Method for quickly diagnosing and positioning open-circuit fault of IGBT (Insulated Gate Bipolar Translator) of submodule of converter

Abstract

According to the converter submodule IGBT open-circuit fault rapid diagnosis and positioning method, each phase of a three-phase converter is composed of an upper bridge arm and a lower bridge arm, the three phases are provided with six bridge arms, and each bridge arm is composed of a plurality of submodules connected in series and an inductor connected in series; the method comprises the steps of fault diagnosis and fault positioning. According to the fault diagnosis, whether a sub-module open-circuit fault occurs or not is judged by optimally estimating three-phase circulating current and output current through an adaptive extended Kalman filtering algorithm, and the fault position of a bridge arm in a fault phase in the three phases is judged; according to fault positioning, sub-modules in a fault bridge arm are grouped and reconstructed through the fault bridge arm corresponding to the fault position of the bridge arm, a reconstructed sub-unit is formed, then the voltage of the reconstructed sub-unit is optimally estimated through an adaptive extended Kalman filtering algorithm, and the fault sub-module is positioned according to the residual error of the voltage of the reconstructed sub-unit. The problems that the number of state observers is large and the diagnosis time is long when the number of sub-modules is large are solved.

Description

Technical Field

[0001] The invention belongs to the technical field of power electronics, and in particular relates to a method for quickly diagnosing and locating an open-circuit fault of an IGBT of a converter submodule. Background Art

[0002] Since the Modular Multilevel Converter (MMC) topology was proposed by German scholar Rainer Marquardt in 2001, it has simplified the topology structure because each bridge arm connects multiple groups of identical power units or submodules in series, does not require a phase-shifting transformer, and has many advantages compared to traditional multilevel converters. MMC has good scalability, simple submodule structure, rich application scenarios, high power density, and other advantages, and its technical and economic advantages are obvious. MMC has become the most promising topology in medium and high power applications such as high-voltage flexible direct current transmission, medium-voltage variable speed drives, power quality improvement, and medium- and high-voltage energy storage.

[0003] Since the MMC circuit topology is composed of a large number of cascaded submodules, a large number of switching power devices will be used, and the power devices themselves are usually the most prone to failure in the application. In the submodule, the power devices are mostly insulated gate bipolar transistors (IGBTs). IGBT faults usually include short circuit and open circuit faults. IGBT short circuit faults will generate strong short circuit currents. Therefore, in actual projects, submodule overcurrent protection devices are generally equipped. Once a short circuit fault occurs, the system will quickly lock the faulty submodule within a few microseconds to avoid direct faults. Therefore, short circuit faults are generally detected and protected by hardware circuits. However, the impact of a single or a few IGBT open circuit faults on the operation of the three-phase converter will not appear quickly and directly in most cases. Appropriate fault diagnosis methods are required to detect IGBT open circuit faults to avoid the situation where the submodule capacitor voltage is too high and the output voltage and current waveforms are seriously distorted, which will eventually cause the system to crash and stop operating.

[0004] The existing detection algorithms for submodule open circuit faults are mainly divided into three categories: hardware detection-based, data deep learning-based, and system mathematical model-based. For the method based on hardware detection, additional hardware circuits need to be added for fault diagnosis. The principle is relatively simple and the diagnosis speed is fast, but this method increases the hardware cost, and the existence of the new hardware itself will also constitute a potential fault point. For the detection method based on data deep learning, although this method has a fast detection speed, it requires a large number of training samples, has limited accuracy, and has extremely high requirements on the computing performance of the main control MCU. Taking machine deep learning as an example, when using this method for submodule fault diagnosis, it is necessary to extract and analyze the characteristic parameters of the data characteristics of normal operation and after the fault occurs, and compare them with the existing characteristic parameters to determine whether the fault occurs. However, this method requires a large number of fault samples to complete the data feature setting before diagnosis, and the failure of the IGBT module is an abnormal state, and there are few relevant sample data, which causes many difficulties in practical applications. In addition, the real-time calculation of a large amount of data requires extremely high MCU performance, which reduces the actual engineering application value. For the method based on system mathematical model, the algorithm is simple and easy to implement, but the accuracy of modeling has a great impact on its robustness. If the modeling accuracy is poor, there may be a model mismatch problem. Summary of the invention

[0005] The present invention provides a method for quickly diagnosing and locating an open-circuit fault of an IGBT of a converter submodule, which is used to solve certain disadvantages of the background technology.

[0006] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a method for rapid diagnosis and location of an open-circuit fault of an IGBT in a converter submodule, comprising: a three-phase converter connected to a power grid, each phase of the three-phase converter consisting of two upper and lower bridge arms, the three phases having a total of six bridge arms, each bridge arm consisting of a plurality of submodules connected in series and an inductor connected in series;

[0007] The method comprises: fault diagnosis and fault location;

[0008] The fault diagnosis is to determine whether a submodule open circuit fault occurs by optimally estimating the three-phase circulating current and output current through an adaptive extended Kalman filter algorithm, and to identify the fault position of the bridge arm in the faulty phase among the three phases;

[0009] The fault location is to group and reconstruct the sub-modules in the faulty bridge arm according to the faulty bridge arm corresponding to the faulty position of the bridge arm to form a reconstructed sub-unit, and then use an adaptive extended Kalman filter algorithm to optimally estimate the voltage of the reconstructed sub-unit, and locate the faulty sub-module based on the residual of the voltage of the reconstructed sub-unit.

[0010] Preferably, the submodule is a half-bridge submodule, and the half-bridge submodule includes: an upper tube and a lower tube connected in series, each of the upper tube and the lower tube is connected to a reverse-parallel diode, and the upper tube and the lower tube connected in series are connected to a parallel capacitor as a whole.

[0011] Preferably, the fault diagnosis specifically includes the following steps:

[0012] Step 1.1, let i be the i-th submodule in the bridge arm of the three-phase converter, initialize i=0, and the counter value t in the fault diagnosis=0; set the initial Kalman gain and reliability coefficient;

[0013] Initialize the noise variables of the converter three-phase circulating current process, the noise variables of the converter three-phase circulating current measurement, the noise variables of the converter three-phase output current process, the noise variables of the converter three-phase output current measurement, the noise variables in the sub-unit grouping reconstruction process, and the noise variables measured in the sub-unit grouping reconstruction process;

[0014] Step 1.2, collecting the three-phase bridge arm current, the DC side voltage of the converter, the three-phase upper bridge arm voltage, the three-phase lower bridge arm voltage, and the three-phase output current of the AC side of the converter;

[0015] Step 1.3, calculate the three-phase circulating current of the converter

[0016] Based on the three-phase bridge arm current, the DC side voltage of the converter, the three-phase upper bridge arm voltage, the three-phase lower bridge arm voltage, and the three-phase grid voltage, an equivalent circulating current control loop of the submodule and an equivalent AC control loop of the converter are obtained according to Kirchhoff's voltage law;

