Parallel inverter circulating current prediction control method

Abstract

The present application relates to the technical field of parallel inverter control, and particularly relates to a parallel inverter circulating current prediction control method. The present application establishes mathematical models of a single inverter and a virtual inverter, and controls the current and zero sequence circulating current through a current prediction model and a circulating current prediction model, wherein the inner layer is used for controlling the current, and the outer layer is used for inhibiting the zero sequence circulating current, so as to calculate the optimal output voltage and common-mode voltage at the next moment in real time, to meet the current tracking and zero sequence circulating current inhibition, and to realize the complete inhibition of the zero sequence circulating current under the premise of ensuring accurate current tracking. Without relying on complex communication or control parameter design between inverters, the present application can actively eliminate the steady-state zero sequence circulating current error, and greatly improves the theoretical inhibition effect of the zero sequence circulating current.

Description

Technical Field

[0001] This invention relates to the technical field of parallel inverter control, and in particular to a method for predictive control of circulating current in a parallel inverter. Background Technology

[0002] As the demand for modern power electronic systems continues to grow across various applications, the need for higher power capacity is also increasing. Multi-parallel inverters, due to their ability to expand system power capacity and meet plug-and-play requirements, are widely used as a key interface between renewable energy systems and the power grid.

[0003] However, the common-mode voltage (CMV) difference between parallel inverters will generate zero-sequence circulating current (ZSCC) in the parallel inverter loop, resulting in output current distortion, increased system power loss, and even system collapse.

[0004] To eliminate zero-sequence circulating current between parallel inverters, various control methods have been proposed in the prior art.

[0005] The literature [T. Liu, A. Chen, and Y. Huang, “Multi-vector model predictive current control for paralleled three-level t-type inverters with circulating current elimination,” IEEE Trans. Ind. Electron., vol. 70, no. 8, pp. 8042-8052, 2023.] proposes a multi-vector zero-average common-mode voltage method based on volt-second balance theory (referred to as Method 1). In each control cycle, three voltage vectors are applied, and a zero common-mode voltage vector is synthesized based on the volt-second balance principle to further improve control accuracy and suppress zero-sequence circulating current. After applying this method, the current tracking accuracy of the parallel inverter is improved to the same level as that of a single inverter. However, Method 1 only ignores the actual steady-state zero-sequence circulating current error caused by factors such as control errors. A zero common-mode voltage only ensures that the circulating current no longer increases, but it cannot actively eliminate existing circulating current, leading to continuous accumulation of circulating current.

[0006] To address the issue that Method 1 cannot eliminate steady-state zero-sequence circulating current, the literature [Y. Li, H. He, Z. Li, and Z. Zhang, “Predictive zero-sequence current control of multiple paralleled power converters,” IEEE Trans. Ind. Electron., vol. 69, no. 12, pp. 11 868-11 878, 2022.] analyzes the influence of different voltage vectors on zero-sequence circulating current and proposes a synthetic voltage vector method (referred to as Method 2) composed of the optimal contracted voltage vector and the zero voltage vector to achieve direct zero-sequence circulating current control, effectively eliminating steady-state zero-sequence circulating current error. However, this method requires the design of voltage vector contraction parameters, resulting in low control accuracy of the circulating current and high dependence on the design of control parameters.

[0007] However, these circulating current predictive control methods rely on dense data exchange between inverters to achieve predictive control of zero-sequence circulating current. This not only increases the complexity of the control process, but also limits the effectiveness of circulating current suppression to communication speed. Summary of the Invention

[0008] The purpose of this invention is to provide a circulating current prediction control method for parallel inverters, which can actively eliminate steady-state zero-sequence circulating current error without relying on complex communication or control parameter design between inverters, simplify the control process, and improve the suppression effect of zero-sequence circulating current.

[0009] To address the aforementioned technical problems, this invention provides a circulating current prediction control method for parallel inverters.

[0010] The parallel inverter circulating current predictive control method of the present invention includes:

[0011] A mathematical model of a single inverter in a parallel inverter system in the α-β stationary coordinate system is established to obtain the circulating current and common-mode voltage of a single inverter.

[0012] The other parallel inverters in the parallel inverter system are equivalent to a virtual inverter, and the equivalent zero-sequence circulating current, equivalent common-mode voltage, and zero-sequence circulating current between a single inverter and the virtual inverter are obtained.

[0013] A current prediction model is established and a prediction model for the desired voltage is derived from it. Based on the prediction model for the desired voltage, the desired voltage vector that satisfies the current tracking target is calculated.

[0014] Calculate the duration of action of the non-zero voltage vector used to synthesize the desired voltage vector;

[0015] Establish a circulating current prediction model, and based on the circulating current prediction model, calculate the optimal common-mode voltage that satisfies the condition of reducing the zero-sequence circulating current to zero in the next cycle;

[0016] Calculate the zero-voltage vector action time that satisfies the optimal common-mode voltage;

[0017] Based on the duration of the non-zero voltage vector and the duration of the zero voltage vector, an output voltage vector is synthesized and applied to the next control cycle of the parallel inverter system.

[0018] Furthermore, the parallel inverter system is defined as an n-parallel inverter system, and the filter inductance L of the virtual inverter is... v for:

[0019] ;

[0020] The equivalent zero-sequence circulating current i of the virtual inverter cc,v for:

[0021] ;

[0022] The equivalent common-mode voltage v of the virtual inverter cmv,v for:

[0023] ;

[0024] Where x refers to the serial number of any inverter in the parallel inverter system, and L j i is the filter inductance for the remaining inverters in the parallel inverter system; cc,x The zero-sequence circulating current of the x-th inverter; i cc,j For the zero-sequence circulating current of any other inverter in the parallel inverter system; v cmv,j This is the common-mode voltage of any other inverter in the parallel inverter system.

