, the original zero-sequence voltage injection quantity v can be calculated by algebraic solution. 0_raw =(-2×i mid_avg -i a -i b -i c +i in ) / i in0 And perform amplitude limiting on it to obtain v0=max(-v 0_lim ,min(v 0_lim ,v 0_raw )), where v 0_lim The preset zero-sequence voltage limit value; if |i in0 |≤i threshold This indicates that the zero-sequence control capability is insufficient or in the dead zone. To avoid ineffective or erroneous control intervention, the zero-sequence voltage injection amount is directly set to v0=0.

[0112] The following section uses a three-level photovoltaic grid-connected inverter as an example to explain in detail the control process of zero-sequence voltage limiting. Figure 3 As shown, the entire process can be divided into the following four steps:

[0113] Step 1: Operating condition identification and initial amplitude limit setting.

[0114] This step sets the control tone, serving as the starting point for the control process. It is based on real-time monitoring of the power grid status, setting an initial limit reference value that matches the operating conditions for subsequent adjustments.

[0115] When the inverter is in the grid-connected startup phase, in order to prevent overly aggressive balance control from causing current surges, the zero-sequence voltage limit v0_lim is preset to a small conservative value (e.g., 0.01) to ensure a smooth startup process.

[0116] After entering steady-state grid-connected operation, in order to obtain sufficient midpoint balance adjustment capability, v0_lim will switch to the designed normal value (0.16).

[0117] If a grid fault is detected and a low-voltage ride-through (LVRT) is triggered, the midpoint voltage fluctuation will be significantly aggravated because a large amount of reactive current needs to be injected into the grid at this time. To address this, the control strategy will proactively increase v0_lim by 0.03 (to 0.19) to enhance balancing capability and suppress voltage deviation.

[0118] Step 2: Dynamic fine-tuning.

[0119] This step involves elastic limiting based on the degree of imbalance. In the main mode of normal grid-connected operation, v 0_lim It is not fixed, but rather dynamically and precisely adjusted based on the actual degree of imbalance in the DC bus voltage. The adjustment logic is as follows:

[0120] First, the original bus voltage difference Δu was sampled. dc_raw Preprocessing is performed to obtain reliable control data:

[0121] Low-pass filtering: High-frequency switching noise is filtered out by using a low-pass filter with a cutoff frequency of 2.5kHz to obtain a smooth Δu. dc_filter .

[0122] Dead zone handling: To eliminate the influence of minute measurement fluctuations, a 5V dead zone is set. Only when |Δu dc_filter A valid voltage imbalance is considered to exist only when the voltage reaches 5V, and Δu is set accordingly. dc =Δu dc_filter Otherwise, consider Δu dc = 0.

[0123] Subsequently, based on the absolute value of the processed bus voltage difference |Δu dc |, for the initial v 0_lim (0.16) Perform hierarchical scaling:

[0124] If there is severe imbalance (|Δu) dc |>50V): Keep v 0_lim = 0.16. At this point, the midpoint voltage deviation is severe, and the controller needs to retain the maximum adjustment authority to quickly correct the imbalance.

[0125] If there is a moderate imbalance (40V <|Δu) dc | ≤ 50V): Reduce the limit value to 0.112 (0.16 × 0.7). While applying strong control, begin to converge moderately to prevent over-adjustment.

[0126] If there is a slight imbalance (30V < |Δu) dc | ≤ 40V): Further reduced to 0.08 (0.16 × 0.5). At this point, the imbalance is limited, and a gentler adjustment is used, prioritizing stability.

[0127] If the basic equilibrium is (|Δu) dc | ≤ 30V): Significantly reduced to 0.016 (0.16×0.1). The midpoint voltage is close to the ideal state, and the controller switches to "fine-tuning" mode, retaining only a very small adjustment capability to smooth out minor disturbances, effectively avoiding frequent actions and energy waste near the equilibrium point.

[0128] Step 3: Zero-sequence voltage calculation and execution.

[0129] This step is the final decision-making stage for the injection amount, and specifically includes the following:

[0130] First, obtain the zero-sequence voltage boundary value v. 0_lim_final In obtaining v 0_lim_final After that, the control process enters the injection volume decision and execution stage.