[0017] The equivalent circulation control loop expression of the submodule is as follows:

[0018]

[0019] In the above formula, L arm is the equivalent inductance of the bridge arm, v dc is the DC side voltage of the converter, i a_cir 、i b_cir 、i c_cir is the three-phase bridge arm circulation, u pa 、u pb 、u pc is the three-phase upper bridge arm voltage, u na 、u nb 、u nc is the three-phase lower bridge arm voltage, t is the counter value in fault diagnosis;

[0020] The equivalent AC control loop of the converter is expressed as follows:

[0021]

[0022] In the above formula, R arm is the equivalent resistance of the bridge arm, i sa 、i sb 、i sc is the three-phase output current of the AC side of the converter, u sa 、u sb 、u sc is the three-phase grid voltage;

[0023] Formula (4) is obtained by back-differentiation discretization of formula (2), and formula (5) is obtained by back-differentiation discretization of formula (3):

[0024]

[0025] In the above formula, f s is the system sampling frequency used in the discretization process; i a_cir (k), i b_cir (k), i c_cir (k) is the three-phase bridge arm circulation at time k, i a_cir (k-1), i b_cir (k-1), i c_cir (k-1) are the three-phase bridge arm circulation at time k-1; i sa (k), i sb (k), i sc (k) is the three-phase output current at time k, i sa (k-1), i sb (k-1), i sc (k-1) is the three-phase output current at time k-1; u pa (k-1),u pb (k-1),u pc (k-1) is the three-phase upper bridge arm voltage at time k-1, u na (k-1),u nb (k-1),u nc (k-1) is the three-phase lower bridge arm voltage at time k-1; u sa (k-1),u sb (k-1),u sc (k-1) are the three-phase grid voltages at time k-1;

[0026] Step 1.4, solving the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation according to formula (4); determining the optimal circulation amount of the three-phase bridge arm based on the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation;

[0027] Solving the adaptive optimal state observer of the extended Kalman filter of the three-phase output current according to equation (5); determining the optimal amount of the three-phase output current based on the adaptive optimal state observer of the extended Kalman filter of the three-phase output current;

[0028] Step 1.5, based on the optimal circulation amount of the three-phase bridge arm and the optimal amount of the three-phase output current, calculate the three-phase bridge arm circulating current residual and calculate the three-phase output current residual;

[0029] If the circulating current residual and the output current residual of a phase bridge arm in the three-phase are both greater than or equal to the preset threshold, then execute step 1.6; otherwise, return to execute step 1.2;

[0030] Step 1.6, clear the counter value in the fault diagnosis, and perform fault diagnosis based on the positive and negative relationship between the circulating current residual and the output current residual.

[0031] Preferably, the fault location specifically includes the following steps:

[0032] Step 2.1, based on the fault diagnosis result, obtain the fault code, and reconstruct the subunit according to the fault code; distinguish and mark the faults located in the upper bridge arm and the lower bridge arm; when the upper bridge arm fails, execute step 2.2, and when the lower bridge arm fails, execute step 2.4;

[0033] Step 2.2, collecting the voltage of the ith submodule of the three-phase upper bridge arm; based on the voltage of the ith submodule of the three-phase upper bridge arm, determining the voltage of the ith reconstruction subunit of the upper bridge arm; specifically expressed as:

[0034] u Subpi =u api +u bpi +u cpi (twenty three)

[0035] In the above formula, u Subpi is the voltage of the i-th reconstructed subunit in the upper bridge arm, u api 、u bpi 、u cpi are the voltages of the ith submodule of the three-phase upper bridge arm respectively;

[0036] Formula (23) can be expressed as:

[0037]

[0038] In the above formula, C sm is the submodule capacitance, S pai , S pbi , S pci are the switch states of the i-th submodule of the three-phase upper bridge arm, i pa 、i pa 、i pbare the three-phase upper bridge arm currents respectively, and t is the counter value in fault diagnosis;

[0039] Establishing a reconstructed subunit voltage optimal estimation observer according to formula (24), and determining the reconstructed subunit voltage optimal estimation value based on the reconstructed subunit voltage optimal estimation observer;

[0040] Step 2.3, based on the optimal estimated value of the reconstructed subunit voltage, calculate the upper bridge arm i-th reconstructed subunit voltage residual according to the following formula;

[0041] U Subpi_err =u Subpi_now -u Subpi (30)

[0042] In the above formula, U Subpi_err is the voltage residual of the i-th reconstructed subunit in the upper bridge arm, u Subpi_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the upper bridge arm, u Subpi is the actual sampling value of the voltage of the i-th reconstructed subunit of the upper bridge arm.

[0043] If the voltage residual of the i-th reconstructed subunit in the upper bridge arm is greater than or equal to the preset threshold of the voltage of the reconstructed subunit, the fault is located in the i-th reconstructed subunit in the upper bridge arm. Combined with the fault phase mark and the fault code, it is determined that the submodule open circuit fault is located in the i-th submodule in the upper bridge arm in one of the three phases, and the fault location is completed. Otherwise, execute step 2.2;

[0044] Step 2.4, calculate the voltage u of the lower bridge arm reconstruction subunit Subni , specifically expressed as:

[0045] u Subni =u ani +u bni +u cni (31)

[0046] In the above formula, u Subni is the voltage of the i-th reconstructed subunit in the lower bridge arm, u ani 、u bni 、u cni is the voltage of the ith submodule of the three-phase lower bridge arm;

[0047] Formula (31) can be expressed as:

[0048]

[0049] In the above formula, C sm is the submodule capacitance, S nai , S nbi , S nci are the switch states of the i-th submodule of the three-phase lower bridge arm, i na 、ina 、i nb They are the three-phase lower bridge arm currents respectively;

[0050] Establishing a second reconstructed subunit voltage optimal estimation observer according to formula (32), and determining a second reconstructed subunit voltage optimal estimation value based on the second reconstructed subunit voltage optimal estimation observer;

[0051] Step 2.5, based on the optimal estimated value of the second reconstructed subunit voltage, calculate the i-th reconstructed subunit voltage residual of the lower bridge arm according to the following formula;

[0052] U Subni_err =u Subni_now -u Subni (38)

[0053] In the above formula, U Subni_err is the voltage residual of the i-th reconstructed subunit in the lower bridge arm, u Subni_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the lower bridge arm, u Subni is the actual sampling value of the i-th reconstruction sub-unit voltage of the lower bridge arm;

[0054] If the voltage residual of the i-th reconstructed subunit in the lower bridge arm is greater than or equal to the preset threshold of the reconstructed subunit voltage, the fault is located to the i-th reconstructed subunit in the lower bridge arm. Combined with the fault phase mark and the fault code, it is determined that the submodule open circuit fault is located in the i-th submodule in the lower bridge arm in one of the three phases, and the fault location is completed. Otherwise, execute step 2.4.