[0025] Furthermore, the voltage equivalent circuit model of the x-th inverter and the virtual inverter is as follows:

[0026] ;

[0027] The zero-sequence circulating current between the x-th inverter and the virtual inverter is:

[0028] ;

[0029] Among them, v cmv,x L is the common-mode voltage of the xth inverter; x Let be the filter inductance of the xth inverter.

[0030] Furthermore, by discretizing the mathematical model of the single inverter in the α-β stationary coordinate system, a current prediction model is established. In the current prediction model, the predicted current for the (k+1)th control cycle is... for:

[0031] ;

[0032] Among them, T s L represents the sampling period of the control system. x Let be the filter inductance of the x-th inverter; Let x be the output voltage of the x-th inverter in the α-β stationary coordinate system; The grid voltage is in the α-β stationary coordinate system;

[0033] Based on the current prediction model, a prediction model for the desired voltage is established. In the voltage prediction model, in the (k+1)th control cycle, the reference value of the output voltage vector is... for:

[0034] ;

[0035] in, Here is the reference value for the predicted current of the x-th inverter in the α-β stationary coordinate system; Let be the grid voltage of the x-th inverter in the α-β stationary coordinate system.

[0036] Furthermore, the current control target for the xth inverter is:

[0037] ;

[0038] in, Let x be the current control target for the x-th inverter. Let be the reference output voltage of the x-th inverter; Let be the output voltage of the xth inverter.

[0039] Furthermore, based on the volt-second balance theory, the non-zero voltage vector v in the α-β coordinate system... m and v n The projection mapping relationship with the output voltage vector is as follows:

[0040] ;

[0041] Non-zero voltage vector v m and v n The duration of action t m and t n They are respectively:

[0042] ;

[0043] Among them, t m The non-zero voltage vector v m The duration of action, t n The non-zero voltage vector v n Duration of action, v xα v is the α component of the output voltage vector; xβ The β component of the output voltage vector; v mα For voltage vector v m α component; v nα For voltage vector v n α component; v mβ For voltage vector v m The β component; v nβ For voltage vector v n The β component.

[0044] Furthermore, by discretizing the circulating current equation and common-mode voltage equation of the x-th inverter in the α-β stationary coordinate system, a circulating current prediction model is established, which is as follows:

[0045] ;

[0046] in, This represents the circulating current of the x-th inverter. This represents the circulating current of other inverters.

[0047] Furthermore, based on the circulating current prediction model, the optimal common-mode voltage for zeroing the zero-sequence circulating current of the x-th inverter and other inverters in the (k+1)th control cycle is:

[0048] ;

[0049] in, Let be the optimal common-mode voltage at which the zero-sequence circulating current of the x-th inverter returns to zero in the (k+1)th control cycle. The optimal common-mode voltage for other inverters to return to zero in the (k+1)th control cycle is the zero-sequence circulating current.

[0050] Furthermore, the circulating current control objective for the xth inverter is:

[0051] ;

[0052] in, For the circulating current control target of the xth inverter, v cmv,x Let be the common-mode voltage of the x-th inverter in the (k+1)th control cycle.

[0053] Furthermore, in the α-β stationary coordinate system, the common-mode voltage of the x-th inverter is:

[0054] ;

[0055] in, The non-zero voltage vector v m The common-mode voltage; The non-zero voltage vector v n The common-mode voltage; The common-mode voltage is the zero-voltage vector v0; The common-mode voltage of the zero voltage vector v7;

[0056] The required durations t0 and t7 for the zero voltage vectors v0 and v7 are respectively:

[0057] ;

[0058] Where t0 is the duration of the zero voltage vector v0, and t7 is the duration of the zero voltage vector v7; v dc This is the DC side voltage.

[0059] Compared with the prior art, the present invention has at least the following beneficial effects:

[0060] This invention establishes mathematical models for a single inverter and a virtual inverter, and performs hierarchical predictive control of current and zero-sequence circulating current through current prediction model and circulating current prediction model. The inner layer is used to control the current, and the outer layer is used to suppress zero-sequence circulating current, thereby calculating the optimal output voltage and common-mode voltage at the next moment in real time to meet the requirements of current tracking and zero-sequence circulating current suppression, and achieving complete suppression of zero-sequence circulating current under the premise of ensuring accurate current tracking.

[0061] This invention actively eliminates steady-state zero-sequence circulating current error without relying on complex inter-inverter communication or control parameter design, simplifying the control process and significantly improving the theoretical suppression effect of zero-sequence circulating current. This invention not only effectively solves the problems of traditional zero-sequence circulating current suppression methods failing to eliminate steady-state zero-sequence circulating current error and incomplete zero-sequence circulating current suppression, but also relies solely on the hardware parameters of each inverter and the current and voltage data sampled by current sensors. It eliminates the need for inter-inverter communication to obtain information from other inverters and the need to design control parameters. Attached Figure Description

[0062] Figure 1 This is a flowchart illustrating an embodiment of the parallel inverter circulating current predictive control method of the present invention;

[0063] Figure 2 This is a topology diagram of the parallel inverter system targeted by the parallel inverter circulating current predictive control method of the present invention.

[0064] Figure 3 for Figure 2 Equivalent circuit diagram of a parallel inverter system in China;

[0065] Figure 4 This is a schematic diagram of the voltage vector and its corresponding common-mode voltage in the parallel inverter circulating current prediction control method of the present invention;

[0066] Figure 5 This is a schematic diagram illustrating the current tracking and zero-sequence circulating current suppression principle of the parallel inverter circulating current prediction control method of the present invention.

[0067] Figure 6 The diagram shows a comparison of the zero-sequence circulating current simulation waveforms of the parallel inverter circulating current prediction control method of the present invention and the prior art under four parallel inverters with different reference currents and line impedances.

