Control strategy and modulation method of three-phase four-wire three-level inverter

Abstract

The invention provides a model prediction control strategy and a modulation method for a three-phase four-wire three-level inverter, and the control strategy comprises the steps: building a mathematical model according to the working principle of the inverter, carrying out the discretization of the model through employing a forward Euler method, calculating the predicted values of the output voltage at the next two moments, and carrying out the prediction of the output voltage at the next two moments; an inverter side bridge arm voltage reference value at the current moment is obtained by combining an output voltage cost function, the inverter side bridge arm voltage reference value is input into a modulation module after passing through a wave trap, a driving signal for driving a switch tube is obtained through a modulation method of the modulation module, and a model filtering inductance value is obtained through a real-time filtering inductance parameter identification method with low-pass filtering; according to the modulation method, dual-modulation wave carrier modulation or sine pulse width modulation is selected by judging the fluctuation amplitude of the neutral-point voltage on the direct-current side of the inverter and the balance capacity of the neutral-point voltage of each phase and comprehensively considering the neutral-point voltage balance and the system efficiency on the direct-current side in combination with a fuzzy control method. According to the invention, the multi-objective optimization control of the output voltage quality and the system efficiency when the inverter is provided with the balanced / unbalanced load can be realized.

Description

Technical Field

[0001] The invention relates to the technical field of power electronic converters, and in particular to a control strategy and a modulation method for a three-phase four-wire three-level inverter. Background Art

[0002] Compared with two-level inverters, three-level inverters have the advantages of low output voltage harmonic content, low device switching stress, and low power loss, and are widely used in industry. When the system is running, it is inevitable to encounter unbalanced load conditions. The three-phase four-wire three-level inverter connects the midpoint of the DC side capacitor and the neutral point of the inverter AC side load to provide a flow path for the zero-sequence current generated when the inverter output terminal is connected to an unbalanced load, so as to ensure the output voltage quality. In the control scheme, model predictive control has the advantages of fast dynamic response, simple structure, and fewer adjustment parameters compared with traditional linear control methods. However, model predictive control is highly dependent on parameters. The filter inductance value in the circuit will change due to the influence of current, temperature, etc. in actual operation, which will affect the output voltage quality. Therefore, real-time inductance identification is particularly important in the output voltage optimization control method based on model predictive control. Continuous set model predictive control can reduce the number of calculations to one through deadbeat model predictive control, which reduces the amount of control system calculations compared to finite set model predictive control. However, in order to achieve multi-objective optimization of output voltage quality and system efficiency, a modulation strategy needs to be combined with the continuous set model predictive control. Summary of the invention

[0003] The present invention proposes a control strategy and a modulation method for a three-phase four-wire three-level inverter, which can achieve multi-objective optimization control of output voltage quality and system efficiency under balanced / unbalanced load conditions.

[0004] The present invention adopts the following technical solutions.

[0005] A modulation method for a three-phase four-wire three-level inverter, wherein the modulation method selects a modulation method of dual modulation wave carrier modulation DMW-CBPWM or sinusoidal pulse width modulation SPWM by judging the fluctuation amplitude of the DC side midpoint voltage of the three-level inverter and the strength of the midpoint voltage balancing ability of each phase, and comprehensively considering the DC side midpoint voltage balance and system efficiency by combining a fuzzy control method, so as to achieve balanced control of the DC side midpoint voltage balance and system efficiency, such as Figure 1 shown.

[0006] A control strategy for a three-phase four-wire three-level inverter, the control strategy is a continuous set model predictive control strategy including a three-phase four-wire three-level inverter modulation method, the continuous set model predictive control strategy first establishes its mathematical model according to the working principle of the three-phase four-wire three-level inverter, and then uses the forward Euler method to discretize the mathematical model, and calculates the output voltage prediction value at the next two moments, and obtains the inverter side bridge arm voltage reference value at the current moment in combination with the output voltage cost function, and inputs it into the modulation module after passing through the notch filter, and obtains the driving signal for driving the three-level inverter switch tube through the modulation method in the modulation module, such as Figure 2 As shown in the figure, the three-phase filter inductance value in the model predictive control strategy is obtained by a real-time filter inductance parameter identification method with low-pass filtering.

[0007] The control strategy achieves multi-objective optimization control of output voltage quality and system efficiency of a three-phase four-wire three-level inverter with balanced / unbalanced loads by suppressing the resonance generated by the interaction between the filter and the load, obtaining a more accurate filter inductance value of model predictive control, eliminating the influence of digital control delay, reducing midpoint voltage fluctuation and reducing the number of switches;

[0008] The three-phase four-wire three-level inverter adopts a split capacitor three-level inverter structure, which consists of an input source, a DC side capacitor composed of two equal capacitors connected in series, a three-bridge arm three-level inverter bridge, a three-phase LC filter and a three-phase load. The neutral point of the inverter AC side load is connected to the neutral point of the filter capacitor, and the midpoint of the DC side capacitor is connected to the neutral point of the inverter AC side load to form a three-phase four-wire system, which provides a flow path for the zero-sequence current generated when the inverter output is connected to an unbalanced load. Figure 3 A three-phase four-wire three-level inverter circuit topology is shown - a three-phase four-wire T-type three-level inverter.

[0009] The control strategy of the three-phase four-wire three-level inverter includes the following steps:

[0010] Step S1: Sampling the output voltage V of the three-phase three-level inverter ca 、V cb 、V cc , inverter side three-phase filter inductor current i a 、i b 、i c and the inverter three-phase load current i oa 、i ob 、i oc , the mathematical model of the system is established as

[0011]

[0012] Where: La , L b , L c are the filter inductors of phases A, B, and C of the three-phase LC filter; R and C are the parasitic resistance and filter capacitance of the filter inductor of the three-phase LC filter; V ao 、V bo 、V co They are the bridge arm voltages of the A, B, and C phase inverter sides respectively;

[0013] Formula 1 and Formula 2 are discretized by the forward Euler method, and the system discrete model is:

[0014]

[0015] Where: T s is a sampling period;

[0016] i a (k), i b (k), i c (k) represents the filter inductor current of the A, B, and C phase inverter side at time k;

[0017] V ao (k), V bo (k), V co (k) represents the A, B, and C phase inverter side bridge arm voltages at time k; V ca (k), V cb (k), V cc (k) represents the output voltage of phases A, B, and C at time k; i oa (k), i ob (k), i oc (k) represents the load current of phases A, B, and C at time k;

[0018] i a (k+1), i b (k+1), i c (k+1) represents the filter inductor current of the A, B, and C phase inverter side at time k+1; V ao (k+1), V bo (k+1), V co (k+1) represents the A, B, and C phase inverter side bridge arm voltages at time k+1; V ca (k+1), V cb (k+1), V cb (k+1) represents the output voltage of phases A, B, and C at time k+1; i oa (k+1), i ob (k+1), i oc (k+1) represents the load current of phases A, B, and C at time k+1.