[0055] Beneficial effects of the present invention:

[0056] Different from the fault diagnosis method in which the traditional state observer observes the current of each bridge arm and the capacitor voltage of all submodules, the present invention introduces the residual feedback compensation of the observed quantity to the Kalman gain on the basis of the extended Kalman filter algorithm, and obtains the adaptive Kalman gain of the observed quantity, which solves the problem of decreased estimation accuracy of the optimal state observer under load mutation and other conditions. The positive and negative relationship of the residuals of the three-phase circulating current and the output current is used to judge the position of the fault phase and the fault bridge arm, and then the fault location link based on the reconstructed subunit voltage is invested, and the three-phase submodules are grouped and reconstructed according to the position of the fault bridge arm, and the voltage of each reconstructed subunit is optimally estimated. The fault submodule is located according to the residual of the reconstructed subunit voltage, which avoids the problems of a large number of state observers, a large amount of data calculation, and a long diagnosis time when the number of submodules is large, and can overcome the problem that the open-circuit fault of the upper tube of the half-bridge submodule in the sorting voltage-sharing control algorithm is difficult to diagnose, and realize the rapid diagnosis and location of the open-circuit fault of a single IGBT. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A main circuit topology diagram of a three-phase converter to which the present invention is applicable;

[0058] Figure 2 A schematic diagram of the variables required for sampling in the present invention;

[0059] Figure 3 This is a schematic diagram of the reconstruction subunit proposed in the present invention;

[0060] Figure 4 This is a structural diagram of the upper bridge arm reconstruction subunit 1 proposed in the present invention;

[0061] Figure 5 The present invention is a flowchart of the steps of the method for diagnosing and locating an IGBT open-circuit fault of a half-bridge submodule of a three-phase converter system. DETAILED DESCRIPTION

[0062] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings. For the purpose of clear description, many physical details will be described together in the following description. However, it should be understood that these physical details should not be used to limit the present invention. That is, in some embodiments of the present invention, these physical details are not necessary. In addition, for the purpose of simplifying the illustration, some conventional structures and components will be depicted in a simple schematic manner in the illustration.

[0063] Example:

[0064] like Figure 1-5 As shown, a method for rapid diagnosis and positioning of an IGBT open-circuit fault of a converter submodule of the present invention comprises: a three-phase converter connected to a power grid, each phase of the three-phase converter is composed of two upper and lower bridge arms, and each bridge arm is composed of a plurality of submodules connected in series and an inductor connected in series;

[0065] The method comprises: fault diagnosis and fault location;

[0066] The fault diagnosis is to determine whether a submodule open circuit fault occurs by optimally estimating the three-phase circulating current and output current through an adaptive extended Kalman filter algorithm, and to identify the fault position of the bridge arm in the faulty phase among the three phases;

[0067] The fault location is to group and reconstruct the sub-modules in the faulty bridge arm according to the faulty bridge arm corresponding to the faulty position of the bridge arm to form a reconstructed sub-unit, and then use an adaptive extended Kalman filter algorithm to optimally estimate the voltage of the reconstructed sub-unit, and locate the faulty sub-module based on the residual of the voltage of the reconstructed sub-unit.

[0068] Each phase of the three-phase converter consists of two upper and lower bridge arms, with a total of six bridge arms for the three phases. Each bridge arm has N MMC (modular multilevel converter, MMC) submodules SM jri(r=p,n;j=a,b,c;i=1,2,3,…,N) and a bridge arm inductor L rj Series composition;

[0069] The submodule is a half-bridge submodule, and the half-bridge submodule includes: an upper tube VT1 and a lower tube VT2 connected in series, each of the upper tube VT1 and the lower tube VT2 is connected to a reverse parallel diode, and the upper tube VT1 and the lower tube VT2 connected in series are connected to a parallel capacitor as a whole.

[0070] exist Figure 1 In each MMC submodule SM jri It is a half-bridge module, which consists of the upper tube VT1 and the lower tube VT2 in series, and each of them is anti-parallel connected with a power diode VD1 and VD2, and then a capacitor C is connected in parallel. sm The half-bridge structure is composed of R, which is the DC side resistance load, C1 and C2 are the DC side support capacitors, and v dc is the DC side voltage. j_cir is the circulation of phase j, u sj is the AC grid voltage of phase j (j = a, b or c), i sj is the grid-side output current of phase j.

[0071] The driving pulse signals of the upper tube VT1 and the lower tube VT2 in the same submodule are defined to be complementary. In the three-phase conversion circuit, the MMC half-bridge submodule has three working states: on, off and locked. When the upper tube VT1 is turned on and the lower tube VT2 is turned off, it indicates that the submodule is in the on state. At this time, the output voltage of the submodule is the voltage of the submodule capacitor, and the capacitor is in the charging and discharging state; when the upper tube VT1 is turned off and the lower tube VT2 is turned on, it indicates that the submodule is in the off state. At this time, the output voltage of the submodule is 0, and the submodule capacitor is in the bypass state; when the upper tube VT1 is turned off and the lower tube VT2 is also turned off, it indicates that the submodule is in the locked state, which will only appear under abnormal operating conditions.

[0072] Figure 2 It is a schematic diagram of the variables that need to be sampled in the present invention. It is necessary to collect the j-phase r bridge arm current i of the three-phase converter rj , the capacitor voltage u of the i-th submodule jrci , the switch status signal S of the i-th submodule jri (j=a, b, c, r=p, n, i=1, 2…N), j-phase AC grid voltage u sj And the grid-side output current i of phase j sj .

[0073] Figure 31 is a schematic diagram of the reconstructed subunit proposed in the present invention, with the positive direction of the upper (p) bridge arm current and the positive direction of the lower (n) bridge arm current as the reference, and the submodules through which the bridge arm current flows in sequence are marked as 1, 2, ..., N. The submodules with the same number in the three-phase upper (p) bridge arm are regarded as the same subunit, and are marked as Sub p1 、Sub p2 、…、Sub pN ; The submodules with the same number in the three-phase lower (n) bridge arm are regarded as the same subunit and are denoted as Sub n1 、Sub n2 、…、Sub nN .

[0074] Figure 4 is a structural diagram of the upper bridge arm reconstruction subunit 1 proposed in the present invention, and the upper (p) bridge arm reconstruction subunit Sub p1 As an example, the upper (p) bridge arm subunit Sub after group reconstruction is shown p1 The specific circuit structure of each sub-module.