[0068] Figure 7 The simulation waveform comparison of the zero-sequence circulating current of the parallel inverter predictive control method of the present invention and the prior art under the condition of a step change in reference current from 12A to 24A is shown in the figure.

[0069] Figure 8 The simulation spatial distribution diagram of the zero-sequence circulating current suppression effect of the parallel inverter circulating current prediction control method of the present invention in the α-β-ZSCC coordinate system is shown.

[0070] Figure 9 The waveforms of phase a current and zero-sequence circulating current are shown in Method 1 when the reference current is different but the line inductance is the same.

[0071] Figure 10 The waveforms of phase a current and zero-sequence circulating current are shown when the reference current is different but the line inductance is the same, using method two.

[0072] Figure 11 The waveforms of phase a current and zero-sequence circulating current are shown when the reference current is different but the line inductance is the same, using the circulating current prediction control method of the parallel inverter of the present invention.

[0073] Figure 12 The waveforms of phase a current and zero-sequence circulating current are shown in Method 1 when the line inductance is different but the reference current is the same.

[0074] Figure 13 The waveforms of phase a current and zero-sequence circulating current are shown when the reference current is the same but the line inductance is different, using method two.

[0075] Figure 14 The waveforms of phase a current and zero-sequence circulating current are shown when the reference current is the same but the line inductance is different, using the circulating current prediction control method of the parallel inverter of the present invention.

[0076] Figure 15The experimental results of the proposed method for predictive control of circulating current in parallel inverters according to the present invention are as follows: the current reference value steps down from 20A to 10A, and then steps up from 10A to 20A. Detailed Implementation

[0077] The circulating current predictive control method for parallel inverters of the present invention will now be described with reference to schematic diagrams, which illustrate preferred embodiments of the invention. It should be understood that those skilled in the art can modify the invention described herein while still achieving its advantageous effects. Therefore, the following description should be understood as being of general knowledge to those skilled in the art and is not intended to limit the invention.

[0078] This invention proposes a circulating current predictive control method for parallel inverters, such as... Figure 1 As shown, it includes the following steps:

[0079] Step S1: Establish a mathematical model of a single inverter in the parallel inverter system in the α-β stationary coordinate system to obtain the circulating current and common-mode voltage of a single inverter;

[0080] Step S2: Equip the other parallel inverters in the parallel inverter system with a virtual inverter to form and obtain the equivalent zero-sequence circulating current, equivalent common-mode voltage, and zero-sequence circulating current between a single inverter and the virtual inverter.

[0081] Step S3: Establish a current prediction model and derive a prediction model for the desired voltage based on it. Based on the prediction model for the desired voltage, calculate the desired voltage vector that satisfies the current tracking target.

[0082] Step S4: Calculate the duration of the non-zero voltage vector used to synthesize the desired voltage vector;

[0083] Step S5: Establish a circulating current prediction model, and based on the circulating current prediction model, calculate the optimal common-mode voltage that satisfies the condition of zero-sequence circulating current in the next cycle being reduced to zero;

[0084] Step S6: Calculate the zero-voltage vector action time that satisfies the optimal common-mode voltage;

[0085] Step S7: Based on the non-zero voltage vector action time and the zero voltage vector action time, synthesize the output voltage vector and apply it to the next control cycle of the parallel inverter system.

[0086] This invention establishes mathematical models for a single inverter and a virtual inverter, and performs hierarchical predictive control of current and zero-sequence circulating current through current prediction model and circulating current prediction model. The inner layer is used to control the current, and the outer layer is used to suppress zero-sequence circulating current, thereby calculating the optimal output voltage and common-mode voltage at the next moment in real time to meet the requirements of current tracking and zero-sequence circulating current suppression, and achieving complete suppression of zero-sequence circulating current under the premise of ensuring accurate current tracking.

[0087] This invention actively eliminates steady-state zero-sequence circulating current error without relying on complex inter-inverter communication or control parameter design, simplifying the control process and significantly improving the theoretical suppression effect of zero-sequence circulating current. This invention not only effectively solves the problems of traditional zero-sequence circulating current suppression methods failing to eliminate steady-state zero-sequence circulating current error and incomplete zero-sequence circulating current suppression, but also relies solely on the hardware parameters of each inverter and the current and voltage data sampled by current sensors. It eliminates the need for inter-inverter communication to obtain information from other inverters and the need to design control parameters.

[0088] In some embodiments, regarding step S1: establish a mathematical model of a single inverter in the parallel inverter system in the α-β stationary coordinate system to obtain the circulating current and common-mode voltage of the single inverter.

[0089] The parallel inverter system refers to a power electronic system in which multiple inverters operate in parallel to supply power to the load, and its topology is shown in the figure below. Figure 2 As shown, in Figure 2 middle, This represents the DC-side capacitor of the x-th inverter. , ,and This represents the output voltage of the x-th inverter; , ,and This represents the output current of the x-th inverter; Indicates the DC side voltage. L represents the mains voltage. x Let x represent the filter inductance of the x-th inverter. This refers to the switching transistor of phase a of the inverter. , and The symbols represent the three-phase currents of the power grid. In these symbols, x∈{1,2} is the inverter number.

[0090] The α-β stationary coordinate system is a commonly used two-phase stationary coordinate system used to transform three-phase AC quantities into two-phase AC quantities, which facilitates analysis and control.

[0091] The circulating current refers to the non-load current flowing between inverters in a parallel inverter system due to differences in the output voltage or impedance of each inverter. Zero-sequence circulating current is a special form of circulating current, mainly caused by common-mode voltage.

[0092] The common-mode voltage refers to the sum of the three-phase voltages divided by three in a three-phase system, reflecting the common component of the three-phase voltages. Common-mode voltage is the primary excitation source causing zero-sequence circulating current.