[0019] Step S2: From step S1, the predicted value V of the output voltage of phases A, B, and C at time k+2 is obtained. ca (k+2), V cb (k+2), V cc (k+2) is

[0020]

[0021] Step S3: Cost function g of output voltage of phases A, B and C ca (k), g cb (k), g cc (k) is:

[0022]

[0023] Where: V * ca (k+2), V * cb (k+2), V * cc (k+2) is the output voltage reference value of phases A, B, and C at time k+2;

[0024] In order to accurately track the inverter three-phase output voltage reference, the cost function is minimized, V ca (k+2)=V * ca (k+2), V cb (k+2)=V * cb (k+2), V cc (k+2)=V * cc (k+2), and since i oa (k+1)≈i oa (k), i ob (k+1)≈i ob (k), i oc (k+1)≈i oc (k), and obtain the A, B, and C phase inverter side bridge arm voltage reference V at time k * ao (k), V * bo (k), V * co (k):

[0025]

[0026] Step S4: The A, B, and C phase inverter side bridge arm voltage reference V generated at time k in step S3 * ao (k), V* bo (k), V * co (k), after passing through the notch filter, we get V * ao_f (k), V * bo_f (k), V * co_f (k), V * ao_f (k), V * bo_f (k), V * co_f (k) is input into the modulation module and combined with the capacitor voltage V on the DC side of the inverter obtained at time k up (k), the capacitor voltage V on the DC side of the inverter low (k) and the three-phase inverter side filter inductor current i a (k), i b (k), i c (k) The driving signal S for driving the three-phase switch tube of the three-level inverter is obtained by the modulation method in the modulation module a1 ~S a4 , S b1 ~S b4 and S c1 ~S c4 ;

[0027] After passing through the notch filter, the A, B, and C phase inverter side bridge arm voltage reference V at time k is obtained * ao_f (k), V * bo_f (k), V * co_f (k), the expression is:

[0028]

[0029] Where:

[0030]

[0031] V * ao (k-1), V * bo (k-1), V * co (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-1, V * ao (k-2), V * bo (k-2), V* co (k-2) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-2 respectively; V * ao_f (k-1), V * bo_f (k-1), V * co_f (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference after passing through the notch filter at time k-1; V * ao_f (k-2), V * bo_f (k-2), V * co_f (k-2) respectively represent the A-phase, B-phase, and C-phase inverter side bridge arm voltage references after passing through the notch filter at time k-2;

[0032] Step S5: If there is a phase lag after the notch filter is used, the output voltage reference values ​​V of the three phases A, B and C are adjusted. ca * (k+2), V cb * (k+2), V cc * (k+2) performs phase correction;

[0033] Step S6: Obtain the three-phase filter inductance value in the model predictive control strategy through a real-time filter inductance parameter identification method with low-pass filtering.

[0034] Considering the lagging effect of the notch filter on the output voltage phase, the voltage reference V * ao (k), V * bo (k), V * co (k) The output voltage reference value of phase A, B, and C at time k+2 in the expression V ca * (k+2), V cb * (k+2), V cc * (k+2) is

[0035]

[0036] Where: V m is the output reference voltage amplitude; ω n is the output reference voltage angular frequency; ψ x is the deviation angle of the fundamental signal after passing through the notch filter; T0 is the fundamental period; ωn =2π / T0.

[0037] In step S4, the transfer function of the notch filter in the continuous domain is:

[0038]

[0039] Where: Q is the quality factor of the notch filter; ω f is the notch frequency;

[0040] The amplitude frequency A(ω) and phase frequency of the notch filter Features:

[0041]

[0042] In order to ensure the resonance suppression effect, ω f Equal to the system resonant angular frequency ω r , notch width Δω=0.1ω r , at frequency ω r ±Δω notch depth A(ω r ±Δω) is less than or equal to the amplitude of the negative system output voltage resonant frequency; the notch filter transfer function is discretized by the bilinear transformation method, and the discretized notch filter transfer function expression is obtained as follows:

[0043]

[0044] Where: T s is the sampling period.

[0045] In step S6, the three-phase filter inductance value L in the model predictive control strategy is a (k), L b (k), L c (k) is obtained by a real-time filter inductor parameter identification method with low-pass filtering, that is, by predicting the filter inductor current i a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c The error of (k) is used to identify the filter inductance parameters, and a low-pass filter is used to increase the identification range of the filter inductance; the specific identification method is as follows: the filter inductance current values ​​i of the A, B, and C phases predicted by using a fixed filter inductance value L0 at time k a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c(k) The following relationship is satisfied:

[0046]

[0047] Where: L' a (k), L' b (k), L' c (k) are the real-time identification values ​​of the filter inductances of phases A, B, and C without low-pass filtering at time k;

[0048] In one sampling period, according to the volt-second balance principle, we can get:

[0049]

[0050] Where: t a_p (k), t b_p (k), t c_p (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) P state action time; t a_n (k), t b_n (k), t c_n (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) N state action time;

[0051] The real-time identification value L' of the filter inductance of phases A, B, and C without low-pass filtering is obtained by formula 13, formula 14, and formula 15. a (k), L' b (k), L' c (k) is:

[0052]

[0053] When the numerator or denominator of formula 16 is 0, the linear extrapolation method is used to recalculate the real-time identification values ​​of the filter inductances of phases A, B, and C without the low-pass filter at time k. The expression is:

[0054]

[0055] Identification value L' of filter inductance for phases A, B, and C a (k), L' b (k), L' c (k) Perform low-pass filtering to obtain the real-time identification value L of the filter inductance of phases A, B, and C a (k), L b(k), L c (k). The transfer function of the low-pass filter discretized by bilinear transformation method is:

[0056]

[0057] Where: c is the -3dB cutoff frequency of the low-pass filter;

[0058] Real-time identification value L of the filter inductance of phase A, B, and C after low-pass filter a (k), L b (k), L c (k) is:

[0059]

[0060] like Figure 1 As shown, a three-phase four-wire three-level inverter modulation method combined with a fuzzy control method, and its implementation method includes the following steps;

[0061] Step A1: The A, B, and C phase inverter side bridge arm voltage reference V * ao_f 、V * bo_f 、V * co_f And the sampled three-phase inverter side filter inductor current i a 、i b 、i c and the inverter DC side midpoint voltage fluctuation ΔV NP , calculate the strength coefficient NPCM of the potential balance capability of the midpoint of phases A, B, and C a 、NPCM b 、NPCM c , the expression is

[0062]

[0063] Where: ΔV NP =V up -V low ; V dc =V up +V low ;

