Rectifier adaptive zero-crossing distortion suppression method and related products

Abstract

The present application relates to the technical field of rectifier control, and particularly relates to a rectifier adaptive zero-crossing distortion suppression method and related products, comprising determining sector misjudgment conditions near input current zero-crossing; establishing a vector error prediction model; constructing a multi-objective cost function of the prediction model; rolling optimization is performed on the multi-objective cost function, the switching state that makes the cost function minimum is selected as the optimal switching state, and the optimal switching state is used to control the rectifier; the technical scheme proposed by the present application analyzes and models the zero-crossing distortion mechanism in depth, realizes dynamic compensation of voltage vector error by using model predictive control, and effectively solves the inherent input current zero-crossing distortion problem of the Vienna rectifier.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rectifier control, in particular to a rectifier adaptive zero-crossing distortion suppression method and related products, especially for a three-phase Vienna rectifier. BACKGROUND

[0002] The Vienna rectifier has the advantages of high reliability, high efficiency and high power density, and is widely used in new energy power generation, electric vehicle charging piles, aviation power supply and communication power supply fields. However, the Vienna rectifier has inherent input current zero-crossing distortion problems, which will lead to increased input current harmonics, decreased power quality, and even threaten the stability of the controller.

[0003] In order to suppress the input current zero-crossing distortion of the Vienna rectifier, the existing technology mainly deals with it from the two aspects of optimizing modulation and control strategy, such as forcing the current at zero-crossing to be clamped to zero to achieve zero-crossing distortion suppression, decoupling input current zero-crossing distortion and midpoint potential balance, dividing sectors to optimize the candidate voltage vector set, and suppressing input current zero-crossing distortion, etc. However, the existing technology still has the following problems:

[0004] (1) The existing method divides the voltage vector sectors according to the input current polarity, ignoring the influence of current ripple and sampling error of the control circuit on sector identification, which leads to repeated identification of sectors near the input current zero-crossing, i.e. sector misjudgment, and ultimately causes the input current zero-crossing distortion phenomenon to be aggravated.

[0005] (2) The existing method lacks quantitative modeling of the vector error generated by the shared switch state when the sector misjudgment occurs, resulting in insufficient zero-crossing distortion suppression capability. SUMMARY

[0006] In order to solve the above technical problems, the present application provides a rectifier adaptive zero-crossing distortion suppression method and related products, which can obtain low-harmonic and zero-crossing-free input current under different current ripples and sampling error conditions, as well as under load mutation conditions.

[0007] The present application is realized by the following technical solutions:

[0008] A rectifier adaptive zero-crossing distortion suppression method, comprising the following steps:

[0009] Determine the sector misjudgment condition near the input current zero-crossing, which is determined by current ripple and sampling error;

[0010] Establish a vector error prediction model, which is used to quantify the voltage vector error generated when the sector misjudgment condition is met and the shared switch state is used;

[0011] constructing a multi-objective cost function for the prediction model, the multi-objective cost function comprising at least a current tracking term, a midpoint potential balancing term and a vector error term; the vector error term being calculated by the vector error prediction model;

[0012] rolling optimization is performed on the multi-objective cost function, and a switching state that makes the cost function minimum is selected as an optimal switching state, and the optimal switching state is used to control the rectifier.

[0013] Optionally, the determination method of the sector misjudgment condition comprises:

[0014] setting a maximum sampling error and a maximum current ripple;

[0015] calculating an actual current considering error based on a sampling current at a current time, the maximum sampling error and the maximum current ripple;

[0016] judging whether the sampling current and the actual current satisfy a polarity reversal condition or a sector boundary crossing condition; if yes, it is determined that the sector misjudgment condition is satisfied;

[0017] wherein the polarity reversal condition refers to a product of components of the sampling current and the actual current in the d-axis and the q-axis being less than zero; and the sector boundary crossing condition refers to a product of absolute values of tangents of the sampling current and the actual current in a two-dimensional coordinate system and a current zero-crossing demarcation line being less than zero.

[0018] Optionally, the method for establishing the vector error prediction model comprises:

[0019] identifying a shared switching state existing in a sector misjudgment region, the shared switching state referring to a state corresponding to different voltage vectors but having the same switching combination in a detected sector and an actual sector;

[0020] determining a shared voltage vector pair corresponding to the shared switching state, the shared voltage vector pair comprising a replaced shared voltage vector located in the detected sector and an actual shared voltage vector located in the actual sector;

[0021] when the sector misjudgment condition is satisfied and the shared switching state is used, calculating a modulus value of a difference between the actual shared voltage vector and the replaced shared voltage vector;

[0022] determining a product of the modulus value and an action time of the shared switching state in a current control period as a voltage vector error.

[0023] Optionally, the determination method of the shared voltage vector pair comprises:

[0024] ​​By comparing the relative positions of the detected sector and the actual sector, the misjudgment type of the sector is determined to be either a lag error type or a lead error type; wherein, the lag error type refers to the detected sector lagging behind the actual sector, and the lead error type refers to the detected sector leading the actual sector;

[0025] Based on the sector misjudgment type and the current shared switch state, a corresponding shared voltage vector pair is selected from a preset mapping relationship; the preset mapping relationship includes: the correspondence between the shared switch state and the replaced shared voltage vector located in the detected sector and the actual shared voltage vector located in the actual sector under different sector boundaries.

[0026] Optionally, based on the volt-second balance principle, the duration of the shared switch state in the current control cycle is calculated using the desired synthesized voltage vector and the base voltage vector in the current sector.