[0093] In step S1, according to Kirchhoff's voltage law, the mathematical model of any x-th inverter in a multi-parallel inverter system in the α-β stationary coordinate system can be expressed as:

[0094] (1);

[0095] Among them, v a,x v b,x v c,x Let i be the output voltage of the x-th inverter. a,x i b,x i c,x These are the grid-side currents, e g,a e g,b e g,c These are the grid voltages, L and L, respectively. x Let be the filter inductance of the xth inverter.

[0096] The sampled three-phase grid current i abc,x The three-phase summation yields the circulating current i of the x-th inverter. cc,x for:

[0097] (2);

[0098] The three-phase power grid voltage v abc,x The common-mode voltage v of the x-th inverter is obtained by summing the three phases. cmv,x for:

[0099] (3);

[0100] By establishing a mathematical model of a single inverter in the α-β stationary coordinate system, the circulating current and common-mode voltage of the single inverter were obtained, and the electrical behavior of the single inverter was described by mathematical formulas, laying the foundation for subsequent analysis.

[0101] In some embodiments, regarding step S2: the other parallel inverters in the parallel inverter system are equivalent to a virtual inverter, and the equivalent zero-sequence circulating current, equivalent common-mode voltage, and zero-sequence circulating current between the virtual inverter and the single inverter are obtained.

[0102] The virtual inverter is an abstract concept used to equate all other parallel inverters in a parallel inverter system, except for the target inverter, to a single inverter with specific parameters, thereby simplifying system analysis and control design.

[0103] In step S2, in some embodiments, such as Figure 3 As shown, the parallel inverter system is defined as an n-parallel inverter system. For any inverter x in the n-parallel inverter system, the influence of the other (n-1) parallel inverters on that inverter can be equivalent to a virtual inverter. According to circuit equivalence, the inductance L of the virtual inverter... v for:

[0104] (4);

[0105] According to Kirchhoff's laws, the equivalent zero-sequence circulating current i of the virtual inverter cc,v for:

[0106] (5);

[0107] Substituting formulas (2), (3), (4), and (5) into formula (1), the equivalent common-mode voltage v of the virtual inverter can be calculated. cmv,v for:

[0108] (6);

[0109] Where x refers to the serial number of any inverter in the parallel inverter system, and L j i is the filter inductance for the remaining inverters in the parallel inverter system; cc,x The zero-sequence circulating current of the x-th inverter; i cc,j For the zero-sequence circulating current of any other inverter in the parallel inverter system; v cmv,j This is the common-mode voltage of any other inverter in the parallel inverter system.

[0110] Through the above formulas, this invention provides a specific and precise mathematical definition for modeling the virtual inverter in a parallel inverter system. This precise parametric modeling significantly improves the accuracy of subsequent circulating current prediction models and common-mode voltage prediction models, thereby enabling more effective prediction and suppression of zero-sequence circulating current in the parallel inverter system. Ultimately, this helps ensure the stability and current output quality of the parallel inverter system under various operating conditions, avoiding problems such as control deviations and poor circulating current suppression due to inaccurate models.

[0111] Further, in step S2, the voltage equivalent circuit model of the x-th inverter and the virtual inverter is as follows:

[0112] (7).

[0113] This model allows for a clear understanding and quantification of the dynamic interaction between a single inverter and a virtual inverter, laying the foundation for subsequent current and circulating current analysis.

[0114] Further, in step S2, the zero-sequence circulating current between the x-th inverter and the virtual inverter is:

[0115] (8);

[0116] Among them, v cmv,x L is the common-mode voltage of the xth inverter; x Let be the filter inductance of the xth inverter.

[0117] By explicitly providing the mathematical expression for the zero-sequence circulation, the zero-sequence circulation at the current moment can be accurately calculated, providing a direct quantitative basis for subsequent circulation prediction and suppression.

[0118] The equivalent circuit model clearly describes the electrical interaction between a single inverter and a virtual inverter, providing a foundation for understanding and analyzing the generation mechanism of circulating current. Based on this, the precise expression for zero-sequence circulating current can directly quantify the zero-sequence circulating current between the single inverter and the virtual inverter, thus providing accurate input for subsequent circulating current prediction models. This enables the system to more accurately predict the dynamic changes of circulating current, and further, to more effectively calculate the optimal common-mode voltage and zero-voltage vector action time, achieving precise zero-sequence circulating current control and significantly improving the accuracy and effectiveness of circulating current suppression in parallel inverter systems.

[0119] In some embodiments, regarding step S3: establishing a prediction model of the desired voltage, and calculating the desired voltage vector that satisfies the current tracking target based on the prediction model of the desired voltage.

[0120] In step S3, in some embodiments, a current prediction model is established by discretizing the mathematical model of the single inverter in the α-β stationary coordinate system. In the current prediction model, the predicted current for the (k+1)th control cycle is... for:

[0121] (9);

[0122] Among them, T s L represents the sampling period of the control system. x Let be the filter inductance of the x-th inverter; Let x be the output voltage of the x-th inverter in the α-β stationary coordinate system; Let be the grid voltage in the α-β stationary coordinate system.

[0123] A current prediction model is established by discretizing the mathematical model of a single inverter in the α-β stationary coordinate system. The discretization process typically employs methods such as forward Euler method, backward Euler method, or bilinear transformation to transform the differential equation into a difference equation. In this way, the predicted current in the (k+1)th control cycle can be obtained. This current prediction model can predict the current response at the next moment based on the current system state and control input, providing an accurate basis for subsequent voltage control.