[0064]

[0065] The SPWM modulation strategy has only two switching states, P, O or N, O, in one sampling cycle, and the midpoint voltage fluctuates; the modulation principle of DMW-CBPWM is to decompose the O state action time of one phase selected to balance the midpoint voltage into equal P and N state action times according to the midpoint voltage fluctuation amplitude in one sampling cycle, so as to achieve midpoint voltage balance control; only when the intensity coefficient of the midpoint potential balancing ability is positive can the midpoint voltage fluctuation be suppressed by decomposing the O state time of the phase, and the greater the positive value of the intensity coefficient of the midpoint potential balancing ability, the stronger the midpoint balancing ability of the corresponding phase; compared with the SPWM modulation strategy, DMW-CBPWM increases the number of switches and the switching loss, but has the ability to balance the midpoint voltage;

[0066] The O state decomposition time Δt of one phase selected to balance the midpoint voltage dis for:

[0067]

[0068] Where: i x is the inverter side filter inductor current of the decomposed phase;

[0069] When modulating phase A, the DWM-CBPWM modulation method is to use the inverter side bridge arm voltage reference V * ao_f Decomposed into positive modulation signal V * ao-fp and negative modulation signal V * ao-fn Two modulation signals are used to perform pulse width modulation with the upper and lower stacked carriers to obtain the action time of the P, O, and N states; in the positive half cycle of the modulation signal, V * ao-fp =V * ao_f +Δv,V * ao-fn =-Δv; The modulation signals and switch states of the two modulation methods are as follows Figure 4 As shown; in the negative half cycle of the modulation signal, V * ao-fp =Δv, V * ao-fn =V * ao_f -Δv; Δv is the change of DMW-CBPWM modulation signal, and its expression is:

[0070]

[0071] Step A2: In one sampling period, the midpoint potential balance strength NPCM of phases A, B, and C is calculated by step A1. a 、NPCMb 、NPCM c , select NPCM a 、NPCM b 、NPCM c The maximum value is the maximum midpoint control margin NPCM max When NPCM max When it is greater than 0, the phase with the strongest midpoint balancing ability is selected and the fuzzy control method is combined to determine whether the phase uses DMW-CBPWM with midpoint voltage balancing control ability or SPWM with less switching times, and the other two phases use SPWM; when NPCM max When it is less than or equal to 0, the three phases A, B, and C all use SPWM;

[0072] Step A3: In step A2, the fuzzy control method is combined to determine the phase with the strongest midpoint balancing ability. The method of using DMW-CBPWM or SPWM is: the midpoint voltage fluctuation ΔV NP With the non-negative maximum midpoint control margin NPCM max As the input of the fuzzy controller, the comparison coefficient h is the output of the fuzzy controller;

[0073] When h is less than NPCM max When h is greater than or equal to NPCM, the phase with the strongest midpoint balancing ability uses DMW-CBPWM; when h is greater than or equal to NPCM max When the midpoint balancing capability is the strongest, SPWM is used.

[0074] Step A4: The fuzzy control process in step A3 is to define the input ΔV NP and NPCM max , the fuzzy subset of the output h and the corresponding domain range, and construct the membership function of the input and output;

[0075] The membership function is used to obtain the membership of each element in the corresponding fuzzy subset for the input quantity, and a fuzzy control rule table of the comparison coefficient h is set; the output of the comparison coefficient h is output based on the membership of each element in the corresponding fuzzy subset and the fuzzy control rule table, and the output of the comparison coefficient h is defuzzified to obtain the value of the comparison coefficient h.

[0076] The fuzzy control method is implemented with a fuzzy controller;

[0077] The fuzzification method of the fuzzy control method is as follows: NP With NPCM max As the input of the fuzzy controller; Based on the system parameters shown in Table 1 and considering the extreme working condition of single-phase load disconnection during operation, ΔV NPThe domain of the word is set to [-14,14], and five fuzzy subsets are defined: negative large (NB), negative small (NS), very small (ZE), positive small (PS), positive large (PB); NPCM max The domain is set to [0,6.5], and four fuzzy subsets are defined: very small (VS), small (S), large (B), and very large (VB); the domain of h is set to [0,9], and three fuzzy subsets are defined: small (S), medium (M), and large (B). Define the membership function to convert the input and output into fuzzy variables. The membership functions of both the input and output use triangular functions, such as Figure 5 to Figure 7 shown.

[0078] The fuzzy control method includes fuzzy reasoning, specifically: according to the input ΔV NP With NPCM max The reasoning principle for establishing fuzzy rules is ΔV NP The main control goal is to reduce switching losses. max is very small, the output comparison coefficient h is large; |ΔV NP | When it is small, if NPCM max If it is very small or small, the output comparison coefficient h is |ΔV NP | When the midpoint voltage is balanced, only in NPCM max When the value is very small, the output comparison coefficient h is small, and in other cases h is 0. The fuzzy control rule table of the comparison coefficient h is set according to fuzzy reasoning, as shown in Table 2.

[0079] The fuzzy control method also includes a defuzzification method. The defuzzification method uses a centroid method to defuzzify the output of the comparison coefficient h obtained according to the fuzzy control rule, and finally obtains the value of the comparison coefficient h.

[0080] The present invention adopts the above control strategy to suppress the resonance generated by the interaction between the filter and the load, obtain a more accurate model predictive control filter inductance value, eliminate the influence of digital control delay, and improve the output voltage quality; the present invention adopts the above modulation method to achieve balanced control of midpoint voltage balance and system efficiency, and then achieve balanced control of output voltage quality and system efficiency. The continuous set model predictive control strategy and modulation method proposed by the present invention can effectively achieve multi-objective optimization control of output voltage quality and system efficiency when the three-phase four-wire three-level inverter is loaded with balanced / unbalanced loads.

[0081] The present invention relates to a continuous set model predictive control strategy and modulation method applied to a three-phase four-wire three-level inverter, belonging to the technical field of power electronic converters. The continuous set model predictive control strategy is to first establish its mathematical model according to the working principle of the three-phase four-wire three-level inverter, and then use the forward Euler method to discretize the mathematical model, and calculate the output voltage prediction value of the next two moments, and obtain the inverter side bridge arm voltage reference value at the current moment in combination with the output voltage cost function, and input it into the modulation module after passing through the notch filter, and obtain the driving signal for driving the three-level inverter switch tube through the modulation method in the modulation module, wherein the three-phase filter inductance value in the model predictive control strategy is obtained by a real-time filter inductance parameter identification method with low-pass filtering. The modulation method is to judge the fluctuation amplitude of the DC side midpoint voltage of the three-level inverter and the balance ability of the midpoint voltage of each phase, and combine the fuzzy control method to comprehensively consider the DC side midpoint voltage balance and system efficiency to select dual modulation wave carrier modulation DMW-CBPWM or sine pulse width modulation SPWM to achieve balanced control of the DC side midpoint voltage balance and system efficiency. The present invention can suppress the resonance generated by the interaction between the filter and the load, obtain a more accurate filter inductance value of the model predictive control, eliminate the influence of digital control delay, reduce the midpoint voltage fluctuation and reduce the number of switches. The present invention can effectively realize the multi-objective optimization control of the output voltage quality and system efficiency of the three-phase four-wire three-level inverter with balanced / unbalanced load.