[0131] The core of this step is to determine whether the control capability of the zero-sequence current path is effective, and based on this, to conduct capability assessment, decision calculation, and final execution.

[0132] (1) Capability assessment: Calculate the absolute value of the zero-sequence injection current |i in0 | and with the preset current dead zone threshold i _threshold (3A) Compare.

[0133] (2) Decision-making and calculation:

[0134] if |i in0 | > i _threshold This indicates that the zero-sequence channel has sufficient controllability. At this point, through formula v... 0_raw =(-2×i mid_avg -i a -i b -i c +i in ) / i in0 The theoretically required zero-sequence voltage is obtained through algebraic solution. For safety, this calculated value is constrained to [-V]. 0_lim_final ,+v 0_lim_final Within the range of ], that is: v0 = max(-v 0_lim_final min(v) 0_lim_final ,v 0_raw )).

[0135] if |i in0 | ≤ i _thresholdThis means that the zero-sequence control capability is insufficient or that the system is in a control dead zone. In this case, to avoid injecting invalid or even harmful zero-sequence voltage, v0 is directly set to 0, and the midpoint balance intervention is suspended.

[0136] (3) Execution: The final determined v0 value is sent to the pulse width modulation (PWM) module to generate the corresponding switching signal, thereby accurately injecting the required zero-sequence voltage component into the inverter output to achieve active control of the midpoint potential.

[0137] Step 4: Closed-loop continuous optimization.

[0138] The entire control process constitutes a high-speed, self-updating dynamic closed loop. Each time a zero-sequence voltage injection is completed, the DC-side bus voltage is immediately resampled to obtain the latest voltage difference information. This updated voltage difference is then sent to the next control cycle, initiating a new round of the complete process: starting with identifying the current operating condition, then dynamically and finely adjusting the limiting boundary based on the real-time degree of voltage imbalance, and finally evaluating the control capability of the zero-sequence channel and deciding on the final voltage injection amount.

[0139] This process repeats itself continuously. It's like an intelligent loop with perception and decision-making capabilities, able to track the imbalance state of the DC bus in real time and adaptively adjust the strength and boundaries of its control strategy accordingly. It is through this continuous self-observation and dynamic adjustment that this control method can consistently and reliably maintain the balance of the midpoint potential under various complex and changing operating conditions.

[0140] Specifically, zero-sequence voltage limiting can be performed using either fixed limiting or dynamic limiting. For dynamic limiting, the segmented thresholds can be flexibly adjusted according to the characteristics of the actual system. Alternatively, a continuous function (such as linear interpolation or an S-curve) can be used instead of a segmented function to obtain more continuous limiting characteristics.

[0141] This control process constructs an adaptive closed-loop regulation mechanism based on operating condition identification and status feedback. First, different zero-sequence voltage limiting benchmarks are preset according to different operating stages, such as grid-connected startup, normal operation, or low-voltage ride-through, to balance startup smoothness and the strong balance requirements during fault ride-through. During normal steady-state operation, further dynamic and fine-tuning is performed based on the degree of DC bus voltage imbalance. By scaling the limiting value in stages, maximum regulation authority is retained when the voltage is severely unbalanced, while a fine-tuning mode is entered when the voltage is close to balance to avoid excessive intervention. Finally, the capability assessment of the zero-sequence current path is introduced in the execution stage. The calculated voltage, constrained by limiting, is injected only when regulation is effective; otherwise, zero output is maintained, thus ensuring the accuracy and safety of control. The entire process repeats continuously, tracking the system status in real time and adaptively adjusting the strategy, achieving stable and reliable control of the midpoint potential under various operating conditions.

[0142] (7) Zero-sequence voltage constraint:

[0143] The calculated zero-sequence voltage is constrained to ensure that the superimposed modulated wave does not exceed the amplitude limit range of [-1, 1]. First, the minimum value m of the original modulated wave is calculated. min =min(m a ,m b ,m c ) and maximum value m max =max(m a ,m b ,m c Then, through formula v 0_final =max(-1-m min ,min(1-m max The final zero-sequence voltage is determined by v0. Where 1-m max For the remaining space where the modulated wave moves upward, -1-m min To allow for the remaining space to move downwards, this step implements overmodulation protection, ensuring that the final modulated wave is always in the linear modulation region.