[0124] After obtaining an accurate current prediction model, a prediction model for the desired voltage is established based on the current prediction model. In the voltage prediction model, the reference value of the output voltage vector is determined in the (k+1)th control cycle. for:

[0125] (10);

[0126] in, Here is the reference value for the predicted current of the x-th inverter in the α-β stationary coordinate system; Let be the grid voltage of the x-th inverter in the α-β stationary coordinate system.

[0127] This model enables the controller to proactively calculate the voltage required to achieve the current tracking target, thereby enabling precise and rapid control of the inverter's output current.

[0128] By discretizing the mathematical model of a single inverter in the α-β stationary coordinate system and establishing a current prediction model based on this, the output current of the inverter in the next control cycle can be accurately predicted. Furthermore, a prediction model for the desired voltage is established, enabling the system to deduce the desired output voltage vector that satisfies the current tracking target based on the accurate current prediction results. This method avoids the lag and errors that may exist in traditional control, significantly improving the accuracy of current tracking and dynamic response speed. By directly calculating the voltage reference that enables the current to reach the desired value in the next control cycle, current fluctuations can be effectively suppressed, ensuring the stability and efficiency of the parallel inverter system under various operating conditions, thereby optimizing the current control performance of the entire system.

[0129] In some embodiments, regarding step S4: calculate the duration of the non-zero voltage vector applied to synthesize the desired voltage vector.

[0130] In step S4, in some embodiments, in order to track the reference output voltage of the x-th inverter, The current control target for the xth inverter is:

[0131] (11);

[0132] in, Let x be the current control target for the x-th inverter. Let be the reference output voltage of the x-th inverter; Let be the output voltage of the xth inverter.

[0133] In this embodiment, a clear current control objective is defined, providing precise guidance for the prediction model of the desired voltage. When the prediction model of the desired voltage is calculated based on a clear current control objective, the reference value of the output voltage vector can be determined more accurately. This significantly improves the current tracking accuracy and dynamic response speed of the parallel inverter system, effectively avoiding current tracking errors and system instability caused by unclear control objectives, thereby optimizing the overall operating performance of the parallel inverter system.

[0134] Furthermore, in some embodiments, the non-zero voltage vector v in the α-β coordinate system is based on the volt-second balance theory. m and v n The projection mapping relationship with the output voltage vector is as follows:

[0135] (12);

[0136] Non-zero voltage vector v m and v n The duration of action t m and t n They are respectively:

[0137] (13);

[0138] Among them, t m The non-zero voltage vector v m The duration of action, t n The non-zero voltage vector v n The duration of action, v xα v is the α component of the output voltage vector; xβ The β component of the output voltage vector; v mα For voltage vector v m α component; v nα For voltage vector v n α component; v mβ For voltage vector v m The β component; v nβ For voltage vector v n The β component.

[0139] The volt-second balance theory is a fundamental principle in the control of power electronic converters. Its core idea is that the voltage-time product across the inductor must be zero within a switching cycle to ensure that the inductor current returns to its initial value after a switching cycle, thereby achieving precise control of the average voltage.

[0140] in, Figure 4 This diagram illustrates the voltage vector and its corresponding common-mode voltage in the parallel inverter circulating current predictive control method of the present invention. The non-zero voltage vector duration refers to the duration of action of the non-zero basic voltage vectors (e.g., v1-v6) used to synthesize the desired output voltage vector in space vector modulation (SVM). Correspondingly, the zero voltage vector duration refers to the duration of action of the zero basic voltage vectors (e.g., v0, v7) used to synthesize the desired output voltage vector in space vector modulation (SVM).

[0141] In an inverter, the desired average output voltage vector can be synthesized by reasonably allocating the duration of different voltage vectors (including non-zero voltage vectors and zero voltage vectors).

[0142] In this embodiment, the non-zero voltage vector v m and v n The time of action t m and t n These represent the two adjacent non-zero fundamental voltage vectors v in each control cycle. m and v n The duration of each action is calculated based on the volt-second balance theory and the projection mapping relationship of the desired output voltage vector in the α-β coordinate system. By accurately calculating and applying these durations, the inverter can synthesize an average output voltage that matches the desired voltage vector, thereby achieving precise control of the output current.

[0143] Through the above technical solution, this embodiment can accurately decompose the desired voltage vector into two adjacent non-zero voltage vectors v. m and v n The time of action t m and t n By accurately calculating the duration of these non-zero voltage vectors, the inverter can precisely synthesize the required output voltage vector, thereby ensuring the achievement of the current tracking target and significantly improving the control accuracy and dynamic response performance of the output current of the parallel inverter system.

[0144] In some embodiments, regarding step S5: establishing a circulating current prediction model, and based on the circulating current prediction model, calculating the optimal common-mode voltage that satisfies the condition of reducing the zero-sequence circulating current to zero in the next cycle.

[0145] In step S5, in some embodiments, a circulating current prediction model is established by discretizing the circulating current equation and common mode voltage equation of the x-th inverter in the α-β stationary coordinate system.

[0146] Specifically, for the circulating flow control objective, the circulating flow i predicted in the (k+1)th control cycle is... cc,x (k+1) can be obtained through discretization formula (8), then the specific expression of the circulation prediction model is:

[0147] (14);

[0148] Substituting formulas (5) and (6) into formula (14) and performing algebraic calculations, the specific expression of the circulation prediction model is as follows:

[0149] ;

[0150] in, This represents the circulating current of the x-th inverter. This represents the circulating current of other inverters.

[0151] The circulating current prediction model directly reflects the dynamic characteristics of the circulating current and is affected by inductance and sampling period. Using this model, the controller can accurately predict the circulating current at the next moment within each sampling period, utilizing real-time measurements and known system parameters, providing a precise basis for subsequent circulating current suppression strategies.