[0082] The present invention adopts a fuzzy control method, does not require an accurate mathematical model, is easy to include nonlinear constraints, has an intuitive implementation method, and is easy to implement system adaptive control. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0084] Attached Figure 1 It is a flow chart of a modulation method combined with a fuzzy control method in one embodiment of the present invention;

[0085] Attached Figure 2 A schematic diagram of a system control strategy in an embodiment of the present invention;

[0086] Attached Figure 3 A schematic diagram of a three-phase four-wire three-level inverter - a three-phase four-wire T-type three-level inverter in one embodiment of the present invention;

[0087] Attached Figure 4 A schematic diagram of the modulation signal positive half cycle A-phase SPWM and DWM-CBPWM modulation signals and switch states in one embodiment of the present invention;

[0088] Attached Figure 5 is the fuzzy control input ΔV in one embodiment of the present inventionNP Schematic diagram of membership function;

[0089] Attached Figure 6 is the fuzzy control input quantity NPCM in one embodiment of the present invention max Schematic diagram of membership function;

[0090] Attached Figure 7 Schematic diagram of the membership function of the fuzzy control output h in one embodiment of the present invention. DETAILED DESCRIPTION

[0091] A modulation method for a three-phase four-wire three-level inverter, wherein the modulation method selects a modulation method of dual modulation wave carrier modulation DMW-CBPWM or sinusoidal pulse width modulation SPWM by judging the fluctuation amplitude of the DC side midpoint voltage of the three-level inverter and the strength of the midpoint voltage balancing ability of each phase, and comprehensively considering the DC side midpoint voltage balance and system efficiency by combining a fuzzy control method, so as to achieve balanced control of the DC side midpoint voltage balance and system efficiency, such as Figure 1 shown.

[0092] A control strategy for a three-phase four-wire three-level inverter, the control strategy is a continuous set model predictive control strategy including a three-phase four-wire three-level inverter modulation method, the continuous set model predictive control strategy first establishes its mathematical model according to the working principle of the three-phase four-wire three-level inverter, and then uses the forward Euler method to discretize the mathematical model, and calculates the output voltage prediction value at the next two moments, and obtains the inverter side bridge arm voltage reference value at the current moment in combination with the output voltage cost function, and inputs it into the modulation module after passing through the notch filter, and obtains the driving signal for driving the three-level inverter switch tube through the modulation method in the modulation module, wherein the three-phase filter inductance value in the model predictive control strategy is obtained by a real-time filter inductance parameter identification method with low-pass filtering, such as Figure 2 shown.

[0093] The control strategy achieves multi-objective optimization control of output voltage quality and system efficiency of a three-phase four-wire three-level inverter with balanced / unbalanced loads by suppressing the resonance generated by the interaction between the filter and the load, obtaining a more accurate filter inductance value of model predictive control, eliminating the influence of digital control delay, reducing midpoint voltage fluctuation and reducing the number of switches;

[0094] The three-phase four-wire three-level inverter adopts a split capacitor three-level inverter structure, which consists of an input source, a DC side capacitor composed of two equal capacitors connected in series, a three-bridge arm three-level inverter bridge, a three-phase LC filter and a three-phase load. The neutral point of the inverter AC side load is connected to the neutral point of the filter capacitor, and the midpoint of the DC side capacitor is connected to the neutral point of the inverter AC side load to form a three-phase four-wire system, which provides a flow path for the zero-sequence current generated when the inverter output is connected to an unbalanced load. Figure 3 A three-phase four-wire three-level inverter circuit topology is shown - a three-phase four-wire T-type three-level inverter.

[0095] The control strategy of the three-phase four-wire three-level inverter includes the following steps:

[0096] Step S1: Sampling the output voltage V of the three-phase three-level inverter ca 、V cb 、V cc , inverter side three-phase filter inductor current i a 、i b 、i c and the inverter three-phase load current i oa 、i ob 、i oc , the mathematical model of the system is established as

[0097]

[0098] Where: L a , L b , L c are the filter inductors of phases A, B, and C of the three-phase LC filter; R and C are the parasitic resistance and filter capacitance of the filter inductor of the three-phase LC filter; V ao 、V bo 、V co They are the bridge arm voltages of the A, B, and C phase inverter sides respectively;

[0099] Formula 1 and Formula 2 are discretized by the forward Euler method, and the system discrete model is:

[0100]

[0101] Where: T s is a sampling period;

[0102] i a (k), i b (k), i c (k) represents the filter inductor current of the A, B, and C phase inverter side at time k;

[0103] V ao (k), V bo (k), V co (k) represents the A, B, and C phase inverter side bridge arm voltages at time k; V ca (k), V cb (k), V cc (k) represents the output voltage of phases A, B, and C at time k; i oa (k), iob (k), i oc (k) represents the load current of phases A, B, and C at time k;

[0104] i a (k+1), i b (k+1), i c (k+1) represents the filter inductor current of the A, B, and C phase inverter side at time k+1; V ao (k+1), V bo (k+1), V co (k+1) represents the A, B, and C phase inverter side bridge arm voltages at time k+1; V ca (k+1), V cb (k+1), V cb (k+1) represents the output voltage of phases A, B, and C at time k+1; i oa (k+1), i ob (k+1), i oc (k+1) represents the load current of phases A, B, and C at time k+1.

