A T-type three-level grid-connected converter open-circuit fault diagnosis method

Abstract

The application provides an open-circuit fault diagnosis method for a T-type three-level grid-connected converter, and the steps are as follows: firstly, the operation condition of the T-type three-level grid-connected converter after an open-circuit fault occurs is analyzed, and the current flowing from the midpoint of the direct-current side capacitor to each phase is calculated according to the switch state; secondly, the average value of the current flowing from the midpoint to each phase and the average value of the converter output current are calculated, the average value of the current flowing from the midpoint to each phase and the average value of the converter output current are compared with a set threshold value respectively, and the fault position is located; finally, according to the fault position, the fault is divided into a type I fault and a type II fault; and the fault-tolerant structure of the type I fault and the type II fault is established respectively, and the model predictive control strategy is used for fault-tolerant control of the two types of faults. The application can quickly locate the fault power device position of the converter, realize continuous operation of fault tolerance, and improve the reliability of the converter.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology for T-type converters, and in particular to a method for diagnosing open-circuit faults in a T-type three-level grid-connected converter. Background Technology

[0002] To accelerate the construction of a green, low-carbon, and sustainable modern energy system, new energy power generation such as wind and solar power has developed rapidly. As a key device for connecting new energy sources to the grid, the reliability of the grid converter directly affects the power quality. The T-type three-level converter has advantages such as high efficiency, low output harmonic content, and more balanced power consumption distribution, and is widely used in the field of new energy power generation.

[0003] During the operation of photovoltaic power plants, converter failure accounts for 60% of common failures. The failure rate of T-type three-level grid-connected converters is increased due to factors such as electromagnetic interference, pulse current, and the operating environment. If a fault can be located and isolated promptly after it occurs, the reliability of grid-connected renewable energy generation will be improved. To ensure the safe and reliable operation of grid-connected power generation systems, much research is dedicated to improving the reliability of grid-connected converters, thereby enhancing the stability of renewable energy grid-connected systems.

[0004] Power device faults in grid-connected converters include open-circuit faults and short-circuit faults. Short-circuit faults can be protected using power device drive circuits. However, open-circuit faults generally do not cause overcurrent, making protection mechanisms less likely to activate. Therefore, open-circuit fault detection in grid-connected converters is crucial for improving system reliability.

[0005] There are two main types of open-circuit fault diagnosis methods: current-based and voltage-based methods. Current-based methods typically use phase current, midpoint current, and current residual to detect faults. The literature [U. Choi, J. Lee, F. Blaabjerg, K. Lee, et al. Open-circuit fault diagnosis and fault-tolerant control for a grid-connected NPC inverter, IEEE Transactions on Power Electronics, 2016, 31(10): 7234-7247.] determines the faulty phase based on the error between the actual output current and the reference current of the converter. Under different grid connection conditions, the degree of output current distortion varies, which may lead to misdiagnosis using this method. The literature [X.Ge, J.Pu, B.Gou, et al. An open-circuit fault diagnosis approach for single-phase three-level neutral-point-clamped converters[J].IEEE Trans.PowerElectron.,2018,33(3):2559-2570.] utilizes the converter output current error to locate faults based on the characteristics of the residual change rate. This method requires injecting specific control signals into the system to reduce the impact on the quality of the converter output current. Voltage-based methods obtain transient fault information through additional hardware circuits or voltage observers. The literature [C.Yang et al.,Voltagedifference residual-based open-circuit fault diagnosis approach for three-level converters in electric traction systems[J].IEEE Trans.Power Electron.,2020,35(3):3012-3028.] proposes a DC bus voltage model and uses residual evaluation functions and current similarity functions for fault diagnosis.The literature [Wang, Y. Tang, and C.-J. Zhang. Open circuit fault diagnosis and tolerance strategy applied to four-wire T-type converter systems[J]. IEEE Trans. Power Electron., 2019, 34(6): 5764–5778.] decomposes the model into positive, negative, and zero-sequence voltage models, and determines the fault location based on different voltage models. The similar fault characteristics caused by different power devices are a challenge in three-level converter fault diagnosis. To address this issue, the literature [J. Chen, C. Zhang, X. Xing, et al. An open-circuit fault diagnosis method for T-type three-level rectifiers[J]. Proc. IEEE Energy Convers. Congr. Expo., 2019: 5502–5506.] locates the faulty device based on the change trend of the DC-side capacitor midpoint voltage. However, using midpoint voltage balance control can affect the midpoint voltage change trend, leading to misdiagnosis. References [Z.Li,B.Zhao,X.Zhang,etal.An IGBT open-circuit fault diagnosis method for grid-tied T-type three-level inverters[J].Proc.IEEE Energy Convers.Congr.Expo.,2020:5324–5327.] and [Z.Li,H.Ma,Z.Bai,etal.Fast transistor open-circuit faultsdiagnosis in grid-tied three-phase VSIs based on average bridge arm pole-to-pole voltages and error-adaptive thresholds[J].IEEE Trans.Power Electron.,2018,33(9):8040–8051.] propose a fault diagnosis method based on a voltage model. This method is related to the modulation strategy and is only applicable to pulse width modulation (PWM) strategies.