[0152] Furthermore, in some embodiments, based on the circulating current prediction model, the optimal common-mode voltage for reducing the zero-sequence circulating current of the x-th inverter and other inverters to zero in the (k+1)th control cycle is:

[0153] (16);

[0154] in, Let be the optimal common-mode voltage at which the zero-sequence circulating current of the x-th inverter returns to zero in the (k+1)th control cycle. The optimal common-mode voltage for other inverters to return to zero in the (k+1)th control cycle is the zero-sequence circulating current.

[0155] The optimal common-mode voltage is a specific value calculated based on the current circulating current magnitude and system parameters. Its function is to actively cancel the voltage difference in the circulating current loop, thereby bringing the zero-sequence circulating current to zero in the next control cycle.

[0156] In some embodiments, regarding step S6: calculate the zero-voltage vector action time that satisfies the optimal common-mode voltage.

[0157] In step S6, in some embodiments, the common-mode voltage tracking target of the x-th inverter can be adjusted from zero common-mode voltage to optimal common-mode voltage, and the circulating current control target of the x-th inverter is:

[0158] (17).

[0159] in, For the circulating current control target of the xth inverter, v cmv,x Let be the common-mode voltage of the x-th inverter in the (k+1)th control cycle.

[0160] The circulating current control objective refers to the ideal state or value that the circulating current is expected to reach in a parallel inverter system. In predictive control strategies, setting a clear circulating current control objective provides the controller with a reference value that needs to be tracked or eliminated.

[0161] Furthermore, in some embodiments, based on the volt-second balance theory, the common-mode voltage of the x-th inverter in the α-β stationary coordinate system is:

[0162] (18);

[0163] in, The non-zero voltage vector v m The common-mode voltage; The non-zero voltage vector v n The common-mode voltage; The common-mode voltage is the zero-voltage vector v0; The common-mode voltage of the zero voltage vector v7;

[0164] The accurate calculation of the common-mode voltage is fundamental to determining the zero-sequence circulating current (ZVDC) action time. It reflects the common-mode component that the inverter needs to generate within the current control cycle to achieve the desired output voltage vector while simultaneously suppressing zero-sequence circulating current. Using this formula, the controller can acquire the inverter's common-mode voltage state in real time, providing necessary feedback information for circulating current predictive control.

[0165] Based on this, the required durations t0 and t7 of the zero voltage vectors v0 and v7 are respectively:

[0166] (19);

[0167] Where t0 is the duration of the zero voltage vector v0, and t7 is the duration of the zero voltage vector v7; v dc This is the DC side voltage.

[0168] In space vector modulation (SVM) technology, zero-voltage vectors v0 and v7 are crucial components of inverter output voltage vector synthesis. In parallel inverter systems, the allocation of the zero-voltage vector duration is critical for common-mode voltage synthesis and zero-sequence circulating current suppression. The formulas for t0 and t7 are derived based on the volt-second balance principle and common-mode voltage control objectives. They precisely quantify the durations t0 and t7 required for zero-voltage vectors v0 and v7 to synthesize an output voltage vector that satisfies the optimal common-mode voltage within each control cycle.

[0169] This embodiment can effectively synthesize the required output voltage vector by precisely adjusting the duration of the zero-sequence circulating current, based on the optimal common-mode voltage calculated by the circulating current prediction model. Specifically, by precisely controlling the durations t0 and t7 of the zero-voltage vectors v0 and v7, the common-mode voltage output of the inverter can be directly affected, thereby achieving precise suppression of the zero-sequence circulating current.

[0170] In some embodiments, regarding step S7: based on the non-zero voltage vector action time and the zero voltage vector action time, an output voltage vector is synthesized and applied to the next control cycle of the parallel inverter system.

[0171] Specifically, based on the calculated non-zero voltage vector action time, zero voltage vector action time, and corresponding basic voltage vector, the corresponding basic voltage vector is applied in the next control cycle according to these action time sequences using space vector modulation (SVM) technology, thereby synthesizing the required output voltage vector.

[0172] In summary, the parallel inverter circulating current predictive control method of the present invention first calculates the non-zero voltage vector v. m and v n The time of action t m and t n To achieve precise current tracking and control, and then, constrained by the inner layer calculation results, calculate the action times t0 and t7 of the zero voltage vectors v0 and v7, synthesize the optimal common-mode voltage, and output the voltage vector acting on the next control cycle of the parallel two-level inverter. The overall idea of ​​this invention can be summarized by the following formula (20):

[0173] (20).

[0174] in, The reference value for the duration of the zero voltage vector v0 is given. This is a reference value for the duration of the zero-voltage vector v7. The non-zero voltage vector v m Reference value for the duration of action. The non-zero voltage vector vn Reference value for the duration of action. This is the reference value for the output voltage vector.

[0175] That is, by minimizing the cost function, the current control objective of formula (11) is optimized to obtain the non-zero voltage vector v. m and v n The time of action t m and t n By minimizing the cost function, for the circulating current control objective of formula (17), the action times t0 and t7 of zero voltage vectors v0 and v7 are obtained, and finally the optimal common-mode voltage is synthesized and the output voltage vector is applied to the next control cycle of the parallel two-level inverter.

[0176] The schematic diagram of the parallel inverter circulating current predictive control method of the present invention is shown below. Figure 5 As shown, specifically, Module I demonstrates the calculation of the reference voltage vector based on the current tracking target in the α-β coordinate system. In this context, VVRef. also represents the reference voltage vector; Module II demonstrates the decomposition of the reference voltage vector into two non-zero voltage vectors v. m and v n By allocating their effective times, the output current can accurately track the reference current. Module III demonstrates the calculation of the optimal common-mode voltage (OCMV) based on the circulating current prediction model, with the goal of achieving zero zero-sequence circulating current in the next cycle. Here, CMV Ref. refers to the reference common-mode voltage. Module IV demonstrates the calculation of the optimal common-mode voltage while maintaining a non-zero voltage vector v. m and v n Action time t m and t n Under the premise of not changing, adjust the action time t0 and t7 of the zero voltage vectors v0 and v7 to synthesize the reference optimal common-mode voltage, so that the zero-sequence circulating current approaches zero in the next cycle. Here, ZCMV refers to the zero common-mode voltage.