[0105] Step S2: From step S1, the predicted value V of the output voltage of phases A, B, and C at time k+2 is obtained. ca (k+2), V cb (k+2), V cc (k+2) is

[0106]

[0107] Step S3: Cost function g of output voltage of phases A, B and C ca (k), g cb (k), g cc (k) is:

[0108]

[0109] Where: V * ca (k+2), V * cb (k+2), V * cc (k+2) is the output voltage reference value of phases A, B, and C at time k+2;

[0110] In order to accurately track the inverter three-phase output voltage reference, the cost function is minimized, V ca (k+2)=V * ca (k+2), V cb (k+2)=V * cb (k+2), Vcc (k+2)=V * cc (k+2), and since i oa (k+1)≈i oa (k), i ob (k+1)≈i ob (k), i oc (k+1)≈i oc (k), and obtain the A, B, and C phase inverter side bridge arm voltage reference V at time k * ao (k), V * bo (k), V * co (k):

[0111]

[0112] Step S4: The A, B, and C phase inverter side bridge arm voltage reference V generated at time k in step S3 * ao (k), V * bo (k), V * co (k), after passing through the notch filter, we get V * ao_f (k), V * bo_f (k), V * co_f (k), V * ao_f (k), V * bo_f (k), V * co_f (k) is input into the modulation module and combined with the capacitor voltage V on the DC side of the inverter obtained at time k up (k), the capacitor voltage V on the DC side of the inverter low (k) and the three-phase inverter side filter inductor current i a (k), i b (k), i c (k) The driving signal S for driving the three-phase switch tube of the three-level inverter is obtained by the modulation method in the modulation module a1 ~S a4 , S b1 ~S b4 and S c1 ~S c4 ;

[0113] After passing through the notch filter, the A, B, and C phase inverter side bridge arm voltage reference V at time k is obtained *ao_f (k), V * bo_f (k), V * co_f (k), the expression is:

[0114]

[0115] Where:

[0116]

[0117] V * ao (k-1), V * bo (k-1), V * co (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-1, V * ao (k-2), V * bo (k-2), V * co (k-2) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-2 respectively; V * ao_f (k-1), V * bo_f (k-1), V * co_f (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference after passing through the notch filter at time k-1; V * ao_f (k-2), V * bo_f (k-2), V * co_f (k-2) respectively represent the A-phase, B-phase, and C-phase inverter side bridge arm voltage references after passing through the notch filter at time k-2;

[0118] Step S5: If there is a phase lag after the notch filter is used, the output voltage reference values ​​V of the three phases A, B and C are adjusted. ca * (k+2), V cb * (k+2), V cc * (k+2) performs phase correction;

[0119] Step S6: Obtain the three-phase filter inductance value in the model predictive control strategy through a real-time filter inductance parameter identification method with low-pass filtering.

[0120] Considering the lagging effect of the notch filter on the output voltage phase, the voltage reference V * ao (k), V * bo (k), V * co (k) The output voltage reference value of phase A, B, and C at time k+2 in the expression V ca * (k+2), V cb * (k+2), V cc * (k+2) is

[0121]

[0122] Where: V m is the output reference voltage amplitude; ω n is the output reference voltage angular frequency; ψ x is the deviation angle of the fundamental signal after passing through the notch filter; T0 is the fundamental period; ω n =2π / T0.

[0123] In step S4, the transfer function of the notch filter in the continuous domain is:

[0124]

[0125] Where: Q is the quality factor of the notch filter; ω f is the notch frequency;

[0126] The amplitude frequency A(ω) and phase frequency of the notch filter Features:

[0127]

[0128] In order to ensure the resonance suppression effect, ω f Equal to the system resonant angular frequency ω r , notch width Δω=0.1ω r , at frequency ω r ±Δω notch depth A(ω r ±Δω) is less than or equal to the amplitude of the negative system output voltage resonant frequency; the notch filter transfer function is discretized by the bilinear transformation method, and the discretized notch filter transfer function expression is obtained as follows:

[0129]

[0130] Where: T s is the sampling period.

[0131] In step S6, the three-phase filter inductance value L in the model predictive control strategy is a (k), L b (k), L c (k) is obtained by a real-time filter inductor parameter identification method with low-pass filtering, that is, by predicting the filter inductor current i a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c The error of (k) is used to identify the filter inductance parameters, and a low-pass filter is used to increase the identification range of the filter inductance; the specific identification method is as follows: the filter inductance current values ​​i of the A, B, and C phases predicted by using a fixed filter inductance value L0 at time k a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c (k) The following relationship is satisfied:

[0132]

[0133] Where: L' a (k), L' b (k), L' c (k) are the real-time identification values ​​of the filter inductances of phases A, B, and C without low-pass filtering at time k;

[0134] In one sampling period, according to the volt-second balance principle, we can get:

[0135]

[0136] Where: t a_p (k), t b_p (k), t c_p (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) P state action time; t a_n (k), t b_n (k), t c_n (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) N state action time;

[0137] The real-time identification value L' of the filter inductance of phases A, B, and C without low-pass filtering is obtained by formula 13, formula 14, and formula 15. a (k), L' b (k), L' c (k) is:

[0138]

[0139] When the numerator or denominator of formula 16 is 0, the linear extrapolation method is used to recalculate the real-time identification values ​​of the filter inductances of phases A, B, and C without the low-pass filter at time k. The expression is:

[0140]

[0141] Identification value L' of filter inductance for phases A, B, and C a (k), L' b (k), L' c (k) Perform low-pass filtering to obtain the real-time identification value L of the filter inductance of phases A, B, and C a (k), L b (k), L c (k). The transfer function of the low-pass filter discretized by bilinear transformation method is:

[0142]

[0143] Where: c is the -3dB cutoff frequency of the low-pass filter;

[0144] Real-time identification value L of the filter inductance of phase A, B, and C after low-pass filter a (k), L b (k), L c (k) is:

[0145]

[0146] like Figure 1 As shown, a three-phase four-wire three-level inverter modulation method combined with a fuzzy control method, and its implementation method includes the following steps;

[0147] Step A1: The A, B, and C phase inverter side bridge arm voltage reference V * ao_f 、V * bo_f 、V * co_f And the sampled three-phase inverter side filter inductor current i a 、i b 、i cand the inverter DC side midpoint voltage fluctuation ΔV NP , calculate the strength coefficient NPCM of the potential balance capability of the midpoint of phases A, B, and C a 、NPCM b 、NPCM c , the expression is

[0148]

[0149] Where: ΔV NP =V up -V low ; V dc =V up +V low ;

[0150]

[0151] The SPWM modulation strategy has only two switching states, P, O or N, O, in one sampling cycle, and the midpoint voltage fluctuates; the modulation principle of DMW-CBPWM is to decompose the O state action time of one phase selected to balance the midpoint voltage into equal P and N state action times according to the midpoint voltage fluctuation amplitude in one sampling cycle, so as to achieve midpoint voltage balance control; only when the intensity coefficient of the midpoint potential balancing ability is positive can the midpoint voltage fluctuation be suppressed by decomposing the O state time of the phase, and the greater the positive value of the intensity coefficient of the midpoint potential balancing ability, the stronger the midpoint balancing ability of the corresponding phase; compared with the SPWM modulation strategy, DMW-CBPWM increases the number of switches and the switching loss, but has the ability to balance the midpoint voltage;

[0152] The O state decomposition time Δt of one phase selected to balance the midpoint voltage dis for:

[0153]

[0154] Where: i x is the inverter side filter inductor current of the decomposed phase;