[0006] The methods for fault-tolerant control of converters are mainly divided into two categories: hardware fault tolerance and software fault tolerance. The literature [KATEBIR, HE Jiangbiao, WEISE N. An advanced three-level active neutral-point-clamped inverter with improved fault-tolerant capabilities[J].IEEE Transactions on Power Electronics, 2018, 33(8): 6897-6909.], [Wang Yufeng, Zhang Ying, Li Shuang, et al. Research on 3N+1 redundancy fault-tolerant strategy for multi-level cascaded H-bridge inverters[J]. High Voltage Apparatus, 2018, 54(02): 208-212.], [WANG Borong, LIZhan, BAI Zhihong, et al. A redundant unit to form T-type three-level inverters tolerant of IGBT open-circuit faults in multiple legs[J].IEEE Transactions on Power Electronics, 2020, 35(1): 924-939.] adopts a hardware fault-tolerant method, using a fourth redundant bridge arm to replace the faulty bridge arm for fault-tolerant control. This method improves the reliability of the converter, but increases the cost. The literature [AZERP,OUNI S,NARIMANI MA novel fault-tolerant technique for active-neutral-point-clamped inverter using carrier-based PWM[J].IEEE Transactions on Industrial Electronics,2020,67(3):1792-1803.] proposes a fault-tolerant control strategy for ANPC converter based on carrier modulation algorithm, which can realize the fault tolerance function of single or multiple devices. However, this method does not consider the capacitor midpoint voltage, and the DC side midpoint potential is unstable after the fault. Summary of the Invention

[0007] To address the shortcomings of the aforementioned background technology, this invention proposes a method for diagnosing open-circuit faults in a T-type three-level grid-connected converter, thus solving the technical problem of low reliability in the operation of the grid-connected converter.

[0008] The technical solution of this invention is implemented as follows:

[0009] A method for diagnosing open-circuit faults in a T-type three-level grid-connected converter, comprising the following steps:

[0010] Step 1: Analyze the operation of the T-type three-level grid-connected converter after an open-circuit fault occurs, and calculate the current flowing from the midpoint of the DC-side capacitor to each phase based on the switch status;

[0011] Step 2: Calculate the average value of the current flowing from the midpoint to each phase and the average value of the converter output current. Compare the average value of the current flowing from the midpoint to each phase and the average value of the converter output current with the set thresholds to locate the fault location.

[0012] Step 3: Based on the location of the fault, classify it into Class I faults and Class II faults;

[0013] Step 4: Establish fault-tolerant structures for Type I and Type II faults respectively, and implement fault-tolerant control for the two types of faults based on model predictive control strategies.

[0014] The open-circuit faults of the T-type three-level grid-connected converter include open-circuit faults in four power devices of phase a, phase b, and phase c.

[0015] The method for calculating the current flowing from the midpoint of the DC-side capacitor to each phase based on the switching state is as follows:

[0016] According to Kirchhoff's current law:

[0017] i o =i ao +i bo +i co (1);

[0018] Among them, i o i represents the sum of the currents flowing from the midpoint of the DC-side capacitor to each phase. ao i is the current flowing from the midpoint of the DC-side capacitor to phase a. bo i is the current flowing from the midpoint of the DC-side capacitor to phase b. co This is the current flowing from the midpoint of the DC-side capacitor to phase c;

[0019] The converter switching state is defined as follows:

[0020]

[0021] Where x = a, b, c; P represents the state S x1 and S x2 On, S x3 and S x4 Closed; O state indicates S x2 and S x3 On, S x1 and Sx4 Closed; N state indicates S x3 and S x4 On, S x1 and S x2 closure;

[0022] According to i xo The relationship between the switch states is derived as follows:

[0023]

[0024] Among them, i a i b and i c Equalizer output current;

[0025] Based on equations (1), (2), and (3), calculate i xo :

[0026]

[0027] The method for locating the fault location is as follows:

[0028] Calculate i xo average current i xo_ave and i xo_ave With the set diagnostic threshold I thr1 Comparison; when i xo_ave Less than -I thr1 When the fault is located in the upper half of the bridge arm; when i xo_ave Greater than I thr1 At that time, the fault was located in the lower half of the bridge arm;

