Modular multilevel converter fault diagnosis method based on kalman filter

Abstract

This invention relates to the field of multilevel power electronic converter technology, specifically to a fault diagnosis method for modular multilevel converters based on a Kalman filter. It includes: S1: acquiring the required data; S2: calculating based on the mathematical relationship between circulating current and submodule capacitor voltage changes and the Kalman filter algorithm. k Optimal estimated value of circulating current, measured value of circulating current, optimal estimated value of capacitor voltage of submodule, and measured value of capacitor voltage; S3: based on k S4: Calculate the circulation error by taking the optimal estimated value of the circulation and the corresponding measured value, and then obtain the variance of the circulation error; r S5: After determining the faulty phase, the variance of the submodule capacitor voltage error is calculated to identify the faulty submodule. The faulty submodule is then replaced, and the process returns to S1. This process simultaneously achieves fault detection and location, enabling more accurate and timely diagnosis of open-circuit faults in submodules.

Description

Technical Field

[0001] This invention relates to the field of multilevel power electronic converter technology, and more specifically to a fault diagnosis method for modular multilevel converters based on Kalman filters. Background Technology

[0002] like Figure 1 The diagram shows the current topology of a modular multilevel converter (MMC) and its submodules. In the implementation shown, the MMC topology consists of six arms, each containing N identical submodules and one arm inductor. The submodules employ a half-bridge structure and include two power switches S. u S l Two diodes D u D l and a DC capacitor C i Among them, power switch S u and diode D u Composed of the upper tube, power switch S l and diode D l Composed of the lower transistor; diode D u Cathode-connected power switch S u The collector of the diode D u Anode-connected power switch S u The emitter of the diode D l Cathode-connected power switch S l The collector of the diode D l Anode-connected power switch S l The emitter; power switch S u emitter, power switch S l The collectors are connected to the current-inflow side of the submodule bridge arm, respectively, and the power switch S u The gate, power switch S l The gates of the power switch are connected to the control circuits that control the power switch to turn on and off, respectively; power switch S l The emitter is connected to the current outflow side of the submodule bridge arm, and the power switch S u The collector is connected to the current outflow side of the submodule bridge arm via a DC capacitor.

[0003] In the above structure, due to the existence of multiple submodules, each of which may fail, the reliability of the entire MMC may be threatened. In this case, designers are required to take effective measures to monitor and manage these submodules to ensure rapid identification and appropriate action in the event of a failure, thereby maintaining the operation of the entire MMC system. To address this issue, corresponding solutions already exist, but they mainly present the following two problems:

[0004] Firstly, in existing methods, fault detection and localization are often performed separately. This means that when the system detects a fault, it is first necessary to determine which submodule the fault occurred in before taking appropriate measures. This separation process may prolong the fault diagnosis time, causing the system to be in an unstable state during the diagnosis period, thereby affecting the overall performance of the system.

[0005] Secondly, in order to achieve accurate fault detection and location, complex algorithms and logic are usually used, which increases the difficulty of system design and implementation and may also complicate the fault diagnosis process. Summary of the Invention

[0006] This invention provides a fault diagnosis method for modular multilevel converters based on Kalman filters, thereby effectively solving the problems in the background art.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: a modular multilevel converter fault diagnosis method based on a Kalman filter, comprising:

[0008] S1: Collect the required data, including the capacitor voltage u of each submodule on the bridge arm. c and circulation i diff ;

[0009] S2: According to the circulating flow i diff and the capacitor voltage u c Using the changing mathematical relationships and the Kalman filter algorithm, the optimal estimate of the circulation i at time k is calculated. diff_c (k), and obtain the circulation measurement value i at this time. diff (k), and calculate the optimal estimate of the capacitor voltage u of the submodule at time k. c_c (k), and obtain the measured value of the capacitor voltage u at this time. c (k);

[0010] S3: Based on the optimal estimate of the circulation i diff_c (k) and the corresponding circulation measurement value i diff (k), calculate the circulation error Δi diff And calculate the variance of the circulation error Δi diff_var ;

[0011] S4: By passing the circulating error variance Δi diff_var With the pre-set threshold Δi diff_var_th Compare the data to determine the phase in which the fault occurred;

[0012] S5: After determining the phase where the fault occurred, based on the optimal estimated value of the capacitor voltage u c_c (k) and the corresponding capacitor voltage measurement value u c(k), the capacitor voltage error Au c , and the capacitor voltage error variance Au c_var of the sub-modules is calculated.

[0013] S6: The faulty sub-module is determined according to the capacitor voltage error variance Au c_var , and the faulty sub-module is replaced, and the process returns to S1.

[0014] Further, the mathematical relationship of the circulating current i diff and the capacitor voltage u c is as follows:

[0015]

[0016]

[0017] wherein, i diff (t) is the circulating current value at time t, i diff (t-1) is the circulating current value at time t-1, U dc (t) is the DC side voltage value at time t, U u (t) and U l (t) are the sum of the capacitor voltage of the upper and lower bridge arm sub-modules at time t; u c (t) is the capacitor voltage value at time t, u c (t-1) is the capacitor voltage value at time t-1, i arm (t) is the bridge arm current value at time t, s(t) is the switching function value at time t, and L is the inductance value of the bridge arm, with the unit of the parameter being H; C is the capacitance value of the sub-module, with the unit of the parameter being F.