[0155] When modulating phase A, the DWM-CBPWM modulation method is to use the inverter side bridge arm voltage reference V * ao_f Decomposed into positive modulation signal V * ao-fp and negative modulation signal V * ao-fn Two modulation signals are used to perform pulse width modulation with the upper and lower stacked carriers to obtain the action time of the P, O, and N states; in the positive half cycle of the modulation signal, V * ao-fp =V *ao_f +Δv,V * ao-fn =-Δv; The modulation signals and switch states of the two modulation methods are as follows Figure 4 As shown; in the negative half cycle of the modulation signal, V * ao-fp =Δv, V * ao-fn =V * ao_f -Δv; Δv is the change of DMW-CBPWM modulation signal, and its expression is:

[0156]

[0157] Step A2: In one sampling period, the midpoint potential balance strength NPCM of phases A, B, and C is calculated by step A1. a 、NPCM b 、NPCM c , select NPCM a 、NPCM b 、NPCM c The maximum value is the maximum midpoint control margin NPCM max When NPCM max When it is greater than 0, the phase with the strongest midpoint balancing ability is selected and the fuzzy control method is combined to determine whether the phase uses DMW-CBPWM with midpoint voltage balancing control ability or SPWM with less switching times, and the other two phases use SPWM; when NPCM max When it is less than or equal to 0, the three phases A, B, and C all use SPWM;

[0158] Step A3: In step A2, the fuzzy control method is combined to determine the phase with the strongest midpoint balancing ability. The method of using DMW-CBPWM or SPWM is: the midpoint voltage fluctuation ΔV NP With the non-negative maximum midpoint control margin NPCM max As the input of the fuzzy controller, the comparison coefficient h is the output of the fuzzy controller;

[0159] When h is less than NPCM max When h is greater than or equal to NPCM, the phase with the strongest midpoint balancing ability uses DMW-CBPWM; when h is greater than or equal to NPCM max When the midpoint balancing capability is the strongest, SPWM is used.

[0160] Step A4: The fuzzy control process in step A3 is to define the input ΔV NP and NPCM max , the fuzzy subset of the output h and the corresponding domain range, and construct the membership function of the input and output;

[0161] The membership function is used to obtain the membership of each element in the corresponding fuzzy subset for the input quantity, and a fuzzy control rule table of the comparison coefficient h is set; the output of the comparison coefficient h is output based on the membership of each element in the corresponding fuzzy subset and the fuzzy control rule table, and the output of the comparison coefficient h is defuzzified to obtain the value of the comparison coefficient h.

[0162] The following is a fuzzy control design taking a specific system as an example. The system parameters are shown in Table 1.

[0163] Table 1

[0164]

[0165] The fuzzy control method is implemented with a fuzzy controller;

[0166] The fuzzification method of the fuzzy control method is as follows: NP With NPCM max As the input of the fuzzy controller; based on the system parameters and considering the extreme conditions of single-phase load disconnection during operation, ΔV NP The domain of the word is set to [-14,14], and five fuzzy subsets are defined: negative large (NB), negative small (NS), very small (ZE), positive small (PS), positive large (PB); NPCM max The domain is set to [0,6.5], and four fuzzy subsets are defined: very small (VS), small (S), large (B), and very large (VB); the domain of h is set to [0,9], and three fuzzy subsets are defined: small (S), medium (M), and large (B). Define the membership function to convert the input and output into fuzzy variables. The membership functions of both the input and output use triangular functions, such as Figure 5 to Figure 7 shown.

[0167] ΔV NP The specific expression of the membership function is:

[0168]

[0169]

[0170] NPCM max The specific expression of the membership function is:

[0171]

[0172]

[0173] The specific expression of h membership function is:

[0174]

[0175] The fuzzy control method includes fuzzy reasoning, specifically: according to the input ΔV NP With NPCM max The reasoning principle for establishing fuzzy rules is ΔV NP The main control goal is to reduce switching losses. max is very small, the output comparison coefficient h is large; |ΔV NP | When it is small, if NPCM max If it is very small or small, the output comparison coefficient h is |ΔV NP | When the midpoint voltage is balanced, only in NPCM max When the value is very small, the output comparison coefficient h is small, and in other cases h is 0. The fuzzy control rule table of the comparison coefficient h is set according to fuzzy reasoning, as shown in Table 2.

[0176] Table 2

[0177]

[0178] The fuzzy control method also includes a defuzzification method. The defuzzification method uses a centroid method to defuzzify the output of the comparison coefficient h obtained according to the fuzzy control rule, and finally obtains the value of the comparison coefficient h.

[0179] The above is only a preferred embodiment of the present invention, and does not limit the present invention in other forms. Any technician familiar with the profession may use the above disclosed technical content to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment according to the technical essence of the present invention without departing from the technical solution of the present invention still belongs to the protection scope of the technical solution of the present invention.

Claims

1. A modulation method for a three-phase four-wire three-level inverter, characterized in that: The modulation method selects the modulation method of dual modulation wave carrier modulation DMW-CBPWM or sinusoidal pulse width modulation SPWM by judging the fluctuation amplitude of the DC side midpoint voltage of the three-level inverter and the strength of the midpoint voltage balancing ability of each phase, and comprehensively considering the DC side midpoint voltage balance and system efficiency by combining the fuzzy control method, so as to achieve balanced control of the DC side midpoint voltage balance and system efficiency.

2. The control strategy of the three-phase four-wire three-level inverter is characterized by: The control strategy is a continuous set model predictive control strategy including a three-phase four-wire three-level inverter modulation method. The continuous set model predictive control strategy first establishes its mathematical model according to the working principle of the three-phase four-wire three-level inverter, and then uses the forward Euler method to discretize the mathematical model, and calculates the output voltage prediction values ​​at the next two moments, and obtains the inverter side bridge arm voltage reference value at the current moment in combination with the output voltage cost function, and inputs it into the modulation module after passing through the notch filter. The driving signal for driving the three-level inverter switch tube is obtained by the modulation method in the modulation module, wherein the three-phase filter inductance value in the model predictive control strategy is obtained by a real-time filter inductance parameter identification method with low-pass filtering.

3. The control strategy of the three-phase four-wire three-level inverter according to claim 2 is characterized in that: The control strategy achieves multi-objective optimization control of output voltage quality and system efficiency of a three-phase four-wire three-level inverter with balanced / unbalanced loads by suppressing the resonance generated by the interaction between the filter and the load, obtaining a more accurate filter inductance value of model predictive control, eliminating the influence of digital control delay, reducing midpoint voltage fluctuation and reducing the number of switches; The three-phase four-wire three-level inverter adopts a split capacitor three-level inverter structure, which consists of an input source, a DC side capacitor composed of two equal upper and lower capacitors connected in series, a three-bridge arm three-level inverter bridge, a three-phase LC filter and a three-phase load. The neutral point of the inverter AC side load is connected to the neutral point of the filter capacitor, and the midpoint of the DC side capacitor is connected to the neutral point of the inverter AC side load to form a three-phase four-wire system, which provides a flow path for the zero-sequence current generated when the inverter output is connected to an unbalanced load.