[0029] When the fault is located in the upper half-bridge arm, calculate i a The average value i x_ave and i x_ave With diagnostic threshold I thr2 Comparison; when i x_ave Less than -I thr2 At that time, power device S x1 An open circuit fault occurs; conversely, the power device S... x2 An open circuit fault occurred;

[0030] When the fault is located in the lower half-bridge arm, calculate i a The average value i x_ave and i x_ave With diagnostic threshold I thr2 Comparison; when i x_ave Greater than I thr2 At that time, power device S x4 An open circuit fault occurs; conversely, the power device S... x3 An open circuit fault has occurred.

[0031] The Class I fault refers to the power device S x1 and S x4 A fault occurs; Class II faults refer to power devices S x2 and S x3 A malfunction has occurred.

[0032] The model predictive control strategy is as follows:

[0033] The mathematical model of the T-type three-level grid-connected converter is:

[0034]

[0035] Among them, u αβ =[u α ,u β ] T u α u represents the α-axis component of the converter output voltage. β Let L be the output voltage component of the converter on the β axis, L be the filter inductance, R be the parasitic resistance of the filter inductance, and i be the voltage across the input voltage. αβ =[i α i β ] T i α i represents the component of the converter output current along the α-axis. β e represents the component of the converter output current on the β axis. αβ =[e α ,e β ] T e α Let i be the component of the grid-side voltage along the α-axis. β This represents the component of the grid-side voltage along the β axis.

[0036] Equation (5) is discretized using Euler methods, resulting in:

[0037]

[0038] Where T is the control period and k is the time.

[0039] Simplifying equation (6), we obtain the predicted current at time k+1:

[0040]

[0041] Substituting the available voltage vectors for different faults into equation (7) yields the predicted current;

[0042] A cost function is constructed based on the predicted current and the predicted DC-side capacitor voltage difference. The voltage vector that minimizes the cost function is selected as the optimal voltage vector and applied to the next time step.

[0043] The converter voltage component after a Type I fault is:

[0044]

[0045] The converter voltage components after a Type II fault are:

[0046]

[0047] The expression for the cost function is:

[0048] g = |i αβref (k+1)-i αβ (k+1)|+λ|△u C (k+1)|(15);

[0049] Among them, i αβref Let λ be the component of the reference current along the αβ axis, and Δu be the weighting coefficient. C To predict the voltage difference between capacitors.

[0050] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0051] 1) After a fault occurs in a T-type three-level grid-connected converter, the current flowing from the midpoint of the DC-side capacitor to the faulty phase and the converter output current are lost. This is determined by calculating the average value of the current flowing from the midpoint of the DC-side capacitor to each phase and comparing it with a threshold I. thr1 After comparison, the faulty bridge arm was identified.

[0052] 2) If the faulty power device is located in the longitudinal bridge arm, the output current of the faulty phase is missing; if the faulty power device is located in the transverse bridge arm, the output current error of the faulty phase increases. The average output current of the converter is calculated and compared with the threshold I. thr2 By comparison, the location of the faulty device is determined, and open-circuit fault detection of the converter is achieved.

[0053] 3) Based on the fault type and fault-tolerant structure, a model prediction fault-tolerant control strategy is proposed, which can achieve fault-tolerant continuous operation without adding extra hardware redundancy and improve the reliability of converter operation. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a structural diagram of the T-type three-level grid-connected converter of the present invention.

[0056] Figure 2This is a voltage vector diagram of the T-type three-level grid-connected converter of the present invention.

[0057] Figure 3 This is a current path diagram after an open-circuit fault; where (a) is the P state, S a1 Fault, i a >0(b)O state, S a2 Fault, i a >0(c)O state, S a3 Fault, i a <0(d)N state, S a4 Fault, i a <0.

[0058] Figure 4 For the converter output current i a and the current i flowing from the midpoint to phase a ao Waveform (a)S a1 Fault (b)S a2 Fault.

[0059] Figure 5 This is a flowchart of the fault diagnosis method of the present invention.

[0060] Figure 6 This is a fault-tolerant structure diagram of the Type I fault T-type three-level converter of the present invention.

[0061] Figure 7 This is a voltage vector distribution diagram for a-phase bridge arm Class I fault according to the present invention.

[0062] Figure 8 This is a fault-tolerant structure diagram of the Type II fault T-type three-level converter of the present invention.

[0063] Figure 9 This is a voltage vector distribution diagram of the a-phase bridge arm under a type II fault according to the present invention.

[0064] Figure 10 This is a hardware-in-the-loop experimental platform for fault tolerance of T-type three-level grid-connected converters.