[0144] (8) Zero-sequence voltage superposition output:

[0145] The constrained final zero-sequence voltage v 0_final Equal amounts are superimposed on the three-phase modulated wave, i.e., m a_new =m a +v 0_final m b_new =m b +v 0_final m c_new =m c +v 0_finalSince the three phases are simultaneously shifted by the same zero-sequence voltage value, the relative relationship between the three phases remains unchanged, thus not affecting the output line voltage. However, it alters the connection time between each phase and the DC midpoint, thereby controlling the direction and magnitude of the midpoint current flow. The superimposed outputs the final modulated wave, which drives the power switching devices of the NPC converter, completing the closed loop of midpoint voltage balance control.

[0146] The midpoint voltage control method for a midpoint clamping three-level converter provided in this embodiment achieves balanced control of the midpoint voltage by acquiring the three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower DC buses, combined with midpoint current decomposition and zero-sequence voltage calculation. In terms of response speed, this method uses a zero-sequence voltage calculation method based on instantaneous current, eliminating the need for complex PI regulators or predictive controllers. It can complete the calculation and output the control quantity within a single sampling period, resulting in extremely fast response. Regarding control accuracy, by segmented judgment and midpoint current decomposition, the required zero-sequence injection amount is accurately calculated, enabling the midpoint voltage to quickly converge to a balanced state. Compared to traditional methods, midpoint voltage fluctuation can be reduced by more than 30%, resulting in superior balancing performance. In terms of robustness, the introduction of current dead-zone thresholds and voltage dead-zone thresholds effectively avoids numerical instability issues under small currents or small voltage differences. Furthermore, by dynamically adjusting the v0 limit, it can adapt to different operating conditions. The algorithm meets the control requirements of various operating conditions. In terms of adaptability, it automatically adjusts the zero-sequence voltage injection level based on the bus voltage difference. When the voltage deviation is small, the injection amount is reduced to decrease harmonic effects; when the deviation is large, the injection amount is increased to quickly balance the midpoint voltage, achieving an optimal trade-off between balancing effect and harmonic performance. Regarding operating condition adaptability, differentiated v0 limiting strategies are set for different operating conditions such as normal operation, low-voltage ride-through, and grid-connected startup, ensuring good midpoint balancing effects under each condition. In terms of computation, the algorithm mainly involves addition, subtraction, multiplication, division, and comparison operations, requiring no trigonometric functions or complex matrix operations. It is easy to implement in real-time on hardware such as DSPs or FPGAs, with a single calculation time of less than 5μs, meeting real-time control requirements. Furthermore, this method is compatible with existing modulation strategies; zero-sequence voltage injection is compatible with various modulation strategies such as SPWM, SVPWM, and DPWM, facilitating upgrades and modifications to existing systems and demonstrating broad applicability.

[0147] Example 2:

[0148] This embodiment provides a midpoint voltage control device for a midpoint clamping three-level converter, the overall structure of which is as follows: Figure 4 As shown. The device is composed of an acquisition unit 10, a first calculation unit 20, a second calculation unit 30, a third calculation unit 40, and a control unit 50 working together. Each unit is connected and cooperates in sequence to jointly complete the balance control function of the midpoint voltage.

[0149] The acquisition unit 10 serves as the information input terminal for the entire device, responsible for collecting the basic data required for system operation, including the current three-phase modulation wave, the actual three-phase current output by the converter, and the voltage difference between the upper and lower bus capacitors on the DC side of the neutral-clamped three-level converter, providing the original data basis for subsequent calculations and processing by each unit.

[0150] The first calculation unit 20 is connected to the acquisition unit 10. Its function is to accurately determine the current midpoint voltage deviation based on the acquired upper and lower bus voltage difference, and further calculate the average midpoint current required to eliminate the midpoint voltage deviation, thereby clarifying the target requirements for midpoint voltage regulation.