[0075] Figure 5 This is a flow chart of the steps of the method for diagnosing and locating the open-circuit fault of the half-bridge submodule IGBT of the three-phase converter system based on the adaptive extended Kalman filter algorithm of the present invention, which illustrates the detailed steps of the method for diagnosing and locating the single-tube open-circuit fault of the three-phase converter submodule IGBT (Insulate-Gate Bipolar Transistor) provided by the present invention, including two parts: fault diagnosis and fault location. Fault diagnosis is to determine whether the submodule IGBT open-circuit fault occurs by optimally estimating the three-phase circulating current a, b, c and the output current through the adaptive extended Kalman filter algorithm, and to distinguish the fault phase and the position of the fault bridge arm; fault location is to group and reconstruct the subunits through the fault bridge arm, and then use the adaptive extended Kalman filter algorithm to optimally estimate the reconstructed subunit voltage, and locate the faulty submodule according to the reconstructed subunit voltage residual.

[0076] The above-mentioned fault diagnosis specifically includes the following steps:

[0077] Step 1.1, let i be the i-th submodule in the bridge arm of the three-phase converter, initialize i=0, and the counter value t in the fault diagnosis = 0; set the initial Kalman gain K a_cir , K b_cir , K c_cir , K a , K b , K c , K Subpi , K Subni , circulation reliability coefficient α cir , α, αSubp , α Subn ;

[0078] Assume that the converter three-phase circulating current threshold is I a_cir_th ,I b_cir_th ,I c_cir_th , the three-phase output current threshold of the converter is I a_th ,I b_th ,I c_th , the fault diagnosis duration threshold is T th , the fault location duration threshold is T th_Sub ;Reconstruct submodule voltage threshold U sub_th , the error autocovariance of the three-phase bridge arm circulating current estimation is P a_cir_mid , P b_cir_mid , P c_cir_mid , the error autocovariance of the optimal estimation of the three-phase bridge arm circulating current is P a_cir_now , P b_cir_now , P c_cir_now , the error autocovariance of the three-phase output current estimation of the converter is P a_mid , P b_mid , P c_mid The error autocovariance of the optimal estimate of the three-phase output current of the converter is P a_now , P b_now , P c_now , the error autocovariance of the voltage estimation of the i-th reconstruction submodule of the upper bridge arm is P Subpi_mid , the error autocovariance of the optimal estimation of the voltage of the i-th reconstruction submodule on the upper bridge arm is P Subpi_now , the error autocovariance of the voltage estimation of the i-th reconstructed subunit in the lower bridge arm is P Subni_mid , the error autocovariance P of the optimal estimate of the voltage of the i-th reconstructed subunit in the lower bridge arm Subni_now ;

[0079] Initialize the converter three-phase circulating current process noise variable Q a_cir , Q b_cir , Q c_cir , converter three-phase circulating current measurement noise variable R a_cir , R a_cir , R a_cir , inverter three-phase output current process noise variable Q a , Q b , Q c , the inverter three-phase output current measurement noise variable R a , R b , R c , the noise variable Q in the subunit grouping reconstruction process Subp , Q Subn , the noise variable R measured during the reconstruction of subunit groups Subp , RSubn ;

[0080] Step 1.2, collect the upper bridge arm current i in the three-phase pa 、i pb 、i pc and the lower bridge arm current i na 、i nb 、i nc , the switch status of each submodule of the upper bridge arm S jpi And the switch status S of each sub-module of the lower bridge arm jni , the capacitor voltage u of each submodule in the upper bridge arm jpci and the capacitor voltage u of each submodule in the lower bridge arm jnci , the DC side voltage of the converter v dc , the output current i of the converter AC side sa 、i sb 、i sc , grid voltage u sa 、u sb 、u sc ;

[0081] Step 1.3, calculate the three-phase circulation of a, b, and c according to formula (1);

[0082]

[0083] According to Kirchhoff's voltage law, the equivalent circulating current control loop of the submodule is obtained as follows (2):

[0084]

[0085] In the above formula, L arm is the equivalent inductance of the bridge arm, v dc is the DC side voltage of the converter, i a_cir 、i b_cir 、i c_cir is the three-phase bridge arm circulation, u pa 、u pb 、u pc is the three-phase upper bridge arm voltage, u na 、u nb 、u nc is the three-phase lower bridge arm voltage;

[0086] The equivalent AC control loop of the converter is also obtained from Kirchhoff's voltage law, as shown in equation (3):

[0087]

[0088] In the above formula, L arm is the equivalent inductance of the bridge arm, R arm is the equivalent resistance of the bridge arm, i sa 、isb 、i sc is the three-phase output current of the AC side of the converter, u pa 、u pb 、u pc is the three-phase upper bridge arm voltage, u na 、u nb 、u nc is the three-phase lower bridge arm voltage, u sa 、u sb 、u sc is the three-phase grid voltage;

[0089] By performing backward difference discretization on equation (2) and equation (3), we can obtain:

[0090]

[0091] In the above formula, L arm is the equivalent inductance of the bridge arm, R arm is the equivalent resistance of the bridge arm, f s is the system sampling frequency used in the discretization process; i a_cir (k), i b_cir (k), i c_cir (k) is the three-phase bridge arm circulation at time k, i a_cir (k-1), i b_cir (k-1), i c_cir (k-1) are the three-phase bridge arm circulation at time k-1; i sa (k), i sb (k), i sc (k) is the three-phase output current at time k, i sa (k-1), i sb (k-1), i sc (k-1) is the three-phase output current at time k-1; u pa (k-1),u pb (k-1),u pc (k-1) is the three-phase upper bridge arm voltage at time k-1, u na (k-1),u nb (k-1),u nc (k-1) is the three-phase lower bridge arm voltage at time k-1; u sa (k-1),u sb (k-1),u sc (k-1) are the three-phase grid voltages at time k-1;

[0092] Step 1.4, solving the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation according to formula (4); determining the optimal circulation amount of the three-phase bridge arm based on the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation; specifically as follows:

[0093] ① The relationship between the output voltage of the i-th submodule of each phase upper and lower bridge arm and the capacitor voltage and switch state is:

[0094]

[0095] In the formula, u jpi 、u jni are the output voltages of the i-th submodule of the upper and lower bridge arms of phase j respectively; S jpi , S jni u are the switch states of the i-th submodule of the upper and lower bridge arms of phase j, 1 means on, 0 means off; jpci 、u jnci are the capacitor voltages of the i-th submodule of the upper and lower bridge arms of phase j respectively.

[0096] And the state equation of the capacitor voltage of the i-th submodule in each bridge arm is:

[0097]

[0098] In the above formula, C sm is the submodule capacitance, i pj 、i nj is the upper and lower bridge arm current.