[0065] Figure 11 The waveform represents the current flowing from the midpoint O to phase a.

[0066] Figure 12 Converter output current and fault diagnosis results when phase a is open-circuited (a)S a1 Fault (b)S a2 Fault (c)S a3 Fault (d)S a4 Fault.

[0067] Figure 13 When it is a type I fault in phase a (S) a1 (Fault) Converter output current and midpoint voltage.

[0068] Figure 14 When it is a type II fault in phase a (S) a2 (Fault) Converter output current and midpoint voltage. Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] This invention provides a method for diagnosing open-circuit faults in a T-type three-level grid-connected converter, the specific steps of which are as follows:

[0071] Step 1: Analyze the operation of the T-type three-level grid-connected converter after an open-circuit fault occurs, and calculate the current flowing from the midpoint of the DC-side capacitor to each phase based on the switch status.

[0072] The structure of a T-type three-level converter is as follows: Figure 1 As shown. U dc For DC side voltage, i a i b and i c i is the output current of the converter. ao i bo and i co This represents the current flowing from the midpoint of the DC-side capacitor to each phase. L is the filter inductance, and R is the parasitic resistance of the filter inductance. a e b and e c This represents the grid-side voltage. F k1 and F k2 For fast-acting fuses. Each phase arm of the T-type three-level grid-connected converter consists of four power devices, namely S... xj (Where x = a, b, c, j = 1, 2, 3, 4). Each phase arm has 3 switching states, namely P state (S... x1 and S x2 On, S x3 and S x4 (Off), O state (S) x2 and S x3 On, S x1 and S x4 (Off) and N state (S x3 and S x4 On, S x1 and S x2 (Closed). Therefore, the T-type three-level grid-connected converter has a total of 3 3=27 switching states. Each switching state corresponds to a voltage vector, such as Figure 2 As shown, long arrows represent large vectors, medium arrows represent medium vectors, and short arrows represent small vectors.

[0073] To determine the location of the open-circuit fault, the scenarios of open-circuit faults occurring at different locations were analyzed. Assuming an open-circuit fault occurs in phase a bridge arm, the converter output current is positive, and there are four possible fault scenarios: S... a1 S a2 S a3 and S a4 The specific details are as follows.

[0074] 1) Power device S a1 Open circuit fault

[0075] like Figure 3 As shown in (a), when the switch state is P and during normal operation, the forward current flows through S. a1 Flow filter. When S a1 During an open-circuit fault, the P state is affected. If the voltage u from the midpoint O of the converter capacitor to the neutral point n... on Greater than the output voltage u a (i.e. u OA >0), the current will pass through S a2 and D a3 Freewheeling continues until the forward current drops to 0, D a3 When subjected to reverse voltage, i is turned off. ao There is no path for the positive current; if u OA <0, the faulty phase outputs negative current that varies with the reference value, and positive current is missing. S a1 An open-circuit fault will not affect the current flow path in the O and N states, and the P and N states will not be affected.

[0076] 2) Power devices S a2 Open circuit fault

[0077] like Figure 3 As shown in (b), when the switch state is O and during normal operation, the forward current flows through S. a2 and D a3 Flow filter. If S a2 In an open-circuit fault, the positive current flow path in state O is affected, while the negative current flow path in state O is unaffected. P and N states can be achieved. The effect of the T-type three-level converter is similar to that of a two-level converter, but the current error increases.

[0078] 3) Power devices S a3 Open circuit fault

[0079] like Figure 3As shown in (c), when the switch state is O and during normal operation, the negative current flows through S. a3 and D a2 Flow filter. If S a3 In an open-circuit fault, the negative current flow path in state O is affected, while the positive current flow path in state O is unaffected. Both state P and state N can be achieved. The effect of the T-type three-level converter is similar to that of a two-level converter, but the current error increases.

[0080] 4) Power devices S a4 Open circuit fault

[0081] like Figure 3 As shown in (d), when the switch state is N and during normal operation, a negative current flows through S. a4 If S a4 An open-circuit fault affects the N state. If the voltage u from the midpoint O of the converter capacitor to the neutral point n... on Less than the output voltage u a (i.e. u OA <0), current flows through S a3 and D a2 Freewheeling continues until the negative current is 0, D a1 Reverse voltage shutdown, i ao The negative current has no path; if u OA >0, the positive current output of the faulty phase varies with the reference value, but the negative current is missing. S a4 An open-circuit fault will not affect the current flow path in the O and P states; the P and N states are unaffected.

[0082] Assuming the open-circuit fault occurs in the other two phases, their operating conditions are the same as when the open-circuit fault occurs in phase a, and will not be analyzed in detail.