[0027] Alternatively, methods for constructing a multi-objective cost function include:

[0028] A combined function comprising an input current tracking term, a midpoint potential balance term, and a vector error term is constructed, and corresponding weighting coefficients are assigned; wherein, the input current tracking term is determined by the absolute value of the difference between the reference current and the predicted current at the next time step;

[0029] The midpoint potential balance term is determined by the absolute value of the difference between the predicted value of the upper capacitor voltage and the predicted value of the lower capacitor voltage on the DC side at the next moment.

[0030] The vector error term is determined by the absolute value of the voltage vector error calculated by the vector error prediction model.

[0031] Optionally, the method for obtaining the predicted current and the predicted voltage values ​​includes:

[0032] Building rectifiers in Current prediction model in coordinate system ;

[0033] Based on the current time grid-side voltage Input current ,resistance ,inductance Control cycle and the optimal voltage vector of the output The next time step is calculated using the current prediction model. Predicted current ;

[0034] Constructing a DC-side capacitor voltage prediction model ;

[0035] based on a voltage value of a capacitor at a current time , a control period , a capacitor value and a current value flowing through the capacitor , a next time capacitor voltage prediction value on the DC side and a capacitor voltage prediction value on the DC side are calculated by using the DC side capacitor voltage prediction model.

[0036] Optionally, the method for obtaining the optimal switching state comprises:

[0037] determining a candidate voltage vector set of the rectifier, the candidate voltage vector set containing all feasible candidate switching states of the rectifier at the current time;

[0038] traversing each candidate switching state in the candidate voltage vector set, and calculating a function value of a multi-objective cost function corresponding to each candidate switching state in combination with the prediction model;

[0039] comparing all calculated function values to identify the minimum function value;

[0040] determining the candidate switching state corresponding to the minimum function value as the optimal switching state, and applying the optimal switching state to the three-phase bidirectional switch tube of the rectifier.

[0041] A computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the rectifier adaptive zero-crossing distortion suppression method as described above.

[0042] A computer program product comprises a computer program / instruction, and the computer program / instruction is executed by a processor to implement the rectifier adaptive zero-crossing distortion suppression method as described above.

[0043] Compared with the prior art, the present application has the following characteristics and beneficial effects:

[0044] The present application first establishes a sector misjudgment condition based on current ripple and sampling error, and then establishes a quantitative prediction model for voltage vector error generated by using a shared switching state in the misjudgment area; on this basis, a multi-objective cost function containing three indexes of current tracking, midpoint potential balance and vector error is constructed, and the optimal switching state is selected through the rolling optimization mechanism of model predictive control to realize the control of the rectifier.

[0045] ​The application can accurately identify the potential misjudgment area near the zero-crossing of the input current by taking the current ripple and sampling error into the judgment logic of the sector misjudgment condition, thereby effectively overcoming the defects of the prior art that only relies on the current polarity for judgment and ignores the interference factors; the use of the shared switch state is optimized by establishing a vector error prediction model to calculate the voltage vector error and then performing rolling optimization through the construction of a multi-objective cost function.

[0046] The technical scheme provided by the application solves the input current zero-crossing distortion problem inherent to the Vienna rectifier by using model predictive control to realize dynamic compensation of the voltage vector error, and additionally, adaptive distortion suppression can be realized through software algorithm optimization without additional hardware circuits, thereby improving the overall reliability and practicality of the device. BRIEF DESCRIPTION OF DRAWINGS

[0047] The accompanying drawings illustrate exemplary embodiments of the present application and together with the description, explain the principles of the application, wherein the drawings are provided to give a further understanding of the application and are incorporated in and constitute a part of this specification, and do not limit the embodiments of the application.

[0048] Figure 1 is a schematic diagram of a three-phase Vienna rectifier topology according to the application.

[0049] Figure 2 is a switch state and voltage vector diagram of the Vienna rectifier according to the application.

[0050] Figure 3 is a schematic diagram of the actual current and the corresponding sector of the sampling current according to the application.

[0051] Figure 4 is a current conduction path diagram corresponding to the shared switch state of sectors II and III according to the application.

[0052] Figure 5 is a vector space distribution diagram near the zero-crossing of the a-phase current between sectors II and III according to the application.

[0053] Figure 6 is a schematic diagram of adaptive zero-crossing distortion suppression of the Vienna rectifier according to the application.

[0054] Figure 7 is a suppression effect diagram of the zero-crossing distortion caused by the current ripple according to the application.

[0055] Figure 8 is a flowchart of a rectifier adaptive zero-crossing distortion suppression method according to the application. DETAILED DESCRIPTION

[0056] In order to make the objectives, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application in combination with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related content, rather than limit the present application.

[0057] Firstly, the key terms are explained. The Vienna rectifier is a high-performance power factor correction (PFC) circuit for three-phase power supply, which has the characteristics of high efficiency, low harmonic, high power density, and low switching tube voltage stress. The THD is the percentage of the sum of the effective values of harmonic components to the effective value of the fundamental component, which is a key indicator to quantify the "purity" of the power waveform. The lower the value, the closer the waveform is to the ideal sine wave, and the better the quality. The higher the value, the more serious the waveform distortion.

[0058] The present application mainly aims to solve the problem that the existing technology for suppressing input current zero-crossing distortion does not analyze the distortion mechanism deeply, ignores the influence of current ripple and sampling error of the control circuit on sector identification, leading to repeated identification of sectors near the input current zero-crossing, i.e. sector misjudgment, and ultimately causing the input current zero-crossing distortion phenomenon to be aggravated. The existing technology lacks quantitative modeling of the vector error generated by the use of shared switch state when sector misjudgment occurs, resulting in insufficient zero-crossing distortion suppression capability. The present application solves the problems of slow dynamic response of traditional control strategies, limitation to static working scenarios, high input current harmonic content, and the need to improve power quality.