[0099] According to formula (6), the voltage output values ​​u of the upper and lower bridge arms of each phase at time k can be obtained: pj 、u nj :

[0100]

[0101] ② Calculate the estimated value of the circulating current state of the three-phase bridge arms a, b, and c at time k:

[0102]

[0103] Among them, i a_cir_mid 、i b_cir_mid 、i c_cir_mid are the estimated circulating currents of the three-phase bridge arms a, b, and c, respectively, a_cir_now 、i b_cir_now 、i c_cir_now They are the optimal estimated values ​​of the circulating currents of the three-phase bridge arms a, b, and c respectively.

[0104] ③Calculate the prediction error covariance at time k:

[0105]

[0106] Among them, P a_cir_mid , P b_cir_mid , P c_cir_mid are the error autocovariance of the circulating current estimation of the three-phase bridge arms a, b, and c, respectively, a_cir_now , P b_cir_now , P c_cir_now are the error autocovariance of the optimal estimation of the circulating current of the three-phase bridge arm a, b, and c, respectively, Q a_cir , Q b_cir , Q c_cir is the noise variable of the three-phase circulating current process of converters a, b and c.

[0107] ④Calculate the Kalman gain at time k:

[0108]

[0109] Among them, K a_cir , K b_cir , K c_cir are the adaptive Kalman gains after the three-phase circulating currents a, b, and c are introduced with their respective circulating current residuals, R a_cir , R b_cir , R c_cir is the noise variable of the three-phase circulating current measurement of converters a, b, and c, α cir is the set circulating current reliability factor.

[0110] ⑤ Calculate the optimal estimated value of the circulating current of the three-phase bridge arms a, b, and c at time k:

[0111]

[0112] Among them, i a_cir 、i b_cir 、i c_cir are the three-phase bridge arm circulating currents a, b, and c calculated according to formula (1).

[0113] ⑥ Update the prediction estimation error covariance at time k:

[0114]

[0115] Where: K a_cir , K b_cir , K c_cir are the adaptive Kalman gains after the three-phase circulating currents a, b, and c are introduced with their respective circulating current residuals, and P a_cir_mid , P b_cir_mid , P c_cir_mid are the error autocovariance of the circulating current estimation of the three-phase bridge arms a, b, and c, respectively, a_cir_now , P b_cir_now , P c_cir_nowThey are the error autocovariance of the optimal estimation of the circulating current of the three-phase bridge arms a, b, and c respectively.

[0116] At this point, an adaptive optimal state observer of the extended Kalman filter is designed to realize the optimal estimation process of the circulating current of the three-phase bridge arms a, b, and c.

[0117] Solve the adaptive optimal state observer of the extended Kalman filter of the three-phase output current according to equation (5); determine the optimal amount of the three-phase output current based on the adaptive optimal state observer of the extended Kalman filter of the three-phase output current; specifically as follows:

[0118] ① Calculate the estimated output current values ​​of three phases a, b, and c at time k:

[0119]

[0120] Among them, i sa_mid 、i sb_mid 、i sc_mid are the estimated output currents of the three phases a, b, and c, respectively, i sa_now 、i sb_now 、i sc_now They are the optimal estimated values ​​of the three-phase output currents a, b, and c respectively.

[0121] ② Calculate the prediction error covariance at time k:

[0122]

[0123] Among them, P a_mid , P b_mid , P c_mid are the error autocovariances of the output current estimates of the three phases a, b, and c, respectively, a_now , P b_now , P c_now are the error autocovariance of the optimal estimation of the three-phase output current a, b, and c, respectively, Q a , Q b , Q c is the process noise variable of the three-phase output current a, b, and c.

[0124] ④Calculate the Kalman gain at time k:

[0125]

[0126] Among them, K a , K b , K c are the adaptive Kalman gains of the three-phase output currents a, b, and c, respectively, and R a , R b , R cis the measurement noise variable of the three-phase output current a, b, c, and α is the set output current reliability coefficient.

[0127] ⑤ Calculate the optimal estimated values ​​of the three-phase output currents a, b, and c at time k:

[0128]

[0129] Among them, i sa 、i sb 、i sc They are the actual sampled three-phase output currents a, b, and c respectively.

[0130] ⑥ Update the prediction estimation error covariance at time k:

[0131]

[0132] Where: K a , K b , K c are the adaptive Kalman gains of the three-phase output currents a, b, and c, respectively, a_mid , P b_mid , P c_mid are the error autocovariances of the output current estimates of the three phases a, b, and c, respectively, a_now , P b_now , P c_now They are the error autocovariance of the optimal estimation of the three-phase output current a, b, and c respectively.

[0133] At this point, an adaptive optimal state observer of the extended Kalman filter is designed to realize the optimal estimation process of the three-phase output currents a, b, and c.

[0134] Step 1.5, based on the converter, calculate the three-phase bridge arm circulating current residual I a_cir_err ,I b_cir_err ,I c_cir_err , calculate the three-phase output current residual I a_err ,I b_err ,I c_err ;

[0135] If the absolute value of the circulating current residual of one of the three phases (j = a, b, c) and the absolute value of the output current residual are equal to the set circulating current threshold I j_cir_th , output current threshold I j_th If the circulating current residual and output current residual of the bridge arm are both greater than or equal to the preset threshold, the corresponding fault phase flag is set to 1 and step 1.6 is executed; otherwise, the fault phase flag is set to 0 and the process returns to step 1.2;

[0136] The residual current I of the three-phase bridge arm a, b, and c is calculated by formula (19): a_cir_err ,Ib_cir_err ,I c_cir_err , calculate the residual output current I of the three phases a, b, and c by formula (20) a_err ,I b_err ,I c_err .

[0137]

[0138] According to formula (21), the absolute value of the circulating current residual of phase j (j = a, b, c), the absolute value of the output current residual and the set circulating current threshold I are determined. j_cir_th , output current threshold I j_th If the j-phase circulating current residual and output current residual are both greater than or equal to the set threshold, the corresponding fault phase flag will be j (j=a, b, c) are set to 1, and step 1.6 is executed; otherwise, the fault phase flag is set to j (j=a, b, c) are all set to 0, and the process returns to step 1.2.

[0139]

[0140] Step 1.6, clear the counter value in the fault diagnosis, and perform fault diagnosis based on the positive and negative relationship between the circulating current residual and the output current residual.