4. The control strategy of the three-phase four-wire three-level inverter according to claim 2, characterized in that: The steps include: Step S1: Sampling the output voltage V of the three-phase three-level inverter ca 、V cb 、V cc , inverter side three-phase filter inductor current i a 、i b 、i c and the inverter three-phase load current i oa 、i ob 、i oc , the mathematical model of the system is established as Where: L a , L b , L c are the filter inductors of phases A, B, and C of the three-phase LC filter respectively; R and C are the parasitic resistance and filter capacitance of the filter inductor of the three-phase LC filter respectively; V ao 、V bo 、V co They are the bridge arm voltages of the A, B, and C phase inverter sides respectively; Formula 1 and Formula 2 are discretized by the forward Euler method, and the system discrete model is: Where: T s is a sampling period; i a (k), i b (k), i c (k) represents the filter inductor current of the A, B, and C phase inverter side at time k; V ao (k), V bo (k), V co (k) represents the A, B, and C phase inverter side bridge arm voltages at time k; V ca (k), V cb (k), V cc (k) represents the output voltage of phases A, B, and C at time k; i oa (k), i ob (k), i oc (k) represents the load current of phases A, B, and C at time k; i a (k+1), i b (k+1), i c (k+1) represents the filter inductor current of the A, B, and C phase inverter side at time k+1; V ao (k+1), V bo (k+1), V co (k+1) represents the A, B, and C phase inverter side bridge arm voltages at time k+1; V ca (k+1), V cb (k+1), V cb (k+1) represents the output voltage of phases A, B, and C at time k+1; i oa (k+1), i ob (k+1), i oc (k+1) represents the load current of phases A, B, and C at time k+1. Step S2: From step S1, the predicted value V of the output voltage of phases A, B, and C at time k+2 is obtained. ca (k+2), V cb (k+2), V cc (k+2) is Step S3: Cost function g of output voltage of phases A, B and C ca (k), g cb (k), g cc (k) is: Where: V * ca (k+2), V * cb (k+2), V * cc (k+2) is the output voltage reference value of phases A, B, and C at time k+2; In order to accurately track the inverter three-phase output voltage reference, the cost function is minimized, V ca (k+2)=V * ca (k+2), V cb (k+2)=V * cb (k+2), V cc (k+2)=V * cc (k+2), and since i oa (k+1)≈i oa (k), i ob (k+1)≈i ob (k), i oc (k+1)≈i oc (k), and obtain the A, B, and C phase inverter side bridge arm voltage reference V at time k * ao (k), V * bo (k), V * co (k): Step S4: The A, B, and C phase inverter side bridge arm voltage reference V generated at time k in step S3 * ao (k), V * bo (k), V * co (k), after passing through the notch filter, we get V * ao_f (k), V * bo_f (k), V * co_f (k), V * ao_f (k), V * bo_f (k), V * co_f (k) is input into the modulation module and combined with the capacitor voltage V on the DC side of the inverter obtained at time k up (k), the capacitor voltage V on the DC side of the inverter low (k) and the three-phase inverter side filter inductor current i a (k), i b (k), i c (k) The driving signal S for driving the three-phase switch tube of the three-level inverter is obtained by the modulation method in the modulation module a1 ~S a4 , S b1 ~S b4 and S c1 ~S c4 ; After passing through the notch filter, the A, B, and C phase inverter side bridge arm voltage reference V at time k is obtained * ao_f (k), V * bo_f (k), V * co_f (k), the expression is: Where: V * ao (k-1), V * bo (k-1), V * co (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-1, V * ao (k-2), V * bo (k-2), V * co (k-2) represents the A, B, and C phase inverter side bridge arm voltage reference at time k-2 respectively; V * ao_f (k-1), V * bo_f (k-1), V * co_f (k-1) represents the A, B, and C phase inverter side bridge arm voltage reference after passing through the notch filter at time k-1; V * ao_f (k-2), V * bo_f (k-2), V * co_f (k-2) respectively represent the A-phase, B-phase, and C-phase inverter side bridge arm voltage references after passing through the notch filter at time k-2; Step S5: If there is a phase lag after the notch filter is used, the output voltage reference values ​​V of the three phases A, B and C are adjusted. ca * (k+2), V cb * (k+2), V cc * (k+2) performs phase correction; Step S6: Obtain the three-phase filter inductance value in the model predictive control strategy through a real-time filter inductance parameter identification method with low-pass filtering.

5. The control strategy of the three-phase four-wire three-level inverter according to claim 4 is characterized in that: Considering the lagging effect of the notch filter on the output voltage phase, the voltage reference V * ao (k), V * bo (k), V * co (k) The output voltage reference value of phase A, B, and C at time k+2 in the expression V ca * (k+2), V cb * (k+2), V cc * (k+2) is Where: V m is the output reference voltage amplitude; ω n is the output reference voltage angular frequency; ψ x is the deviation angle of the fundamental signal after passing through the notch filter; T0 is the fundamental period; ω n =2π / T0.

6. The control strategy of the three-phase four-wire three-level inverter according to claim 4, characterized in that: In step S4, the transfer function of the notch filter in the continuous domain is: Where: Q is the quality factor of the notch filter; ω f is the notch frequency; The amplitude frequency A(ω) and phase frequency of the notch filter Features: In order to ensure the resonance suppression effect, ω f Equal to the system resonant angular frequency ω r , notch width Δω=0.1ω r , at frequency ω r ±Δω notch depth A(ω r ±Δω) is less than or equal to the amplitude of the negative system output voltage resonant frequency; the notch filter transfer function is discretized by the bilinear transformation method, and the discretized notch filter transfer function expression is obtained as follows: Where: T s is the sampling period.