[0059] In the field of aviation power supply, by applying passive control combined with zero sequence component injection, modulation wave interval clamping control and other strategies, the input current zero-crossing distortion can be effectively suppressed to ensure stable operation of the system under adverse conditions such as unbalanced grid voltage.

[0060] In new energy power generation systems and battery pack charging applications of electric vehicles, the Vienna rectifier may work in a dual-load mode (e.g., charging a series-connected battery pack). In this case, the voltage and power on the upper and lower DC sides may not be equal, making vector synthesis complex and possibly exacerbating current zero-crossing distortion. By injecting a zero sequence voltage, power decoupling and active control of the midpoint voltage can be achieved, simplifying the calculation and effectively suppressing current zero-crossing distortion when power is uneven.

[0061] In industrial power supply systems such as coal mines, there are a large number of nonlinear loads, leading to power quality problems. Some studies have proposed a control method for implementing static reactive power compensation using Vienna rectifiers in parallel. This method allows a portion of the rectifiers to provide reactive power, while another portion compensates for harmonics, effectively suppressing the input current zero-crossing distortion problem that occurs when the rectifiers compensate for reactive power, and improving the system's ability to govern power quality.

[0062] For various three-phase rectifier devices widely used in industry (such as star-connected controllable rectifiers), the real-time compensation method based on modulation waves shows advantages. This method generates a compensation voltage by calculating the maximum and minimum values ​​of the three-phase modulation waves in real time and superimposing it on the original modulation waves. This method has low computational complexity, high reliability, and, as long as the control system is stable, can effectively suppress or eliminate zero-crossing distortion of the input current by adjusting the modulation waves in real time.

[0063] This invention, by shifting the analytical focus to the aspects of current ripple and sampling error, clarifies the impact of sector misjudgment and the vector error generated by the use of shared switch states during sector misjudgment on the zero-crossing distortion of the Vienna rectified input current. This provides a high-performance control scheme and a corresponding mathematical model. By applying a predictive model of vector error and optimizing the use of shared switch states, the zero-crossing distortion of the input current can be effectively suppressed under the condition of achieving midpoint potential balance, and the harmonic content of the input current can be significantly reduced, thereby improving system reliability.

[0064] The inventors conducted zero-crossing distortion analysis on the Vienna rectifier, and the detailed analysis steps are as follows:

[0065] Vienna rectifier topology as follows Figure 1 As shown in the figure. and These are the grid-side input voltage and current, respectively. For grid-side filter inductance; The equivalent resistance on the grid side; It is a freewheeling diode; and It is a bidirectional switching transistor composed of IGBTs; and These are upper and lower filter capacitors with equal DC-side capacitance, and their output voltages are respectively... and ; The DC-side equivalent load has a terminal voltage of .

[0066] Due to the topology characteristics of the Vienna rectifier, the bridge arm output voltage It depends on the switching state of the bidirectional switch and the polarity of the input current: ,in, for Phase output level; for The switching states of the bidirectional switching transistor satisfy the following relationship: ,in, .

[0067] Define the space voltage vector of the Vienna rectifier as ,in, for Bridge output voltage, for Bridge output voltage, for Bridge output voltage, It is a complex unit.

[0068] Thus, the switching state of the Vienna rectifier can be obtained. and the corresponding voltage vector distribution as follows Figure 2 As shown. Based on the three-phase input current and The polarity of the voltage vector's working space can be divided into six sectors (I to VI), and each sector's voltage vector has a corresponding switching state.

[0069] However, due to current ripple and sampling errors in the control circuit, the input current will exhibit a polarity "reversal" near the zero crossing, meaning the polarity of the detected current is opposite to that of the actual current, resulting in inconsistency between the detected sector and the actual sector, i.e., sector misjudgment.

[0070] Near the zero-crossing of phase a current between sectors II and III, the sector misjudgment caused by current ripple and sampling error is as follows: Figure 3 As shown in the diagram, the red shaded area represents the sector misjudgment area caused by current ripple. When the detected sector is sector II, the current polarity is positive, but due to the influence of current ripple, the actual current polarity is negative, and the actual sector is sector III, leading to sector misjudgment. When the detected sector is sector III, the current polarity is negative, but the actual current polarity is positive, and the actual sector is sector II, also causing sector misjudgment. Similarly, the green shaded area represents the sector misjudgment area caused by sampling error leading to current polarity reversal.

[0071] When a sector is misjudged near the zero-crossing current of a certain phase, and the power switching transistor of that phase is off, the output voltage of that phase will change abruptly due to the reversal of current polarity. This will result in two voltage vectors, one in the detection sector and the other in the actual sector, corresponding to the same state. This paper defines a switch state that meets the above conditions as a shared switch state, and defines the two voltage vectors corresponding to this switch state as a shared voltage vector pair. Specifically, the method for determining the shared voltage vector pair includes: selecting the corresponding shared voltage vector pair from a preset mapping relationship based on the sector misjudgment type and the current shared switch state; the preset mapping relationship includes: the correspondence between the shared switch state and the replaced shared voltage vector located in the detected sector and the actual shared voltage vector located in the actual sector under different sector boundaries.