[0141] According to the positive and negative relationship between the circulating current residual and the output current residual, the fault diagnosis is performed. The fault code is recorded as Fault_flag j , the specific fault diagnosis judgment relationship is shown in formula (22):

[0142]

[0143] Where, Fault_flag j =1 means that the VT1 switch tube of the submodule of the bridge arm on phase j (p) has an open circuit fault; Fault_flag j =2 means that the switch tube of the submodule VT2 of the bridge arm on phase j (p) has an open circuit fault; Fault_flag j =3 means that the switch tube of the submodule VT1 of the lower (n) bridge arm of phase j has an open circuit fault; Fault_flag j =4 means that the VT2 switch tube of the lower (n) bridge arm of phase j has an open circuit fault. So far, the open circuit fault can be initially located in the r (r=p, n) bridge arm of phase j, and it can be determined whether it is a VT1 open circuit fault or a VT2 open circuit fault, and then the fault location link is entered.

[0144] The fault location specifically includes the following steps:

[0145] Step 2.1, based on the fault diagnosis result, obtain the fault code, and reconstruct the subunit according to the fault code; distinguish and mark the faults located in the upper bridge arm and the lower bridge arm; when the upper bridge arm fails, execute step 2.2, and when the lower bridge arm fails, execute step 2.4;

[0146] According to the fault code Fault_flag j Reconstruct the three-phase MMC subunit. If the fault code Fault_flag j =1 or 2, the fault is located in the upper bridge arm. The positive direction of the upper bridge arm current is taken as the reference, and the submodules through which the bridge arm current flows are marked as 1, 2, ..., N. The submodules with the same number in the upper bridge arm of phase a, b, and c are regarded as the same subunit and are marked as Sub p1 、Sub p2 、…、Sub pN ; If the fault code Fault_flag j =3 or 4, the fault is located in the lower bridge arm. The positive direction of the lower bridge arm current is taken as the reference, and the submodules through which the bridge arm current flows in sequence are marked as 1, 2, ..., N. The submodules with the same number in the lower bridge arm of phase a, b, and c are regarded as the same subunit and are marked as Sub n1 、Sub n2 、…、Sub nN .

[0147] Step 2.2, collecting the voltage of the ith submodule of the three-phase upper bridge arm; based on the voltage of the ith submodule of the three-phase upper bridge arm, determining the voltage of the ith reconstruction subunit of the upper bridge arm; specifically expressed as:

[0148] u Subpi =u api +u bpi +u cpi (twenty three)

[0149] In the above formula, u Subpi is the voltage of the i-th reconstructed subunit in the upper bridge arm, u api 、u bpi 、u cpi are the voltages of the ith submodule of the three-phase upper bridge arm respectively;

[0150] Formula (23) can be expressed as:

[0151]

[0152] In the above formula, C sm is the submodule capacitance, S pai , S pbi , S pci are the switch states of the i-th submodule of the three-phase upper bridge arm, i pa 、i pa、i pb are the three-phase upper bridge arm currents respectively, and t is the counter value in fault diagnosis;

[0153] According to formula (24), a reconstructed sub-unit voltage optimal estimation observer is established, and based on the reconstructed sub-unit voltage optimal estimation observer, the optimal estimation value of the reconstructed sub-unit voltage is determined; specifically, as follows:

[0154] ① Calculate the estimated value of the voltage of the i-th reconstructed subunit at time k:

[0155]

[0156] Among them, u Subpi_mid is the estimated value of the voltage of the i-th reconstructed subunit in the upper bridge arm, u Subpi_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the upper bridge arm.

[0157] ② Calculate the prediction error covariance at time k:

[0158] P Subpi_mid (k) = P Subpi_now (k-1)+Q Subp (26)

[0159] Among them, P Subpi_mid is the error autocovariance of the voltage estimation of the i-th reconstructed subunit in the upper bridge arm, P Subpi_now is the error autocovariance of the optimal estimate of the voltage of the i-th reconstructed subunit in the upper bridge arm, Q Subp Process noise variation for reconstructing the subcell voltage for the upper arm.

[0160] ③Calculate the Kalman gain at time k:

[0161]

[0162] Among them, K Subpi is the adaptive Kalman gain of the i-th reconstruction subunit voltage of the upper bridge arm, R Subp is the measurement noise variable for reconstructing the sub-unit voltage on the upper bridge arm, α Subp Reconstruct the sub-unit voltage reliability factor for the set upper bridge arm.

[0163] ④ Calculate the optimal estimated value of the voltage of the i-th reconstructed subunit at time k:

[0164] u Subpi_now (k) = u Subpi_mid (k)+K Subpi (k)·[u Subpi (k)-u Subpi_mid (k)] (28)

[0165] Among them, u Subpiis the voltage of the i-th reconstructed subunit of the upper bridge arm calculated according to formula (23).

[0166] ⑤ Calculate the prediction estimation error covariance at time k:

[0167] P Subpi_now (k) = [1-K Subpi (k)]·P Subpi_mid (k) (29)

[0168] Where: K Subpi is the adaptive Kalman gain of the i-th reconstruction subunit voltage of the upper bridge arm, P Subpi_mid is the error autocovariance of the voltage estimation of the i-th reconstructed subunit in the upper bridge arm, P Subpi_now is the error autocovariance of the optimal estimate of the voltage of the i-th reconstructed subunit in the upper bridge arm.

[0169] At this point, the optimal state observer for reconstructing the sub-unit voltage in the upper bridge arm is designed.

[0170] Step 2.3, based on the optimal estimated value of the reconstructed subunit voltage, calculate the upper bridge arm i-th reconstructed subunit voltage residual according to the following formula;

[0171] U Subpi_err =u Subpi_now -u Subpi (30)

[0172] In the above formula, U Subpi_err is the voltage residual of the i-th reconstructed subunit in the upper bridge arm, u Subpi_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the upper bridge arm, u Subpi is the actual sampling value of the voltage of the i-th reconstructed subunit of the upper bridge arm.

[0173] If the voltage residual U of the i-th reconstruction subunit in the upper bridge arm Subpi_err Greater than or equal to the reconstructed subunit voltage preset threshold U sub_th , and the duration exceeds T th_Sub , then the fault is located at the i-th reconstruction subunit in the upper bridge arm, combined with the fault phase flag flag and the fault code Fault_flag j , it is determined that the submodule open circuit fault is located in the i-th submodule of the upper bridge arm in one phase of the j (j=a, b, c) three-phase, and the fault location is completed. Otherwise, execute step 2.2;

[0174] Step 2.4, calculate the voltage u of the lower bridge arm reconstruction subunit Subni , specifically expressed as:

[0175] u Subni =u ani +u bni +ucni (31)

[0176] In the above formula, u Subni is the voltage of the i-th reconstructed subunit in the lower bridge arm, u ani 、u bni 、u cni is the voltage of the ith submodule of the three-phase lower bridge arm;

[0177] Formula (31) can be expressed as:

[0178]

[0179] In the above formula, C sm is the submodule capacitance, S nai , S nbi , S nci are the switch states of the i-th submodule of the three-phase lower bridge arm, i na 、i na 、i nb They are the three-phase lower bridge arm currents respectively;

[0180] According to formula (32), a second reconstructed subunit voltage optimal estimation observer is established, and based on the second reconstructed subunit voltage optimal estimation observer, an optimal estimation value of the second reconstructed subunit voltage is determined; specifically, as follows:

[0181] ① Calculate the estimated voltage value of the i-th reconstructed subunit at time k:

[0182]

[0183] Among them, u Subni_mid is the estimated value of the voltage of the i-th reconstructed subunit in the lower bridge arm, u Subni_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the lower bridge arm.