[0177] To verify the effectiveness of the parallel inverter circulating current predictive control method of the present invention, it was verified by Matlab / Simulink simulation, and the parameters are shown in Table 1 below.

[0178] Table 1 Parameter Configuration

[0179]

[0180] in, Figure 6The waveforms of the zero-sequence circulating current are shown for four parallel inverters (Inv.1 to Inv.4) operating under different reference currents and line inductances. Methods one and two in the background art can suppress the amplitude of the zero-sequence circulating current to 1.1A. The present invention (method three) further reduces the amplitude of the zero-sequence circulating current to 0.3A while also reducing ripple, fully demonstrating the effectiveness of the present invention.

[0181] Figure 7 The diagram shows the zero-sequence circulating current waveform of a single inverter when the reference current changes from 12A to 24A. The reference current for other inverters is 12A, and the line inductance of each inverter is 4.5mH. It can be seen that each method effectively suppresses the zero-sequence circulating current. However, since the suppression effect of zero-sequence circulating current is mainly referenced to the effective value RMS and peak-to-peak value PP, where RMS represents the effective value and PP represents the peak-to-peak value, i.e., the difference between the highest point (positive peak) and the lowest point (negative peak) of the signal waveform, in… Figure 7 In the comparison, Method 1, lacking zero-sequence circulating current feedback, saw its zero-sequence circulating current rise from 0.8A to 1.6A, with the largest peak-to-peak RMS, thus performing the worst. Method 2, while appearing to have many circulating current spikes and its zero-sequence circulating current amplitude fluctuating between 0.6A and 1.4A, did so because of the active adjustment of the voltage vector, resulting in an overall circulating current closer to zero, with lower RMS and peak-to-peak PP, thus performing better than Method 1. The present invention (Method 3) exhibits the best zero-sequence circulating current suppression capability, strictly limiting the zero-sequence circulating current to below 0.5A.

[0182] Figure 8 This is a simulation spatial distribution diagram of the zero-sequence circulating current suppression effect of the present invention in the α-β-ZSCC coordinate system. On the α-β plane, the data points are dispersed, indicating that the system can output the required grid-connected current normally, with good dynamic performance and is not affected by the circulating current suppression. On the ZSCC axis, the data points converge to zero, indicating that the control accuracy of the circulating current suppression loop is extremely high, and the system operates stably without continuous circulating current oscillations or large pulsations.

[0183] To verify the circulation suppression capability of the present invention Figure 9 , Figure 10 and Figure 11 The graph shows the waveforms of phase a current and zero-sequence circulating current under different methods when the reference current is different but the line inductance is the same. The reference circuits for inverters Inv.1 and Inv.2 are configured with 10A and 20A respectively. In the graph, RMS represents the effective value, and PP represents the peak-to-peak value, which is the difference between the highest point (positive peak value) and the lowest point (negative peak value) of the signal waveform. The phase a current represents the three-phase output current of the first inverter in the experiment. The a-phase current represents the three-phase output current of the second inverter in the experiment. THD represents the harmonic distortion rate, which is usually the ratio of the sum of the power (or effective value of voltage and current) starting from the second harmonic to the power (or effective value) of the fundamental signal.

[0184] in, Figure 9 The waveforms of phase a current and zero-sequence circulating current when using method one are shown. Figure 10 The waveforms of phase a current and zero-sequence circulating current when using method two are shown. Figure 11 The diagram shows the waveforms of phase a current and zero-sequence circulating current when using the circulating current predictive control method for parallel inverters according to the present invention. Although the reference currents are not equal, Method 1, Method 2, and the present invention all effectively suppress the zero-sequence circulating current. However, when Method 1 and Method 2 are executed, the RMS (effective value) of the zero-sequence circulating current is limited to 0.85A and 0.3A, respectively. In contrast, the present invention exhibits excellent zero-sequence circulating current suppression performance, with the RMS of the zero-sequence circulating current limited to within 0.11A.

[0185] Figure 12 , Figure 13 and Figure 14 The waveforms of phase a current and zero-sequence circulating current are shown under different methods when the line inductance is different but the reference current is the same. The line inductances of inverters Inv.1 and Inv.2 are set to 3.5mH and 7mH, respectively. Figure 12 The waveforms of phase a current and zero-sequence circulating current when using method one are shown. Figure 13 The waveforms of phase a current and zero-sequence circulating current when using method two are shown. Figure 14 The waveforms of phase a current and zero-sequence circulating current when using the parallel inverter circulating current prediction control method of the present invention are shown.

[0186] When Method 1 is implemented, the RMS of the zero-sequence circulating current is as high as 0.76A. Method 2 reduces the zero-sequence circulating current amplitude to 0.29A, with a relatively large peak value. In contrast, when this invention is implemented, the zero-sequence circulating current amplitude is significantly reduced and limited to within 0.1A. Therefore, this invention can more effectively suppress zero-sequence circulating current under unequal current and inductance conditions.

[0187] To verify the steady-state performance and dynamic response of the present invention, Figure 15 The dynamic response to current and zero-sequence circulating current is described. The current decreases from 20A to 10A, and then recovers from 10A to 20A. Figure 11 As shown, each inverter rapidly tracks the step-change reference current, with the response time for current reduction and recovery both remaining below 1.2 ms. The zero-sequence circulating current shows no significant change during the transient process, indicating that this invention has good dynamic zero-sequence circulating current suppression capability.