[0083] Based on the above analysis of converter operation under power device failure, it can be seen that power device S x1 and S x4 The fault operation is similar, S x2 and S x3 The fault operation is similar. To accurately locate the fault, each phase bridge arm is divided into an upper half and a lower half. The upper half includes S... x1 and S x2 Two power devices, the lower half-bridge arm includes S x3 and S x4 Two power devices (x = a, b, c).

[0084] Taking phase a as an example, assume the fault location is in the upper half of phase a. If power device S... a1 When an open-circuit fault occurs, the converter output phase a current waveform is as follows: Figure 4As shown in (a), the P state is affected, and the forward current of the fault phase is lost. Due to u OA If the value is greater than 0, the current will pass through S. a2 and D a3 Freewheeling occurs when, after a fault, the converter output phase a current still briefly exists during the positive half-wave. This freewheeling current, i, is due to this freewheeling. a The current flows from the midpoint of the DC-side capacitor of the converter to the faulty phase, and the converter output current i a with i ao coincide.

[0085] If power device S a2 When an open-circuit fault occurs, the converter output phase a current waveform is as follows: Figure 4 As shown in (b), the forward current flow path in state O is affected, i ao The positive current is completely absent. S a2 An open-circuit fault will not affect the P and N states, and the forward current of phase a of the converter output will not be lost.

[0086] In summary, when a fault occurs in the upper half of phase a (S) bridge arm, a1 and S a2 ), i ao The positive current is missing in the non-faulty phase, and the current i in the non-faulty phase is... bo and i co It will not be missing. Similarly, if the open-circuit fault occurs in the lower half-arm of phase a (S... a3 and S a4 ), i ao The negative current is missing, and the current i in the non-faulty phase is... bo and i co It will not be missing.

[0087] Compare the current i flowing from the midpoint of the DC-side capacitor to each phase. ao i bo and i co The faulty phase can then be identified through the following steps:

[0088] According to Kirchhoff's current law:

[0089] i o =i ao +i bo +i co (1);

[0090] Among them, i o i represents the sum of the currents flowing from the midpoint of the DC-side capacitor to each phase. ao i is the current flowing from the midpoint of the DC-side capacitor to phase a. bo i is the current flowing from the midpoint of the DC-side capacitor to phase b. co This is the current flowing from the midpoint of the DC-side capacitor to phase c;

[0091] The converter switching state is defined as follows:

[0092]

[0093] Where x = a, b, c; P represents the state S x1 and S x2 On, S x3 and S x4 Closed; O state indicates S x2 and S x3 On, S x1 and S x4 Closed; N state indicates S x3 and S x4 On, S x1 and S x2 closure;

[0094] According to i xo The relationship between the switch states is derived as follows:

[0095]

[0096] Among them, i a i b and i c Equalizer output current;

[0097] Based on equations (1) and (2), calculate i xo :

[0098]

[0099] Step 2: Calculate the average value of the current flowing from the midpoint to each phase and the average value of the converter output current. Compare the average value of the current flowing from the midpoint to each phase and the average value of the converter output current with the set thresholds to locate the fault location.

[0100] When the open-circuit fault is located in the upper half-arm, the positive current flowing from the midpoint of the DC-side capacitor to the faulty phase is lost. When the open-circuit fault is located in the lower half-arm, the negative current flowing from the midpoint of the DC-side capacitor to the faulty phase is lost. This is determined by calculating i... xo average current i xo_ave and i xo_ave With the set diagnostic threshold I thr1 Comparison; when i xo_ave Less than -I thr1 When the fault is located in the upper half of the bridge arm; when i xo_ave Greater than I thr1 At that time, the fault was located in the lower half of the bridge arm.

[0101] When the upper half-bridge power device S a1When an open-circuit fault occurs, the forward current of the faulty phase of the converter is lost, such as... Figure 4 As shown in (a). If the upper half-bridge power device S... a2 When a fault occurs, the output current error of the faulty phase increases, such as... Figure 4 As shown in (b). Calculate i a The average value i x_ave and i x_ave With diagnostic threshold I thr2 Comparison; the fault is located in the upper half-bridge arm, when i x_ave Less than -I thr2 At that time, power device S x1 An open circuit fault occurs; conversely, the power device S... x2 An open circuit fault occurred; the fault was located in the lower half-bridge arm, when i x_ave Greater than I thr2 At that time, power device S x4 An open circuit fault occurs; conversely, the power device S... x3 An open circuit fault has occurred.

[0102] In summary, i x and i xo By comparing the average value with the diagnostic threshold, accurate location of the faulty device can be achieved. The flowchart of the proposed fault diagnosis method is shown below. Figure 5 As shown, S nj This indicates the specific fault switch device.