[0151] The second calculation unit 30 is connected to the first calculation unit 20 and undertakes a more refined current component analysis task. This unit first extracts the polarity sign of the three-phase modulation wave, and then integrates the three-phase modulation wave itself, the polarity sign, and the actual three-phase currents to accurately calculate two specific current indices: the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence component. This lays the foundation for the accurate solution of the subsequent zero-sequence voltage injection.

[0152] The third calculation unit 40 is connected to the second calculation unit 30, and its core function is to integrate and calculate the current parameters of the aforementioned stages. This unit gathers four pieces of information: the average neutral point current, the actual three-phase current, the current flowing into the neutral point from the non-injected zero-sequence component, and the current flowing into the neutral point from the injected zero-sequence component. Through comprehensive calculation of these parameters, the required zero-sequence voltage injection amount is finally calculated, thereby realizing the quantitative determination of the control quantity.

[0153] The control unit 50 is connected to the third calculation unit 40 and the acquisition unit 10, respectively, as the final execution link of the device. This unit accurately superimposes the calculated zero-sequence voltage injection amount onto the original three-phase modulation wave to generate a new target modulation wave. By correcting the modulation wave, the switching state of the converter is regulated, thereby effectively achieving the neutral point voltage balance control target.

[0154] It should be noted that the device in this embodiment, through the synergistic effect of the aforementioned units, can completely and effectively realize all the functions of the method described in Embodiment 1. The device adopts a modular design, decomposing the complex control algorithm into multiple relatively independent but interconnected calculation steps, which ensures control accuracy while facilitating engineering implementation, debugging, and maintenance.

[0155] Example 3:

[0156] like Figure 5As shown, this embodiment provides a midpoint clamping three-level converter. The overall architecture of the converter mainly consists of three core parts: the converter main circuit, the sampling and detection circuit, and the midpoint voltage control device described in Embodiment 2 above.

[0157] In terms of hardware configuration, the DC side of the converter's main circuit includes an upper bus capacitor and a lower bus capacitor connected in series. Their common connection point forms the circuit's neutral point, serving as a crucial potential reference. To achieve closed-loop control, a sampling and detection circuit is connected to the main circuit. Its function is to acquire the actual three-phase current during the main circuit's operation in real time and accurately detect the voltage difference between the upper and lower bus capacitors on the DC side, providing feedback information for control.

[0158] As the core control unit of the entire midpoint clamping three-level converter system, the midpoint voltage control device is connected to both the converter's main circuit and the sampling and detection circuit. During operation, the device first receives voltage difference and three-phase current data from the sampling and detection circuit. Then, based on this information, it generates a target three-phase modulation wave superimposed with a specific zero-sequence voltage injection. Finally, based on this target modulation wave, the device generates a corresponding drive signal and precisely modulates the converter's main circuit, thereby effectively achieving balanced control of the midpoint voltage and ensuring stable and reliable converter operation.

[0159] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for controlling the midpoint voltage of a midpoint-clamped three-level converter, characterized in that, include: Obtain the current three-phase modulation waveform, three-phase actual current, and the voltage difference between the upper and lower busbars on the DC side of the neutral-point clamped three-level converter; The midpoint voltage deviation is determined based on the voltage difference between the upper and lower busbars, and the average midpoint current used to eliminate the midpoint voltage deviation is calculated. Extract the polarity sign of the three-phase modulated wave, and calculate the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component based on the three-phase modulated wave, the polarity sign, and the actual three-phase current. The zero-sequence voltage injection amount is calculated based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence component. The zero-sequence voltage injection amount is superimposed on the three-phase modulation wave to achieve balanced control of the midpoint voltage.

2. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 1, characterized in that, The determination of the midpoint voltage deviation based on the voltage difference between the upper and lower busbars specifically includes: The voltage difference between the upper and lower busbars is subjected to low-pass filtering to obtain the filtered voltage difference; Based on the filtered voltage difference, the midpoint voltage deviation is determined: if the absolute value of the filtered voltage difference is less than or equal to a preset voltage dead zone threshold, the midpoint voltage deviation is set to zero; if the absolute value of the filtered voltage difference is greater than the voltage dead zone threshold, the filtered voltage difference is used as the midpoint voltage deviation.

3. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 1, characterized in that, The calculation of the average midpoint current used to eliminate the midpoint voltage deviation specifically includes: The average midpoint current is calculated based on the midpoint voltage deviation, the half-bus capacitor parameters of the midpoint clamping three-level converter, and the sampling period. The target direction of the average midpoint current is opposite to the polarity of the midpoint voltage deviation.

4. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 1, characterized in that, The extraction of the polarity sign of the three-phase modulated wave specifically includes: The three-phase modulated waves are each subjected to amplitude limiting processing to ensure that each phase modulated wave is within a preset amplitude range, thereby obtaining the amplitude-limited modulated wave. Based on the sign of the modulation wave after amplitude limiting in each phase of the three-phase modulation wave, the polarity sign of each phase is determined respectively; wherein, if the modulation wave after amplitude limiting of a certain phase is greater than or equal to zero, the polarity sign of that phase is determined to be positive; if the modulation wave after amplitude limiting of a certain phase is less than zero, the polarity sign of that phase is determined to be negative.

5. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 1, characterized in that, The step of calculating the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint corresponding to the injected zero-sequence component based on the three-phase modulation wave, the polarity sign, and the three-phase actual current specifically includes: calculating the current flowing into the midpoint of the non-injected zero-sequence component based on the three-phase modulation wave, the polarity sign, and the three-phase actual current; and calculating the current flowing into the midpoint corresponding to the injected zero-sequence component based on the polarity sign and the three-phase actual current. The step of calculating the current flowing into the midpoint from the non-injected zero-sequence component based on the three-phase modulation wave, the polarity sign, and the actual three-phase current specifically includes: The three-phase modulated wave is subjected to amplitude limiting to obtain the amplitude-limited three-phase modulated wave; For each phase of the three phases of the midpoint clamped three-level converter, the actual current of that phase, the three-phase modulation wave after limiting of that phase, and the corresponding polarity sign are multiplied together to obtain the non-injected phase product of that phase. The non-injection phases of the three phases are multiplied and accumulated to obtain the current flowing into the midpoint by the non-injection zero-sequence component. The step of calculating the current flowing into the midpoint corresponding to the zero-sequence injection, based on the polarity sign and the actual three-phase current, specifically includes: For each phase of the three-phase neutral-clamped three-level converter, the actual current of that phase is multiplied by the corresponding polarity sign to obtain the injected phase product. By multiplying and summing the injections of the three phases, the current flowing into the midpoint corresponding to the zero-sequence injection is obtained.

6. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 1, characterized in that, The calculation of the zero-sequence voltage injection amount based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint from the non-injected zero-sequence component, and the current flowing into the midpoint from the injected zero-sequence component specifically includes: The absolute value of the current flowing into the midpoint corresponding to the zero-sequence injection is compared with a preset current dead zone threshold, and the zero-sequence voltage injection amount is calculated based on the comparison result: If the absolute value of the current flowing into the midpoint corresponding to the injected zero sequence is greater than the preset current dead zone threshold, then the sum of the three-phase actual currents is calculated, and the preset midpoint current equation relationship satisfied by the average midpoint current, the sum of the three-phase actual currents, the current flowing into the midpoint of the non-injected zero sequence component, and the current flowing into the midpoint corresponding to the injected zero sequence is solved algebraically to obtain the original zero sequence voltage injection amount, and the original zero sequence voltage injection amount is limited to obtain the zero sequence voltage injection amount. If the absolute value of the current flowing into the midpoint corresponding to the injected zero sequence is less than or equal to the current dead zone threshold, then the zero sequence voltage injection amount is determined to be zero.

7. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to any one of claims 1 to 6, characterized in that, The step of superimposing the zero-sequence voltage injection amount onto the three-phase modulation wave to achieve neutral point voltage balance control specifically includes: Extract the minimum and maximum values ​​from the three-phase modulated wave; Based on the minimum and maximum values, boundary constraints are applied to the zero-sequence voltage injection amounts to be superimposed, resulting in constrained zero-sequence voltage injection amounts, so that the superimposed phase results are all within a preset amplitude range. The constrained zero-sequence voltage injection amount is superimposed on the three-phase modulation wave to obtain the target three-phase modulation wave; The neutral point voltage is balanced and controlled according to the target three-phase modulation wave.