7. The control strategy of the three-phase four-wire three-level inverter according to claim 4, characterized in that: In step S6, the three-phase filter inductance value L in the model predictive control strategy is a (k), L b (k), L c (k) is obtained by a real-time filter inductor parameter identification method with low-pass filtering, that is, by predicting the filter inductor current i a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c (k) The error of the filter inductor is used to identify the filter inductor parameters, and a low-pass filter is used to increase the identification range of the filter inductor; The specific identification method is as follows: The filter inductance current values ​​i of phases A, B, and C predicted at time k using a fixed filter inductance value L0 a_p (k), i b_p (k), i c_p (k) and the actual sampled filter inductor current i a (k), i b (k), i c (k) The following relationship is satisfied: Where: L' a (k), L' b (k), L' c (k) are the real-time identification values ​​of the filter inductances of phases A, B, and C without low-pass filtering at time k; In one sampling period, according to the volt-second balance principle, we can get: Where: t a_p (k), t b_p (k), t c_p (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) P state action time; t a_n (k), t b_n (k), t c_n (k) represents a sampling period T at time k s Inner V ao (k), V bo (k), V co (k) N state action time; The real-time identification value L' of the filter inductance of phases A, B, and C without low-pass filtering is obtained by formula 13, formula 14, and formula 15. a (k), L' b (k), L' c (k) is: When the numerator or denominator of formula 16 is 0, the linear extrapolation method is used to recalculate the real-time identification values ​​of the filter inductances of phases A, B, and C without the low-pass filter at time k. The expression is: Identification value L' of filter inductance for phases A, B, and C a (k), L' b (k), L' c (k) Perform low-pass filtering to obtain the real-time identification value L of the filter inductance of phases A, B, and C a (k), L b (k), L c (k). The transfer function of the low-pass filter discretized by bilinear transformation method is: Where: c is the -3dB cutoff frequency of the low-pass filter; Real-time identification value L of the filter inductance of phase A, B, and C after low-pass filter a (k), L b (k), L c (k) is:

8. The modulation method of a three-phase four-wire three-level inverter according to claim 1, characterized in that: A three-phase four-wire three-level inverter modulation method combined with a fuzzy control method, wherein the implementation method comprises the following steps: Step A1: The A, B, and C phase inverter side bridge arm voltage reference V * ao_f 、V * bo_f 、V * co_f And the sampled three-phase inverter side filter inductor current i a 、i b 、i c and the inverter DC side midpoint voltage fluctuation ΔV NP , calculate the strength coefficient NPCM of the potential balance capability of the midpoint of phases A, B, and C a 、NPCM b 、NPCM c , the expression is Where: ΔV NP = V up - V low ; V dc = V up + V low ; The SPWM modulation strategy has only two switching states, P, O or N, O, in one sampling cycle, and the midpoint voltage fluctuates; the modulation principle of DMW-CBPWM is to decompose the O state action time of one phase selected to balance the midpoint voltage into equal P and N state action times according to the midpoint voltage fluctuation amplitude in one sampling cycle, so as to achieve midpoint voltage balance control; only when the intensity coefficient of the midpoint potential balancing ability is positive can the midpoint voltage fluctuation be suppressed by decomposing the O state time of the phase, and the greater the positive value of the intensity coefficient of the midpoint potential balancing ability, the stronger the midpoint balancing ability of the corresponding phase; compared with the SPWM modulation strategy, DMW-CBPWM increases the number of switches and the switching loss, but has the ability to balance the midpoint voltage; The O state decomposition time Δt of one phase selected to balance the midpoint voltage dis for: Where: i x is the inverter side filter inductor current of the decomposed phase; When modulating phase A, the DWM-CBPWM modulation method is to use the inverter side bridge arm voltage reference V * ao_f Decomposed into positive modulation signal V * ao-fp and negative modulation signal V * ao-fn Two modulation signals are used to perform pulse width modulation with the upper and lower stacked carriers to obtain the action time of the P, O, and N states; In the positive half cycle of the modulation signal, V * ao-fp =V * ao_f +Δv,V * ao-fn =-Δv; in the negative half cycle of the modulation signal, V * ao-fp =Δv, V * ao-fn =V * ao_f -Δv; Δv is the change of DMW-CBPWM modulation signal, and its expression is: Step A2: In one sampling period, the midpoint potential balance strength NPCM of phases A, B, and C is calculated by step A1. a 、NPCM b 、NPCM c , select NPCM a 、NPCM b 、NPCM c The maximum value is the maximum midpoint control margin NPCM max ; When NPCM max When it is greater than 0, the phase with the strongest midpoint balancing ability is selected and the fuzzy control method is combined to determine whether the phase uses DMW-CBPWM with midpoint voltage balancing control ability or SPWM with less switching times, and the other two phases use SPWM; when NPCM max When it is less than or equal to 0, the three phases A, B, and C all use SPWM; Step A3: In step A2, the fuzzy control method is combined to determine the phase with the strongest midpoint balancing ability. The method of using DMW-CBPWM or SPWM is: the midpoint voltage fluctuation ΔV NP With the non-negative maximum midpoint control margin NPCM max As the input of the fuzzy controller, the comparison coefficient h is used as the output of the fuzzy controller; when h is less than NPCM max When h is greater than or equal to NPCM, the phase with the strongest midpoint balancing ability uses DMW-CBPWM; when h is greater than or equal to NPCM max When the midpoint balancing capability is the strongest, SPWM is used. Step A4: The fuzzy control process in step A3 is to define the input ΔV NP and NPCM max , the fuzzy subset of the output h and the corresponding domain range, and construct the membership function of the input and output; The membership function is used to obtain the membership of each element in the corresponding fuzzy subset for the input quantity, and a fuzzy control rule table of the comparison coefficient h is set; the output of the comparison coefficient h is output based on the membership of each element in the corresponding fuzzy subset and the fuzzy control rule table, and the output of the comparison coefficient h is defuzzified to obtain the value of the comparison coefficient h.

9. The modulation method of the three-phase four-wire three-level inverter according to claim 8, characterized in that: The fuzzy control method is implemented with a fuzzy controller; The fuzzification method of the fuzzy control method is as follows: NP With NPCM max As the input of the fuzzy controller; based on the system parameters and considering the extreme conditions of single-phase load disconnection during operation, ΔV NP In the domain of , five fuzzy subsets are defined: negative large NB, negative small NS, very small ZE, positive small PS, positive large PB; NPCM max In the domain, four fuzzy subsets are defined: very small VS, small S, large B, and very large VB; in the domain of h, three fuzzy subsets are defined: small S, medium M, and large B; membership functions are defined to convert input and output quantities into fuzzy variables, and both input and output membership functions use triangular functions.

10. The modulation method of a three-phase four-wire three-level inverter according to claim 8, characterized in that: Fuzzy control methods include fuzzy reasoning and defuzzification. Fuzzy reasoning is as follows: according to the input ΔV NP With NPCM max The reasoning principle for establishing fuzzy rules is ΔV NP The main control goal is to reduce switching losses. max is very small, the output comparison coefficient h is large; |ΔV NP | When it is small, if NPCM max If it is very small or small, the output comparison coefficient h is |ΔV NP | When the midpoint voltage is balanced, only in NPCM max When the comparison coefficient h is very small, the output comparison coefficient h is small, and in other cases h is 0; the fuzzy control rule table of the comparison coefficient h is set according to fuzzy reasoning; The defuzzification method is to use the centroid method to defuzzify the output of the comparison coefficient h obtained according to the fuzzy control rule, and finally obtain the value of the comparison coefficient h.

Images