[0184] ② Calculate the prediction error covariance at time k:

[0185] P Subni_mid (k) = P Subni_now (k-1)+Q Subn (34)

[0186] Among them, P Subni_mid is the error autocovariance of the voltage estimation of the i-th reconstructed subunit in the lower bridge arm, P Subni_now is the error autocovariance of the optimal estimate of the voltage of the i-th reconstructed subunit in the lower bridge arm, Q Subn Process noise variation for reconstructing the subcell voltage for the lower arm.

[0187] ③Calculate the Kalman gain at time k:

[0188]

[0189] Among them, K Subni is the adaptive Kalman gain of the i-th reconstructed subunit voltage of the lower bridge arm, R Subn is the measurement noise variable for reconstructing the subcell voltage in the lower bridge arm, α Subn The reliability coefficient of the sub-unit voltage reconstructed for the set lower bridge arm.

[0190] ④ Calculate the optimal estimated value of the voltage of the i-th reconstructed subunit at time k:

[0191] u Subni_now (k) = u Subni_mid (k)+K Subni (k)·[u Subni (k)-u Subni_mid (k)] (36)

[0192] Among them, u Subni is the i-th reconstructed subunit voltage of the lower bridge arm calculated according to formula (31).

[0193] ⑤ Calculate the prediction estimation error covariance at time k:

[0194] P Subni_now (k) = [1-K Subni (k)]·P Subni_mid (k) (37)

[0195] Where: K Subni is the adaptive Kalman gain of the i-th reconstruction subunit voltage of the lower bridge arm, P Subni_mid is the error autocovariance of the voltage estimation of the i-th reconstructed subunit in the lower bridge arm, P Subni_now is the error autocovariance of the optimal estimate of the voltage of the i-th reconstructed subunit in the lower bridge arm.

[0196] At this point, the optimal state observer for reconstructing the sub-unit voltage in the lower bridge arm is designed.

[0197] Step 2.5, based on the optimal estimated value of the second reconstructed subunit voltage, calculate the i-th reconstructed subunit voltage residual of the lower bridge arm according to the following formula;

[0198] U Subni_err =u Subni_now -u Subni (38)

[0199] In the above formula, U Subni_err is the voltage residual of the i-th reconstructed subunit in the lower bridge arm, u Subni_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the lower bridge arm, u Subni is the actual sampling value of the i-th reconstruction sub-unit voltage of the lower bridge arm;

[0200] If the voltage residual of the i-th reconstructed subunit in the lower bridge arm is greater than or equal to the preset threshold of the reconstructed subunit voltage, the fault is located in the i-th reconstructed subunit in the lower bridge arm, and the fault phase flag flag and the fault code Fault_flag are combined. j , it is determined that the submodule open circuit fault is located in the i-th submodule of the lower bridge arm in one phase of the j (j=a, b, c) three-phase, and the fault location is completed. Otherwise, execute step 2.4.

[0201] The invention discloses a method for quickly diagnosing an IGBT open-circuit fault of a three-phase converter half-bridge submodule based on an adaptive optimal state observer of an extended Kalman filter and a method for locating an IGBT open-circuit fault submodule by using the residual of a reconstructed subunit voltage. The fault diagnosis method improves the dynamic response performance and robustness performance of the optimal estimation observer by introducing the residual of the optimal estimation value and the measured value in the Kalman gain. The observer is used to estimate the optimal state of the three-phase circulating current and the output current, and the positive and negative relationship between the residual of the optimal estimation value and the actual sampling value is used to diagnose the fault phase and the fault bridge arm position. After determining the position of the fault bridge arm, the submodules are grouped and reconstructed according to the position of the fault bridge arm, and the fault submodule is located by using the residual value of the reconstructed subunit voltage. By using the method proposed by the invention, when a single or multiple submodule IGBT open-circuit fault occurs in a three-phase MMC, the fault submodule can be diagnosed and located quickly and accurately, and the problem of being unable to identify the fault submodule when the sorting algorithm is used to control the submodule capacitor voltage balance can be solved.

[0202] The above description is only a specific embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for rapid diagnosis and location of IGBT open circuit fault in a converter submodule, characterized in that: include: A three-phase converter connected to a power grid, wherein each phase of the three-phase converter is composed of two upper and lower bridge arms, and each bridge arm is composed of a plurality of sub-modules connected in series and an inductor connected in series; The method comprises: fault diagnosis and fault location; The fault diagnosis is to determine whether a submodule open circuit fault occurs by optimally estimating the three-phase circulating current and output current through an adaptive extended Kalman filter algorithm, and to identify the fault position of the bridge arm in the faulty phase among the three phases; The fault location is to group and reconstruct the sub-modules in the faulty bridge arm according to the faulty bridge arm corresponding to the faulty position of the bridge arm to form a reconstructed sub-unit, and then use an adaptive extended Kalman filter algorithm to optimally estimate the voltage of the reconstructed sub-unit, and locate the faulty sub-module based on the residual of the voltage of the reconstructed sub-unit.

2. The method according to claim 1, characterized in that The submodule is a half-bridge submodule, and the half-bridge submodule includes: an upper tube and a lower tube connected in series, each of the upper tube and the lower tube is connected to a diode in reverse parallel, and the upper tube and the lower tube connected in series are connected to a capacitor in parallel as a whole.