8. The method for controlling the midpoint voltage of a midpoint-clamped three-level converter according to claim 7, characterized in that, The step of applying boundary constraints to the zero-sequence voltage injection amount to be superimposed based on the minimum and maximum values ​​to obtain the constrained zero-sequence voltage injection amount specifically includes: Based on the polarity sign combination of the three-phase modulation wave, the target sector where the space voltage vector of the current midpoint clamping three-level converter is located is determined, and the reference boundary value corresponding to the target sector is obtained according to the preset correspondence table between the target sector and the zero-sequence voltage boundary. With the constraint that the superimposed three-phase modulated wave does not exceed the preset amplitude range, the reference boundary value is corrected using the maximum and minimum values ​​to obtain the physical allowable boundary, and the maximum allowable absolute value of the physical allowable boundary is extracted as the upper limit of the hardware topology constraint. The dynamic limiting threshold is calculated based on the numerical range of the absolute value of the current midpoint voltage deviation. When the absolute value of the midpoint voltage deviation is within a preset first numerical range, the dynamic limiting threshold is determined as the initial limiting value. When the absolute value of the midpoint voltage deviation is within a preset second numerical range, the dynamic limiting threshold is reduced according to the magnitude of the absolute value of the midpoint voltage deviation and a preset decreasing rule. The upper limit of the second numerical range is less than the lower limit of the first numerical range. The dynamic limiting threshold is clamped within the upper limit of the hardware topology constraint to obtain the final limiting threshold. The zero-sequence voltage injection amount to be superimposed is symmetrically truncated and limited using the final limiting threshold to obtain the constrained zero-sequence voltage injection amount.

9. A midpoint voltage control device for a midpoint clamping three-level converter, characterized in that, include: The acquisition unit is used to acquire the current three-phase modulation wave, the actual three-phase current, and the voltage difference between the upper and lower busbars on the DC side of the neutral-point clamped three-level converter. The first calculation unit, connected to the acquisition unit, is used to determine the midpoint voltage deviation based on the voltage difference between the upper and lower busbars, and to calculate the average midpoint current used to eliminate the midpoint voltage deviation. The second calculation unit, connected to the first calculation unit, is used to extract the polarity sign of the three-phase modulation wave, and calculate the current flowing into the midpoint of the non-injected zero-sequence component and the current flowing into the midpoint of the injected zero-sequence component based on the three-phase modulation wave, the polarity sign and the actual three-phase current. The third calculation unit, connected to the second calculation unit, is used to calculate the zero-sequence voltage injection amount based on the average midpoint current, the actual three-phase current, the current flowing into the midpoint of the non-injected zero-sequence component, and the current flowing into the midpoint corresponding to the injected zero-sequence component. The control unit, connected to the third calculation unit and the acquisition unit respectively, is used to superimpose the zero-sequence voltage injection amount onto the three-phase modulation wave to achieve balanced control of the midpoint voltage.

10. A midpoint clamping three-level converter, characterized in that, include: The converter main circuit, the sampling and detection circuit, and the midpoint voltage control device of the midpoint clamping three-level converter as described in claim 9; The DC side of the converter main circuit includes an upper bus capacitor and a lower bus capacitor connected in series, and the connection point of the upper bus capacitor and the lower bus capacitor constitutes a neutral point. The sampling and detection circuit is connected to the main circuit of the converter and is used to collect the actual three-phase current of the main circuit of the converter and the voltage difference between the upper and lower busbars on the DC side. The midpoint voltage control device is connected to the converter main circuit and the sampling and detection circuit respectively. It is used to generate a target three-phase modulation wave with zero-sequence voltage injection superimposed on it based on the voltage difference between the upper and lower bus and the actual three-phase current collected by the sampling and detection circuit. It also generates a drive signal based on the target three-phase modulation wave to modulate the converter main circuit, thereby realizing the balance control of the midpoint voltage.

Images