[0103] Step 3: Based on the fault location, faults are classified into Class I and Class II. Analysis of the operation of the power device in phase a after an open-circuit fault reveals that the power device S... a1 and S a4 The impact of an open-circuit fault on the converter is similar, affecting power devices S. a2 and S a3 The impact of an open-circuit fault on the converter is similar. Based on the fault location, fault types are divided into two categories: Class I faults (power device S...) x1 and S x4 Type I faults and Type II faults (power device S) x2 and S x3 A fault occurs. Fault tolerance control is implemented for both types of faults.

[0104] The mathematical model of the T-type three-level grid-connected converter is:

[0105]

[0106] Among them, u αβ =[u α ,u β ] T u αu represents the α-axis component of the converter output voltage. β Let L be the output voltage component of the converter on the β axis, L be the filter inductance, R be the parasitic resistance of the filter inductance, and i be the voltage across the input voltage. αβ =[i α i β ] T i α i represents the component of the converter output current along the α-axis. β e represents the component of the converter output current on the β axis. αβ =[e α ,e β ] T e α Let i be the component of the grid-side voltage along the α-axis. β This represents the component of the grid-side voltage along the β axis.

[0107] Equation (5) is discretized using Euler methods, resulting in:

[0108]

[0109] Where T is the control period and k is the time.

[0110] Simplifying equation (6), we obtain the predicted current at time k+1:

[0111]

[0112] Step 4: Establish fault-tolerant structures for Type I and Type II faults respectively, and implement fault-tolerant control for the two types of faults based on model predictive control strategies.

[0113] When a Type I fault occurs in the T-type converter, the faulty phase passes through the power device S. x2 and S x3 It is directly connected to the midpoint of the DC-side capacitor. After a Type I fault occurs in the converter, the faulty phase has only a 0 state, while the non-faulty phases each have three states, for a total of 9 switching states. The voltage component u after the fault... αβ As shown in equation (8).

[0114]

[0115] Taking phase a as an example, the fault-tolerant structure after a Type I fault is as follows: Figure 6 As shown. After a fault, the voltage vector can be used as follows: Figure 7 As shown, the voltage vectors are shown in Table 1.

[0116] Table 1a shows the fault-tolerant structure voltage vector under Class I faults in phase bridge arm.

[0117]

[0118] When a Type II fault occurs in the converter, the faulty phase's O state is affected, leaving only the P and N states for the faulty phase. The non-faulty phases each have three states, resulting in a total of 18 switching states. The voltage component u after the fault... αβ As shown in equation (9).

[0119]

[0120] Taking phase a as an example, after a Type II fault occurs, the power device S a2 and S a3 The voltage component u remains in the off state after a fault. αβ As shown in equation (9), when phase a is faulty, S a Take 1 or -1, S b and S c Take 1, 0, or -1.

[0121] When a phase bridge arm experiences a Class II fault, the available voltage vectors are shown in Table 2, and their distribution is as follows: Figure 9 As shown.

[0122] Table 2a shows the fault-tolerant structure voltage vector during Class II faults in phase bridge arms.

[0123]

[0124]

[0125] Based on the predictive control theory of the finite control set model, the available voltage vectors after different faults are substituted into (7) to obtain the predicted current, the cost function is evaluated, and the voltage vector that minimizes it is selected as the optimal voltage vector and applied to the next moment.

[0126] After a fault, the DC-side current flows directly into the faulty phase through the midpoint of the capacitor, exacerbating the problem of DC-side midpoint voltage imbalance. This invention predicts the voltage difference between the upper and lower capacitors on the DC side by controlling the switching state and incorporates the prediction result into the cost function to achieve the goal of balancing the midpoint voltage.

[0127] The expression for the DC-side capacitor voltage is:

[0128]

[0129] Among them, I C1 and I C2 These are the currents flowing through the DC-side capacitors C1 and C2, respectively.

[0130] Discretizing (10) yields the capacitor voltage at time k+1:

[0131]

[0132] The two capacitors on the DC side have the same value, and the voltage difference between them is:

[0133]

[0134] Where C represents the values ​​of capacitors C1 and C2.

[0135] From (1) and (3), we can conclude that:

[0136] i o =(1-|S a |)×i a +(1-|S b |)×i b +(1-|S c |)×i c (13)

[0137] Substituting (13) into (12), we obtain the predicted capacitor voltage difference:

[0138]

[0139] Substitute the predicted current and the predicted DC-side capacitor voltage difference into the cost function (15).