[0072] In the sector misjudgment region near the zero-crossing of phase-a current between sector II and sector III, there are two sharing switch states, which are

[001] and

[010] respectively. When the sharing switch state is

[001] , the phase-a switch state is 0 (off). Because of the current polarity reversal and the bridge arm voltage The output characteristic, voltage vector [1, 1, 0] will be replaced by

[001] as Figure 4 shown in Figs. 4(a) and 4(b). In which, voltage vector [1, 1, 0] is located in the detected sector (sector II), while voltage vector [-1, 1, 0] is located in the actual sector (sector III). Therefore, [1, 1, 0] and [-1, 1, 0] are the sharing voltage vector pair of sharing switch state

[001] . When the sharing switch state is

[001] , the sharing voltage vector pair is [1, 0, -1] and [-1, 0, -1], and the corresponding current conduction diagram is shown in Figs. 4(c) and 4(d). Figure 4

[0073] When the sector misjudgment and the sharing switch state are used, the replacement phenomenon will occur between the corresponding sharing voltage vector pair, which leads to the reference voltage vector cannot be accurately synthesized, and finally causes the input current zero-crossing distortion. In the sector misjudgment region near the zero-crossing of phase-a current between sector II and sector III, the voltage and current vector space distribution is shown in Fig. 3 when the sector misjudgment occurs in a control period during the SVPWM modulation process. The detected sector of this control period is sector II, and the expected synthesized reference voltage vector Figure 5 (red) is located in the triangular region composed of voltage vectors [1, 1, 0], [0, 0, -1], [0, 1, 0] and [0, 1, 1]. The switch states corresponding to these voltage vectors are

[001] ,

[110] ,

[101] and

[100] respectively, in which

[001] is the sharing switch state, and the corresponding sharing voltage vector pair is [1, 1, 0] and [-1, 1, 0]. Due to the sector misjudgment and the sharing switch state being used, the sharing voltage vector [1, 1, 0] will be replaced by [-1, 1, 0], which will lead to the inconsistency between the expected synthesized vector (red) and the actual synthesized vector (blue), and the vector error occurs, as shown in Fig. 4(e). Because the reference voltage vector cannot be accurately synthesized, the a-phase current will be distorted near the zero-crossing, and the analysis method of other sector current zero-crossing distortion is the same as above. Therefore, based on the above analysis, it can be concluded that the input current zero-crossing distortion is caused by the vector error due to the sector misjudgment when the sharing switch state is used. Figure 4

[0074] Based on the above analysis, the causes of Vienna rectifier input current zero-crossing distortion can be summarized as follows:

[0075] (1) When the Vienna rectifier identifies the control sector by the input current polarity, due to the current ripple and sampling error, the input current polarity near the zero-crossing will be reversed, which leads to the sector misjudgment.​​​

[0076] (2) When the sector misjudgment occurs and the sharing switch state is used, the replacement of the sharing voltage vectors occurs between the voltage vectors, resulting in the error of the reference voltage vector synthesis, the vector error, and the zero-crossing distortion of the input current. Therefore, the zero-crossing distortion of the input current can be inhibited by avoiding the vector error.

[0077] Embodiment one

[0078] The embodiment provides a rectifier adaptive zero-crossing distortion inhibition method, which is mainly applied to a three-phase Vienna rectifier. In view of the problem that the sector misjudgment occurs near the zero-crossing of the input current of the Vienna rectifier due to the current ripple and the sampling error, and then the input current distortion is caused, the embodiment realizes the adaptive inhibition of the distortion by establishing a vector error model and incorporating the vector error model into the multi-objective collaborative optimization of the model prediction control.

[0079] The method in the embodiment comprises the following steps:

[0080] S1, determining a sector misjudgment condition near the zero-crossing of the input current, the sector misjudgment condition being determined by the current ripple and the sampling error.

[0081] When the input current is near the zero-crossing point, various interference factors can cause the detected current polarity to be opposite to the actual current polarity (i.e., the polarity is reversed), so that the "detected sector" determined by the control system is inconsistent with the "actual sector" corresponding to the actual physical state, that is, the sector misjudgment occurs. By analyzing the influence of the current ripple and the sampling error on the polarity determination, a mathematical inequality or a logical judgment condition is constructed as the sector misjudgment condition, and the region in which the misjudgment is likely to occur is identified.

[0082] S2, establishing a vector error prediction model, the vector error prediction model being used for quantifying the voltage vector error generated when the sector misjudgment condition is met and the sharing switch state is used.

[0083] When the system is in the sector misjudgment region determined in step S1 and the sharing switch state is output, since the detected sector is different from the actual sector, the actual sharing voltage vector generated by the switch state in the actual circuit is inconsistent with the voltage vector (the replaced sharing voltage vector) expected to be generated by the controller. According to the current sector misjudgment state and the switch state, the specific vector error value is calculated through the established vector error prediction model.

[0084] S3, constructing a multi-objective cost function of the prediction model, the multi-objective cost function at least comprising a current tracking term, a midpoint potential balance term and a vector error term; the vector error term is obtained by the vector error prediction model.

[0085] The current tracking term ensures that the input current closely tracks the reference current, guaranteeing the power factor correction effect. The midpoint potential balancing term controls the balance of the voltages of the upper and lower capacitors on the DC side, preventing midpoint potential drift.

[0086] S4. Perform rolling optimization on the multi-objective cost function, select the switching state that minimizes the cost function as the optimal switching state, and use the optimal switching state to control the rectifier.

[0087] Within each control cycle, all candidate switching states are traversed, and each candidate switching state is substituted into the multi-objective cost function constructed in step S3 for calculation. The obtained function values ​​are compared. The switching state that minimizes the cost function value is selected as the optimal switching state at the current moment, and it is applied to the bidirectional switching transistor of the rectifier, thereby completing the control of the rectifier.

[0088] Example 2

[0089] Based on Example 1, this embodiment, combined with the specific topology of the three-phase Vienna rectifier, provides specific mathematical modeling and detailed methodological steps for each step.