3. The method according to claim 1, characterized in that The fault diagnosis specifically includes the following steps: Step 1.1, let i be the i-th submodule in the bridge arm of the three-phase converter, initialize i=0, and the counter value t in the fault diagnosis=0; set the initial Kalman gain and reliability coefficient; Initialize the converter three-phase circulating current process noise variable, the converter three-phase circulating current measurement noise variable, the converter three-phase output current process noise variable, the converter three-phase output current measurement noise variable, the noise variable in the sub-unit grouping reconstruction process, and the measurement noise variable in the sub-unit grouping reconstruction process; Step 1.2, collecting the three-phase bridge arm current, the DC side voltage of the converter, the three-phase upper bridge arm voltage, the three-phase lower bridge arm voltage, and the three-phase output current of the AC side of the converter; Step 1.3, calculating the three-phase circulating current of the converter; based on the three-phase bridge arm current, the DC side voltage of the converter, the three-phase upper bridge arm voltage, the three-phase lower bridge arm voltage, and the three-phase grid voltage, according to Kirchhoff's voltage law, the equivalent circulating current control loop of the submodule and the equivalent AC control loop of the converter are obtained; The equivalent circulation control loop expression of the submodule is as follows: In the above formula, L arm is the equivalent inductance of the bridge arm, v dc is the DC side voltage of the converter, i a_cir 、i b_cir 、i c_cir is the three-phase bridge arm circulation, u pa 、u pb 、u pc is the three-phase upper bridge arm voltage, u na 、u nb 、u nc is the three-phase lower bridge arm voltage; The equivalent AC control loop of the converter is expressed as follows: In the above formula, R arm is the equivalent resistance of the bridge arm, i sa 、i sb 、i sc is the three-phase output current of the AC side of the converter, u sa 、u sb 、u sc is the three-phase grid voltage; Formula (4) is obtained by back-differentiation discretization of formula (2), and formula (5) is obtained by back-differentiation discretization of formula (3): In the above formula, f s is the system sampling frequency used in the discretization process; i a_cir (k), i b_cir (k), i c_cir (k) is the three-phase bridge arm circulation at time k, i a_cir (k-1), i b_cir (k-1), i c_cir (k-1) are the three-phase bridge arm circulation at time k-1; i sa (k), i sb (k), i sc (k) is the three-phase output current at time k, i sa (k-1), i sb (k-1), i sc (k-1) is the three-phase output current at time k-1; u pa (k-1),u pb (k-1),u pc (k-1) is the three-phase upper bridge arm voltage at time k-1, u na (k-1),u nb (k-1),u nc (k-1) is the three-phase lower bridge arm voltage at time k-1; u sa (k-1),u sb (k-1),u sc (k-1) are the three-phase grid voltages at time k-1; Step 1.4, solving the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation according to formula (4); determining the optimal circulation amount of the three-phase bridge arm based on the adaptive optimal state observer of the extended Kalman filter of the three-phase bridge arm circulation; Solving the adaptive optimal state observer of the extended Kalman filter of the three-phase output current according to equation (5); determining the optimal amount of the three-phase output current based on the adaptive optimal state observer of the extended Kalman filter of the three-phase output current; Step 1.5, based on the optimal circulation amount of the three-phase bridge arm and the optimal amount of the three-phase output current, calculate the three-phase bridge arm circulating current residual and calculate the three-phase output current residual; If the circulating current residual and the output current residual of a phase bridge arm in the three-phase are both greater than or equal to the preset threshold, then execute step 1.6; otherwise, return to execute step 1.2; Step 1.6, clear the counter value in the fault diagnosis, and perform fault diagnosis based on the positive and negative relationship between the circulating current residual and the output current residual.

4. The method according to claim 1, characterized in that The fault location specifically includes the following steps: Step 2.1, based on the fault diagnosis result, obtain the fault code, and reconstruct the subunit according to the fault code; distinguish and mark the faults located in the upper bridge arm and the lower bridge arm; when the upper bridge arm fails, execute step 2.2, and when the lower bridge arm fails, execute step 2.4; Step 2.2, collecting the voltage of the ith submodule of the three-phase upper bridge arm; based on the voltage of the ith submodule of the three-phase upper bridge arm, determining the voltage of the ith reconstruction subunit of the upper bridge arm; specifically expressed as: in Subpi =in api +in bpi +in cpi (23) In the above formula, u Subpi is the voltage of the i-th reconstructed subunit in the upper bridge arm, u api 、u bpi 、u cpi are the voltages of the ith submodule of the three-phase upper bridge arm respectively; Formula (23) can be expressed as: In the above formula, C sm is the submodule capacitance, S pai , S pbi , S pci are the switch states of the i-th submodule of the three-phase upper bridge arm, i pa 、i pa 、i pb They are the three-phase upper arm currents respectively; Establishing a reconstructed subunit voltage optimal estimation observer according to formula (24), and determining the reconstructed subunit voltage optimal estimation value based on the reconstructed subunit voltage optimal estimation observer; Step 2.3, based on the optimal estimated value of the reconstructed subunit voltage, calculate the upper bridge arm i-th reconstructed subunit voltage residual according to the following formula; IN Subpi_err =in Subpi_now -in Subpi (30) In the above formula, U Subpi_err is the voltage residual of the i-th reconstructed subunit in the upper bridge arm, u Subpi_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the upper bridge arm, u Subpi is the actual sampling value of the voltage of the i-th reconstructed subunit of the upper bridge arm. If the voltage residual of the i-th reconstructed subunit in the upper bridge arm is greater than or equal to the preset threshold of the voltage of the reconstructed subunit, the fault is located in the i-th reconstructed subunit in the upper bridge arm. Combined with the fault phase mark and the fault code, it is determined that the submodule open circuit fault is located in the i-th submodule in the upper bridge arm in one of the three phases, and the fault location is completed. Otherwise, execute step 2.2; Step 2.4, calculate the voltage u of the lower bridge arm reconstruction subunit Subni , specifically expressed as: in Subni =in ani +in npii +in cni (31) In the above formula, u Subni is the voltage of the i-th reconstructed subunit in the lower bridge arm, u ani 、u bni 、u cni is the voltage of the ith submodule of the three-phase lower bridge arm; Formula (31) can be expressed as: In the above formula, C sm is the submodule capacitance, S nai , S nbi , S nci are the switch states of the i-th submodule of the three-phase lower bridge arm, i na 、i na 、i nb They are the three-phase lower bridge arm currents respectively; Establishing a second reconstructed subunit voltage optimal estimation observer according to formula (32), and determining a second reconstructed subunit voltage optimal estimation value based on the second reconstructed subunit voltage optimal estimation observer; Step 2.5, based on the optimal estimated value of the second reconstructed subunit voltage, calculate the i-th reconstructed subunit voltage residual of the lower bridge arm according to the following formula; IN Subni_err =in Subni_now -in Subni (38) In the above formula, U Subni_err is the voltage residual of the i-th reconstructed subunit in the lower bridge arm, u Subni_now is the optimal estimated value of the voltage of the i-th reconstructed subunit in the lower bridge arm, u Subni is the actual sampling value of the i-th reconstruction sub-unit voltage of the lower bridge arm; If the voltage residual of the i-th reconstructed subunit in the lower bridge arm is greater than or equal to the preset threshold of the reconstructed subunit voltage, the fault is located to the i-th reconstructed subunit in the lower bridge arm. Combined with the fault phase mark and the fault code, it is determined that the submodule open circuit fault is located in the i-th submodule in the lower bridge arm in one of the three phases, and the fault location is completed. Otherwise, execute step 2.4.

Images