[0140] g = |i αβref (k+1)-i αβ (k+1)|+λ△u C (k+1)| (15)

[0141] Among them, i abref =[i aref i bref ] T i aref and i bref These represent the components of the reference current along the ab axis. λ is the weighting coefficient.

[0142] Experimental Analysis

[0143] To verify the effectiveness of the proposed fault diagnosis and fault-tolerant control strategy for the T-type three-level grid-connected converter, a system was built as follows: Figure 10 The StarSim hardware-in-the-loop experimental platform shown has its control circuit implemented by the Rapid Control Prototyping (RCP) prototype, and its hardware circuitry implemented by the Hardware-in-the-Loop (HIL) test platform. The experimental parameters are shown in Table 3.

[0144] Table 3. Fault-Tolerant Experimental Parameters for Type T Three-Level Grid-Connected Converter

[0145]

[0146] Figure 11The current i flowing from the midpoint O of the converter capacitor to phase a is... ao The actual and calculated values ​​of the experimental results. When the converter is running normally, i ao The calculated value is consistent with the actual value. When the power device S a2 When an open-circuit fault occurs, the positive half-wave of the current flowing from the capacitor midpoint to phase a is missing, consistent with theoretical analysis. At this time, the calculated value i... ao Equally accurate. Experimental results verified the correctness of formula (4).

[0147] Figure 12 The output current of the T-type three-level converter when an open-circuit fault occurs in phase a bridge arm and the fault diagnosis results are presented. Figure 12 (a) indicates that when the power device S a1 After an open circuit fault occurs, due to u OA If the value is greater than 0, the forward current of phase a flows through S. a2 and D a3 Freewheeling continues until the current drops to D. a3 When subjected to reverse voltage, the faulty phase is shut off, and its forward current is zero. After the fault occurs, the flag changes from 0 to 1, indicating that S... a1 A malfunction has occurred.

[0148] Figure 12 (b) indicates the power device S a2 Converter output current and fault diagnosis results after an open-circuit fault occurs. When power device S... a2 After an open-circuit fault occurs, the fault phase current ripple increases. After the fault occurs, the flag changes from 0 to 2, indicating that S... a2 A malfunction has occurred.

[0149] Figure 12 (c) and Figure 12 (d) represents the power device S a3 and S a4 The converter output current waveform and fault diagnosis results after an open-circuit fault occur. Power device S a3 After an open-circuit fault occurs, the negative current ripple of phase a increases, and flag jumps from 0 to 3, indicating that S a3 A malfunction has occurred. Power device S a4 After an open-circuit fault occurs, the negative current output by the faulty phase drops to 0, and the flag changes from 0 to 4, indicating that S a4 A malfunction has occurred.

[0150] Diagnostic threshold I thr1 =0.35A, I thr2 =2A. The experimental results of fault diagnosis are consistent with the theoretical analysis, and the proposed method can quickly locate the fault location.

[0151] Figure 13This represents the converter output current and the voltages of the upper and lower capacitors after a Type I fault occurs in phase a. a1 After the fault, the forward current of phase a is lost, and the current of the non-faulty phases is distorted. After adopting the proposed fault-tolerant control strategy, the three-phase current is close to a sine wave, and the THD is 4.02%. After a Class I fault, without the midpoint voltage balancing strategy, the voltage difference between the upper and lower capacitors is about 400V. After the midpoint voltage balancing strategy, the voltages of the upper and lower capacitors are basically consistent.

[0152] Figure 14 The waveform represents the converter output current under a Type II fault in phase a, with a waveform distortion rate of 4.12%. The T-type three-level grid-connected converter has a certain degree of self-fault tolerance for Type II faults, and the converter output current waveform remains intact after the fault. After a Type II fault occurs, without the midpoint voltage balancing strategy, the voltage difference between the upper and lower capacitors is approximately 500V; after midpoint voltage balancing, the voltages of the two capacitors remain essentially consistent.

[0153] This invention first analyzes the operation of a T-type three-level grid-connected converter after an open-circuit fault, calculating the current flowing from the midpoint of the DC-side capacitor to each phase based on the switching state. The average output current and the average current flowing from the midpoint to each phase are calculated and compared with set thresholds to locate the fault. Based on the fault location, faults are classified into Type I faults (vertical arm faults) and Type II faults (horizontal arm faults). Fault-tolerant structures are established for Type I and Type II faults respectively, and fault-tolerant control is implemented for each type of fault based on a model predictive control (MPC) strategy. Finally, an experimental platform for a T-type three-level grid-connected converter is built to verify the effectiveness of the proposed fault diagnosis and fault-tolerant control strategy.