[0090] In step S1, the method for determining the sector misjudgment condition includes: setting the maximum sampling error and the maximum current ripple; calculating the actual current after considering the error based on the current sampling current, the maximum sampling error, and the maximum current ripple at the current moment; determining whether the sampling current and the actual current satisfy the polarity reversal condition or the sector boundary crossing condition; if so, determining that the sector misjudgment condition is satisfied; wherein, the polarity reversal condition refers to the sampling current and the actual current being in polarity reversal condition. The product of the components on the axis is less than zero; the sector boundary crossing condition refers to the difference between the sampled current and the actual current. The product of the difference between the absolute value of the tangent in the coordinate system and the absolute value of the tangent of the current zero-crossing boundary is less than zero.

[0091] The specific mathematical modeling and detailed methodological steps are explained below:

[0092] Based on the above analysis, sector misjudgment is caused by the polarity reversal of the input current near its zero-crossing point due to the influence of current ripple and sampling error. Figure 2 It can be seen that in the sector misjudgment region near the zero-crossing of phase b and c currents, the product of the absolute values ​​of the tangents of the sampled current and the actual current in the coordinate system and the absolute values ​​of the tangents of the boundary lines of the zero-crossing of phase b and c currents is less than "0". In the sector misjudgment region near the zero-crossing of phase a current, the product of the components of the sampled current and the actual current on the axis is less than "0".

[0093] Based on this, the conditions for sector misjudgment caused by sampling error near the zero-crossing of the input current can be expressed as follows: The conditions for sector misjudgment caused by current ripple are expressed as follows: In the formula, and For the maximum sampling error Quantity; and For the maximum current ripple Quantity; For the first Constant sampling current Quantity; For the first The actual current value should always take into account the sampling error. For the first Always consider the actual current value of current ripple.

[0094] Combining the above formulas, the boundary conditions for sector misjudgment can be expressed as follows: In the formula, and for Considering both current ripple and sampling error in the coordinate system The actual current at a given time can be used to calculate the conditions under which sector misjudgment occurs near the zero crossing of the input current due to current ripple and sampling error.

[0095] In step S2, the method for establishing the vector error prediction model includes: identifying the shared switch state existing in the misjudged sector area. The shared switch state refers to the state in which different voltage vectors are corresponding in the detected sector and the actual sector, but the switch combination is the same.

[0096] Determine the shared voltage vector pair corresponding to the shared switch state. The shared voltage vector pair includes: the replaced shared voltage vector located in the detection sector and the actual shared voltage vector located in the actual sector.

[0097] When the sector misjudgment condition is met and the shared switch state is used, calculate the magnitude of the difference between the actual shared voltage vector and the replaced shared voltage vector.

[0098] The voltage vector error is determined by multiplying the magnitude of the difference with the duration of the shared switch state in the current control cycle. The duration of the shared switch state in the current control cycle is calculated based on the volt-second balance principle and using the expected synthesized voltage vector and the base voltage vector in the current sector.

[0099] The specific mathematical modeling and detailed methodological steps are explained below:

[0100] The relative positions of the detected sector and the actual sector are compared to determine whether the sector misjudgment type is a lag error type or a lead error type; the lag error type refers to the detected sector lagging behind the actual sector, and the lead error type refers to the detected sector leading the actual sector.

[0101] When a lag error occurs near the zero-crossing of the a-phase current between sectors II and III, the detected sector is sector II, and the actual sector is sector III. For the desired synthesized voltage vector, located in the triangular sub-sector enclosed by voltage vectors [0, 1, 1], [0, 1, 0] and [1, 1, 0] and [0, 0-1], the corresponding switch states are

[100] ,

[101] ,

[001] and

[110] , respectively. Among them,

[001] is a shared switch state, and the corresponding shared voltage vector is [1, 1, 0] at this time. According to the vector synthesis rule, the action time of each switch state can be calculated , wherein is the action time of the switch state

[100] ; is the action time of the switch state

[101] ; is the common action time of the switch states

[001] and

[110] ; and is the voltage vector in coordinate axis; is the control period.

[0102] The action time of the switch state is , if the DC midpoint potential is balanced, the voltage vectors [1, 1, 0] and [0, 1, -1] have equal action time , and the desired synthesized voltage vector is obtained by substitution .

[0103] However, due to sector misjudgment, the actual sector is sector III, and the shared voltage vector corresponding to the actual shared switch state

[100] is [-1, 1, 0]. Therefore, the actual synthesized voltage vector is obtained .

[0104] Combining the desired synthesized voltage vector and the actual synthesized voltage vector, the voltage vector error is obtained .

[0105] Similarly, when a lead error occurs near the zero-crossing of the a-phase current between sectors II and III, the detected sector is sector III, and the actual sector is sector II.

[010] is a shared switch state, and the corresponding shared error voltage vector pair is [1, 0, -1] and [-1, 0, -1]. At this time, the generated vector error is , is the action time of the shared switch state

[010] .

[0106] It can be found that the vector error is actually the product of the modulus of the difference of the shared voltage vector pair and the action time of the corresponding shared switching state. However, the vector deviation occurs under the condition that the sector misjudgment and the shared switching state are used. Therefore, the calculation formula of the vector error of each sector due to the use of the shared voltage vector pair can be obtained: wherein, in the formula: and are the activation functions of sector misjudgment and shared voltage vector used respectively, and satisfy:

[0107] , .

[0108] is the voltage vector located in the actual sector under the shared switching state, is the voltage vector located in the detected sector under the shared switching state, that is, the shared voltage vector to be replaced.

[0109] Based on the above analysis, the shared switching state of each sector and the corresponding shared voltage vector pair are summarized as shown in Table 1. By calculating the action time of the shared switching state of each sector, the corresponding vector error can be calculated.