[0188] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A circulating current predictive control method for a parallel inverter, characterized in that, include: A mathematical model of a single inverter in a parallel inverter system in the α-β stationary coordinate system is established to obtain the circulating current and common-mode voltage of a single inverter. The other parallel inverters in the parallel inverter system are equivalent to a virtual inverter, and the equivalent zero-sequence circulating current, equivalent common-mode voltage, and zero-sequence circulating current between a single inverter and the virtual inverter are obtained. A current prediction model is established and a prediction model for the desired voltage is derived from it. Based on the prediction model for the desired voltage, the desired voltage vector that satisfies the current tracking target is calculated. Calculate the duration of action of the non-zero voltage vector used to synthesize the desired voltage vector; Establish a circulating current prediction model, and based on the circulating current prediction model, calculate the optimal common-mode voltage that satisfies the condition of reducing the zero-sequence circulating current to zero in the next cycle; Calculate the zero-voltage vector action time that satisfies the optimal common-mode voltage; Based on the duration of the non-zero voltage vector and the duration of the zero voltage vector, an output voltage vector is synthesized and applied to the next control cycle of the parallel inverter system.

2. The parallel inverter circulating current predictive control method according to claim 1, characterized in that, The parallel inverter system is defined as an n-parallel inverter system, and the filter inductance L of the virtual inverter is... v for: ； The equivalent zero-sequence circulating current i of the virtual inverter cc,v for: ； The equivalent common-mode voltage v of the virtual inverter cmv,v for: ； Where x refers to the serial number of any inverter in the parallel inverter system, and L j i is the filter inductance for the remaining inverters in the parallel inverter system; cc,x The zero-sequence circulating current of the x-th inverter; i cc,j For the zero-sequence circulating current of any other inverter in the parallel inverter system; v cmv,j This is the common-mode voltage of any other inverter in the parallel inverter system.

3. The parallel inverter circulating current predictive control method according to claim 2, characterized in that, The voltage equivalent circuit model of the x-th inverter and the virtual inverter is as follows: ； The zero-sequence circulating current between the x-th inverter and the virtual inverter is: ； Among them, v cmv,x L is the common-mode voltage of the xth inverter; x Let be the filter inductance of the xth inverter.

4. The parallel inverter circulating current predictive control method according to claim 3, characterized in that, By discretizing the mathematical model of the single inverter in the α-β stationary coordinate system, a current prediction model is established. In the current prediction model, the predicted current for the (k+1)th control cycle is... for: ； Among them, T s L represents the sampling period of the control system. x Let be the filter inductance of the x-th inverter; Let x be the output voltage of the x-th inverter in the α-β stationary coordinate system; The grid voltage is in the α-β stationary coordinate system; Based on the current prediction model, a prediction model for the desired voltage is established. In the voltage prediction model, in the (k+1)th control cycle, the reference value of the output voltage vector is... for: ； in, Here is the reference value for the predicted current of the x-th inverter in the α-β stationary coordinate system; Let be the grid voltage of the x-th inverter in the α-β stationary coordinate system.

5. The parallel inverter circulating current predictive control method according to claim 4, characterized in that, The current control target for the xth inverter is: ； in, Let x be the current control target for the x-th inverter. Let be the reference output voltage of the x-th inverter; Let be the output voltage of the xth inverter.

6. The parallel inverter circulating current predictive control method according to claim 5, characterized in that, Based on the volt-second balance theory, the non-zero voltage vector v in the α-β coordinate system m and v n The projection mapping relationship with the output voltage vector is as follows: ； Non-zero voltage vector v m and v n The duration of action t m and t n They are respectively: ； Among them, t m The non-zero voltage vector v m The duration of action, t n The non-zero voltage vector v n Duration of action, v xα v is the α component of the output voltage vector; xβ The β component of the output voltage vector; v mα For voltage vector v m α component; v nα For voltage vector v n α component; v mβ For voltage vector v m The β component; v nβ For voltage vector v n The β component.

7. The parallel inverter circulating current predictive control method according to claim 6, characterized in that, By discretizing the circulating current equation and common-mode voltage equation of the x-th inverter in the α-β stationary coordinate system, a circulating current prediction model is established. The circulating current prediction model is as follows: ; in, This represents the circulating current of the x-th inverter. This represents the circulating current of other inverters.

8. The parallel inverter circulating current predictive control method according to claim 7, characterized in that, Based on the aforementioned circulating current prediction model, the optimal common-mode voltage for zeroing the zero-sequence circulating current of the x-th inverter and other inverters in the (k+1)th control cycle is: ； in, Let be the optimal common-mode voltage at which the zero-sequence circulating current of the x-th inverter returns to zero in the (k+1)th control cycle. The optimal common-mode voltage for other inverters to return to zero in the (k+1)th control cycle is the zero-sequence circulating current.

9. The parallel inverter circulating current predictive control method according to claim 8, characterized in that, The circulating current control objective for the xth inverter is: ； in, For the circulating current control target of the xth inverter, v cmv,x Let be the common-mode voltage of the x-th inverter in the (k+1)th control cycle.

10. The parallel inverter circulating current predictive control method according to claim 9, characterized in that, In the α-β stationary coordinate system, the common-mode voltage of the x-th inverter is: ； in, The non-zero voltage vector v m The common-mode voltage; The non-zero voltage vector v n The common-mode voltage; The common-mode voltage is the zero-voltage vector v0; The common-mode voltage of the zero voltage vector v7; The required durations t0 and t7 for the zero voltage vectors v0 and v7 are respectively: ； Where t0 is the duration of the zero voltage vector v0, and t7 is the duration of the zero voltage vector v7; v dc This is the DC side voltage.

Images