[0154] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for diagnosing open-circuit faults in a T-type three-level grid-connected converter, characterized in that, The steps are as follows: Step 1: Analyze the operation of the T-type three-level grid-connected converter after an open-circuit fault occurs, and calculate the current flowing from the midpoint of the DC-side capacitor to each phase based on the switch status; Step 2: Calculate the average value of the current flowing from the midpoint to each phase and the average value of the converter output current. Compare the average value of the current flowing from the midpoint to each phase and the average value of the converter output current with the set thresholds to locate the fault location. Step 3: Based on the location of the fault, classify it into Class I faults and Class II faults; Step 4: Establish fault-tolerant structures for Type I and Type II faults respectively, and implement fault-tolerant control for the two types of faults based on model predictive control strategies. The method for calculating the current flowing from the midpoint of the DC-side capacitor to each phase based on the switching state is as follows: According to Kirchhoff's current law: （1）； Among them, i o i represents the sum of the currents flowing from the midpoint of the DC-side capacitor to each phase. ao i is the current flowing from the midpoint of the DC-side capacitor to phase a. bo i is the current flowing from the midpoint of the DC-side capacitor to phase b. co This is the current flowing from the midpoint of the DC-side capacitor to phase c; The converter switching state is defined as follows: （2）； Where x = a, b, c; P represents the state S x1 and S x2 On, S x3 and S x4 Closed; O indicates S x2 and S x3 On, S x1 and S x4 Closed; N state indicates S x3 and S x4 On, S x1 and S x2 closure; According to i xo The relationship between the switch states is derived as follows: （3）； Among them, i a i b and i c Equalizer output current; Based on equations (1), (2), and (3), calculate i xo : （4）； The method for locating the fault location is as follows: Calculate i xo average current i xo_ave and i xo_ave With the set diagnostic threshold I thr1 Comparison; when i xo_ave Less than - I thr1 When the fault is located in the upper half of the bridge arm; when i xo_ave Greater than I thr1 At that time, the fault was located in the lower half of the bridge arm; When the fault is located in the upper half-bridge arm, calculate i a The average value i x_ave and i x_ave With diagnostic threshold I thr2 Comparison; when i x_ave Less than - I thr2 At that time, power device S x1 An open circuit fault occurs; conversely, the power device S... x2 An open circuit fault occurred; When the fault is located in the lower half-bridge arm, calculate i a The average value i x_ave and i x_ave With diagnostic threshold I thr2 Comparison; when i x_ave Greater than I thr2 At that time, power device S x4 An open circuit fault occurs; conversely, the power device S... x3 An open circuit fault has occurred.

2. The method for diagnosing open-circuit faults in a T-type three-level grid-connected converter according to claim 1, characterized in that, The open-circuit faults of the T-type three-level grid-connected converter include open-circuit faults in four power devices of phase a, phase b, and phase c.

3. The method for diagnosing open-circuit faults in a T-type three-level grid-connected converter according to claim 1 or 2, characterized in that, The Class I fault refers to the power device S x1 and S x4 A fault occurs; Class II faults refer to power devices S x2 and S x3 A malfunction has occurred.

4. The method for diagnosing open-circuit faults in a T-type three-level grid-connected converter according to claim 3, characterized in that, The model predictive control strategy is as follows: The mathematical model of the T-type three-level grid-connected converter is: （5）； Among them, u αβ =[u α , u β ] T u α u represents the α-axis component of the converter output voltage. β Let L be the output voltage of the converter on the β-axis, L be the filter inductance, and R be the parasitic resistance of the filter inductance. =[i α i β ] T i α i represents the component of the converter output current along the α-axis. β This represents the component of the converter output current along the β axis. =[e α , e β ] T e α Let i be the component of the grid-side voltage along the α-axis. β This represents the component of the grid-side voltage along the β axis. Equation (5) is discretized using Euler methods, resulting in: （6）； Where T is the control period and k is the time. Simplifying equation (6), we obtain the predicted current at time k+1: （7）； Substituting the available voltage vectors for different faults into equation (7) yields the predicted current; A cost function is constructed based on the predicted current and the predicted DC-side capacitor voltage difference. The voltage vector that minimizes the cost function is selected as the optimal voltage vector and applied to the next time step.

5. The method for diagnosing open-circuit faults in a T-type three-level grid-connected converter according to claim 4, characterized in that, The converter voltage components after a Type I fault are: （8）； The converter voltage components after a Type II fault are: （9）。 6. The method for diagnosing open-circuit faults in a T-type three-level grid-connected converter according to claim 5, characterized in that, The expression for the cost function is: （15）； in, The reference current component along the αβ axis, These are the weighting coefficients. To predict the voltage difference between capacitors.

Images