[0110] Table 1 Correspondence table of shared switching state of each sector and shared error voltage vector and vector error

[0111]

[0112] In step S3, the method for constructing the multi-objective cost function includes:

[0113] A combined function including an input current tracking term, a midpoint potential balance term and a vector error term is constructed, and corresponding weight coefficients are assigned; wherein the input current tracking term is determined by the absolute value of the difference between the reference current and the predicted current at the next moment; the midpoint potential balance term is determined by the absolute value of the difference between the predicted value of the capacitor voltage on the direct current side and the predicted value of the lower capacitor voltage at the next moment; and the vector error term is determined by the absolute value of the voltage vector error calculated by the vector error prediction model.

[0114] The specific mathematical modeling and detailed method steps are as follows:

[0115] It can be seen from Figure 6 that the control objectives of the Vienna rectifier include input current tracking, direct current side midpoint potential control and input current zero-crossing distortion suppression.

[0116] According to Figure 1 , the current prediction model of the Vienna rectifier based on coordinate system can be obtained , based on the current moment grid-side voltage Input current ,resistance ,inductance Control cycle and the optimal voltage vector of the output The next moment is calculated using a current prediction model. Predicted current .

[0117] Right now: In the formula: for The first coordinate system Input the predicted current value at any time; For the first Current is constantly being input; For the first The voltage on the grid side at any given time; For the first The optimal voltage vector output at time 1 is Coordinate axis components.

[0118] According to the Lagrange interpolation theorem, the input current can be obtained at... Coordinates Reference value of time for The cost function for obtaining current tracking is .

[0119] Constructing a DC-side capacitor voltage prediction model Based on the current moment The voltage value of the capacitor Control cycle capacitance value and the current flowing through the capacitor The predicted values ​​of the upper and lower capacitor voltages on the DC side at the next moment are obtained by using the DC side capacitor voltage prediction model.

[0120] That is, according to Figure 1 ,get Prediction model of DC-side capacitor voltage in coordinate system In the formula: and The first time and The voltage value; and The first time and The voltage value; and respectively the current value flowing through the rectifier at the moment and .

[0121] Therefore, the cost function of the DC side midpoint potential balance control is , and the cost function of the vector error is .

[0122] Therefore, the multi-objective cost function containing the vector error is . and are weight coefficients of input current tracking, midpoint potential balance control and voltage vector error, respectively.

[0123] In step S4, the method for obtaining the optimal switching state comprises:

[0124] determining a candidate voltage vector set of the rectifier, the candidate voltage vector set containing all feasible candidate switching states of the rectifier at the current moment; according to the topological structure characteristics of the Vienna rectifier, determining the candidate voltage vector set available at the current moment. The Vienna rectifier has multiple switching states, corresponding to different space voltage vectors (as shown in FIG. 1). The candidate voltage vector set refers to the set of all feasible candidate switching states of the rectifier at the current moment. Figure 2

[0125] traversing each candidate switching state in the candidate voltage vector set, combining the prediction model (the current prediction model and the capacitor voltage prediction model) to predict the predicted current and the predicted voltage that the system will produce if the switching state is applied at the next moment, and judging whether the switching state will trigger the vector error. Subsequently, all the predicted values (current deviation, voltage imbalance degree, vector error modulus) are substituted into the multi-objective cost function to calculate the function value of the multi-objective cost function corresponding to each candidate switching state;

[0126] comparing all the calculated function values to identify the minimum function value; if a certain candidate switching state (i.e., the shared switching state) will produce a larger vector deviation under the current sector misjudgment condition, the total cost value calculated by the candidate switching state will significantly increase. Therefore, by finding the minimum value, the states that cause distortion are avoided, or the state with the smallest distortion impact is selected.

[0127] The candidate switching state corresponding to the minimum function value is determined as the optimal switching state, and the optimal switching state is applied to the three-phase bidirectional switch tube of the rectifier, so as to improve the use of the shared switching state to dynamically compensate the vector error, thereby realizing the input current zero-crossing distortion suppression.

[0128] Embodiment three

[0129] ​The embodiment provides specific experimental effects after zero-crossing distortion suppression by using the suppression method of embodiment one and embodiment two.

[0130] In this paper, Simulink simulation and physical experiment are combined to verify the effectiveness of the proposed method. The parameter values involved in the experiment are shown in Table 2.

[0131] Table 2 Vienna rectifier parameters

[0132]

[0133] To verify the correctness and feasibility of the constructed mechanism model, the suppression of current ripple and its input zero-crossing distortion caused by sector misjudgment due to current ripple and sampling error by PI control method and the method proposed in this paper are analyzed.

[0134] When the PI control method is used, there is sector misjudgment near the current zero-crossing, and the input current has obvious zero-crossing distortion (0.9ms), and the current THD is 5.98%. In contrast, when the method proposed in this paper is used, the input current smoothly crosses zero without zero-crossing distortion. In addition, the midpoint potential and DC output voltage under the two control methods remain balanced, and the peak value of the bridge arm interphase voltage stabilizes at the preset DC value of 600V.

[0135] When the PI control method is used, and the current ripple and sampling error exist at the same time, the sector misjudgment near the current zero-crossing is more obvious, which leads to the further aggravation of the input current zero-crossing distortion (THD=7.89%), and the distortion duration rises to 1.3ms. In contrast, when the method proposed in this paper is used, even in the case of aggravated sector misjudgment, the input current can still smoothly cross zero without distortion. Similarly, the bridge arm interphase voltage , the midpoint potential and the DC output voltage under the two control methods have no obvious changes. Therefore, it is proved that the zero-crossing distortion mechanism model constructed in this paper is correct, and the proposed method can effectively suppress the input current zero-crossing distortion under the midpoint potential balance.

[0136] To verify the superiority of the proposed control strategy in zero-crossing distortion suppression, based on the experimental parameters shown in Table 2, this paper carries out comparative analysis experiments of four control methods (M1, M2, M3, M4) under different current ripple and sampling error conditions, and focuses on analyzing the zero-crossing distortion suppression effect under different methods from the perspective of harmonic content, such as Figure 7As shown. The actual current ripple is simulated by changing the load value, and the sampling error is simulated by injecting different noises into the control circuit. The specific comparison methods are as follows: Method M1: using the traditional PI control method; Method M2: using the traditional model predictive control method; Method M3: the traditional model predictive control method without using redundant small vectors; Method M4: the method proposed in this paper.

[0137] Depend on Figure 7 It can be seen that method M1 has the worst suppression effect on input current zero-crossing distortion caused by sector misjudgment, with a maximum THD of 13.09%. Method M2 improves the suppression capability of input current zero-crossing distortion, reducing the maximum THD of input current to 11.00%. Method M3, by not using redundant small vectors, significantly improves the suppression capability of input current zero-crossing distortion, further reducing the maximum THD of input current to 6.58%. Compared with the above methods, the proposed method (M4) has the most significant suppression of input current zero-crossing distortion, and even under extreme conditions with large current ripple and sampling errors, the THD of input current remains at a low level. Based on the above experimental analysis, the proposed method can effectively suppress input current zero-crossing distortion by accurately modeling the voltage vector error caused by sector misjudgment and optimizing the use of shared switching states.

[0138] To further verify the effectiveness of the proposed method, a 3kW Vienna rectifier hardware experimental platform was built based on the parameters shown in Table 2. The control circuit of the experimental platform uses TMS320F28335PGFA as the central control chip, the grid-side power supply is simulated by Chroma61705, the load is simulated by Chroma63203A electronic load, the waveform is displayed using a TekMDO34 oscilloscope, and the input current harmonic analysis (THD) is measured using a power quality analyzer.

[0139] When sector misjudgment is caused by current ripple, PI control results in significant distortion of the input current near zero crossing, with a THD of 6.00% and a distortion time of 1.01 ms. In contrast, the proposed method in this paper allows the input current to smoothly cross zero without distortion, reducing the THD to 2.97%. The bridge arm phase-to-phase voltage under both control methods is also compared. The DC midpoint potential and output voltage remain in balance. Therefore, it is proven that the method proposed in this paper can significantly suppress zero-crossing distortion of the input current while achieving the DC midpoint potential.

[0140] When sector misjudgment is caused by current ripple and sampling error, when PI control is adopted, and current ripple and sampling error exist at the same time, input current zero-crossing distortion is further intensified, distortion time increases to 1.42 ms, and THD increases to 7.95%. When the method proposed in the present application is adopted, input current can still smoothly pass through zero point, without distortion, and THD does not change significantly. Similarly, under the two control methods, the bridge arm inter-phase voltage, DC midpoint potential and output remain balanced. Therefore, it is further proved that the method proposed in the present application can effectively suppress input current zero-crossing distortion.

[0141] When the load is suddenly changed, when PI control is adopted, and the load is suddenly changed from 120Ω to 150Ω, input current zero-crossing distortion phenomenon becomes more obvious, distortion time increases from 1.40 ms to 2.59 ms, and THD rises to 10.06%. There is an obvious transient process (20 ms) in the DC output voltage, and the ripple amplitude is as high as 17V. Under the method proposed in the present application, input current has no zero-crossing distortion, the transient process of the DC output voltage is reduced to 10 ms, and the ripple amplitude is also significantly reduced. In addition, under the two control methods, the DC midpoint potential and the bridge arm inter-phase voltage do not change significantly before and after the load is suddenly changed. Based on the above experimental analysis, the method proposed in the present application can effectively suppress input current zero-crossing distortion under different load conditions, and has strong dynamic performance.

[0142] The execution time of PI control algorithm is 24.12µs, while the execution time of the method proposed in the present application is reduced to 14.25µs. In combination with the suppression of input current zero-crossing distortion and the execution efficiency of the algorithm, the method proposed in the present application has better control performance.

[0143] Embodiment Four

[0144] A computer readable storage medium, the computer readable storage medium storing a computer program, the computer program being executed by a processor to implement the above rectifier adaptive zero-crossing distortion suppression method.

[0145] Without loss of generality, the computer readable medium can include computer storage media and communication media. Computer storage media include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer readable instructions data structures, program modules or other data. Computer storage media include RAM, ROM, EPROM, EEPROM, flash memory or other solid state storage technology, CD-ROM, DVD or other optical storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage device. Of course, those skilled in the art can know that computer storage media are not limited to the above. The system memory and mass storage device mentioned above can be collectively referred to as memory.

[0146] A computer program product comprising computer programs / instructions which, when executed by a processor, implement the above rectifier adaptive zero-crossing distortion suppression method.

[0147] A computer program product comprises a computer program or set of instructions designed to perform a specific task or implement a specific function. These programs or instructions are designed to be executable by a processor, thereby implementing a series of predefined steps or operations. The program product can be stored in various forms of computer storage media, such as memory, hard disk, solid state drive, optical disc, or other forms of digital storage devices. It can exist in the form of compiled binary code or in the form of scripts or bytecodes executable by an interpreter. The program product, through carefully designed algorithms and logical instructions, enables the processor to process data in a specific order and manner, completing various functions such as data analysis, user interaction, device control, etc.

[0148] In the description of the present specification, the description of the terms "one embodiment / way", "some embodiments / ways", "example", "specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment / way or example are included in at least one embodiment / way or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment / way or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in any one or more embodiments / ways or examples. In addition, the skilled person in the art can combine and combine the different embodiments / ways or examples described in the present specification and the features of the different embodiments / ways or examples, without contradiction.

[0149] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "a plurality of" is at least two, for example, two, three, etc., unless otherwise specifically limited.

[0150] Those skilled in the art should understand that the above embodiments are only for the purpose of clearly illustrating the present application, and are not intended to limit the scope of the present application. Other changes or modifications can be made to the above invention by those skilled in the art, and these changes or modifications are still within the scope of the present application.

Claims

1. A method for adaptive zero-crossing distortion suppression in a rectifier, characterized in that, Includes the following steps: Determine the sector misjudgment conditions near the zero crossing of the input current, wherein the sector misjudgment conditions are determined by current ripple and sampling error; A vector error prediction model is established, which is used to quantify the voltage vector error generated when the sector misjudgment condition is met and the shared switch state is used. The shared switch state refers to the state in which the detection sector and the actual sector correspond to different voltage vectors but the switch combination is the same. A multi-objective cost function is constructed for the prediction model, wherein the multi-objective cost function includes at least a current tracking term, a midpoint potential balance term, and a vector error term; The vector error term is calculated using the vector error prediction model. The multi-objective cost function is subjected to rolling optimization, and the switching state that minimizes the cost function is selected as the optimal switching state. The rectifier is then controlled using the optimal switching state. The method for determining the sector misjudgment conditions includes: Set the maximum sampling error and maximum current ripple; Based on the current sampling current, the maximum sampling error, and the maximum current ripple, calculate the actual current after taking the error into account; Determine whether the sampled current and the actual current satisfy the polarity reversal condition or the sector boundary crossing condition; if so, determine that the sector misjudgment condition is satisfied. Wherein, the polarity reversal condition refers to the fact that the sampled current and the actual current are in the same polarity. The product of the components on the axis is less than zero; the sector boundary crossing condition refers to the sampling current and the actual current being in... The product of the difference between the absolute value of the tangent in the coordinate system and the absolute value of the tangent of the current zero-crossing boundary is less than zero.

2. The rectifier adaptive zero-crossing distortion suppression method according to claim 1, characterized in that, The method for establishing the vector error prediction model includes: Identify the status of shared switches within the misjudged sector area; Determine a shared voltage vector pair corresponding to the shared switch state, the shared voltage vector pair including: a replaced shared voltage vector located in the detection sector and an actual shared voltage vector located in the actual sector; When the sector misjudgment condition is met and the shared switch state is used, calculate the magnitude of the difference between the actual shared voltage vector and the replaced shared voltage vector. The product of the magnitude of the difference and the duration of the shared switch state within the current control cycle is determined as the voltage vector error.

3. The rectifier adaptive zero-crossing distortion suppression method according to claim 2, characterized in that, The method for determining the shared voltage vector pair includes: By comparing the relative positions of the detected sector and the actual sector, the misjudgment type of the sector is determined to be either a lag error type or a lead error type; wherein, the lag error type refers to the detected sector lagging behind the actual sector, and the lead error type refers to the detected sector leading the actual sector; Based on the sector misjudgment type and the current shared switch state, a corresponding shared voltage vector pair is selected from a preset mapping relationship; the preset mapping relationship includes: the correspondence between the shared switch state and the replaced shared voltage vector located in the detected sector and the actual shared voltage vector located in the actual sector under different sector boundaries.

4. The rectifier adaptive zero-crossing distortion suppression method according to claim 2, characterized in that, Based on the volt-second balance principle, the duration of the shared switch state in the current control cycle is calculated using the expected synthesized voltage vector and the base voltage vector in the current sector.

5. The rectifier adaptive zero-crossing distortion suppression method according to claim 1, characterized in that, Methods for constructing multi-objective cost functions include: A combined function comprising an input current tracking term, a midpoint potential balance term, and a vector error term is constructed, and corresponding weighting coefficients are assigned; wherein, the input current tracking term is determined by the absolute value of the difference between the reference current and the predicted current at the next time step; The midpoint potential balance term is determined by the absolute value of the difference between the predicted value of the upper capacitor voltage and the predicted value of the lower capacitor voltage on the DC side at the next moment. The vector error term is determined by the absolute value of the voltage vector error calculated by the vector error prediction model.

6. The rectifier adaptive zero-crossing distortion suppression method according to claim 5, characterized in that, The method for obtaining the predicted current and the predicted voltage values ​​includes: Building rectifiers in Current prediction model in coordinate system ; Based on the current time grid-side voltage Input current ,resistance ,inductance Control cycle and the optimal voltage vector of the output The next time step is calculated using the current prediction model. Predicted current ; Constructing a DC-side capacitor voltage prediction model ; Based on the current time The voltage value of the capacitor Control cycle capacitance value and the current value flowing through the capacitor The predicted values ​​of the upper and lower capacitor voltages on the DC side at the next moment are calculated using the DC-side capacitor voltage prediction model.

7. The rectifier adaptive zero-crossing distortion suppression method according to claim 1, characterized in that, The method for obtaining the optimal switching state includes: Determine a candidate voltage vector set for the rectifier, the candidate voltage vector set containing all feasible candidate switching states of the rectifier at the current moment; Traverse each candidate switch state in the candidate voltage vector set, and calculate the function value of the multi-objective cost function corresponding to each candidate switch state in combination with the prediction model; Compare all calculated function values ​​and identify the minimum function value; The candidate switch state corresponding to the minimum function value is determined as the optimal switch state, and the optimal switch state is applied to the three-phase bidirectional switch tubes of the rectifier.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the rectifier adaptive zero-crossing distortion suppression method as described in any one of claims 1-7.

9. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the rectifier adaptive zero-crossing distortion suppression method as described in any one of claims 1-7.

Images