[0018] Further, according to the mathematical relationship of the circulating current i diff and the capacitor voltage u c and the Kalman filter algorithm, the optimal estimation value i diff_c (k) of the circulating current and the optimal estimation value u c_c (k) of the capacitor voltage of the sub-module at time k are calculated, including:

[0019] According to the optimal estimation value i diff_c (k-1) of the circulating current and the optimal estimation value u c_c (k-1) of the capacitor voltage of the sub-module at the previous time, the mathematical relationship of the circulating current i diff and the capacitor voltage u c is used to calculate the predicted state estimation value i diff_p (k) of the circulating current and the predicted state estimation value u c_p (k) of the capacitor voltage of the sub-module at time k by using the state prediction step of the Kalman filter.

[0020] According to the optimal state estimation error covariance matrix P c (k-1) of the previous moment, a prior error covariance matrix P p (k) of the predicted state is calculated.

[0021] The Kalman gain K(k) at the k moment is calculated through the prior error covariance matrix P p (k) of the predicted state.

[0022] The state is updated by using the Kalman gain K(k), the loop current measurement i diff (k) and the capacitor voltage measurement u c (k), to obtain the optimal estimation value i diff_c (k) of the loop current at the k moment and the optimal estimation value u c_c (k) of the capacitor voltage of the sub-module.

[0023] The prior error covariance matrix P p (k) of the predicted state is corrected and updated according to the Kalman gain K(k), to obtain the self-covariance matrix P c (k) of the optimal state estimation error.

[0024] Further, the loop current predicted state estimation value i diff_p (k), the capacitor voltage predicted state estimation value u c_p (k) of the sub-module and the prior error covariance matrix P p (k) of the predicted state at the k moment are calculated according to the following formulas respectively:

[0025]

[0026]

[0027] P p (k)=A·P c (k-1)·A T +Q

[0028] Wherein, i diff_p (k) is the loop current predicted state estimation value at the k moment, i diff_c (k-1) is the optimal estimation value of the loop current at the k-1 moment, U dc (k) is the DC side voltage value at the k moment, U u (k) and U l (k) are the sum of the capacitor voltage of the upper and lower bridge arm sub-modules at the k moment respectively; u c_p (k) is the capacitor voltage predicted state estimation value of the sub-module at the k moment, u c_c(k-1) is the optimal estimation of capacitor voltage at k-1 time, C is the capacitor value of sub-module, i arm (k) is the bridge arm current at k time, s(k) is the switch function value at k time, ΔT is the sampling interval, Q is the variance of process noise, A and B are both 1.

[0029] Further, according to the updating step of Kalman filter, the optimal estimation of loop current at k time i diff_c (k) and the optimal estimation of capacitor voltage of the sub-module u c_c (k) and the self-covariance of optimal state estimation error P c (k), and the calculation formulae are respectively:

[0030] i diff_c (k) = i diff_p (k) + K(k)·(i diff (k) - i diff_p (k))

[0031] u c_c (k) = u c_p (k) + K(k)·(u c (k) - u c_p (k))

[0032] K(k) = P p (k) / (P p (k) + R)

[0033] P c (k) = [1-K(k)]·P p (k)

[0034] Wherein, i diff_c (k) is the optimal estimation of loop current at k time, i diff (k) is the measured value of loop current at k time, u c_c (k) is the optimal estimation of capacitor voltage of the sub-module at k time, u c (k) is the measured value of capacitor voltage at k time, and R is the variance of measurement noise.

[0035] Further, the loop current error variance Δi diff_var , and the calculation formula is:

[0036]

[0037] Wherein, Δi diff_var is the loop current error variance, Δi diff_i is the loop current error, Δi diff_avg is the average value of Δi diff_i , and n is the number of sampling points.

[0038] Further, the capacitor voltage error variance Δu c_var , the calculation formula is:

[0039]

[0040] Wherein, Δu c_var is the capacitor voltage error variance, Δu c_i is the capacitor voltage error, Δu c_avg is the average value of Δu c_i , and n is the sampling point number.

[0041] Further, the operation of judging the fault occurring phase is: comparing the loop current error variance Δi diff_var of the any phase with the pre-set threshold value Δi diff_var_th , when the loop current error variance Δi diff_var of the phase is greater than the threshold value Δi diff_var_th and lasts for the time ΔT1, the phase is considered to have a fault.

[0042] Further, the lasting time ΔT1 is [5ms, 10ms].

[0043] Further, the operation of determining the fault sub-module is: integrating the capacitor voltage error variance Δu c_var of each sub-module of the fault phase, the integration time is ΔT2, and the sub-module with the maximum integration value is determined as the fault sub-module.

[0044] Further, the integration time ΔT2 is [5ms, 10ms].

[0045] Further, the pre-set threshold value Δi diff_var_th is [1kA, 2kA].

[0046] The application also comprises a computer device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, when the processor executes the computer program, the method as above is realized.

[0047] The application also comprises a storage medium, which stores a computer program, when the processor executes the computer program, the method as above is realized.

[0048] The beneficial effects of the application are: 1, in the traditional sub-module open circuit fault diagnosis algorithm, the fault detection and fault positioning are generally carried out separately, which not only complicates the fault diagnosis process, but also prolongs the required fault diagnosis time; the application can realize the fault detection and positioning at the same time, and can more accurately and timely diagnose the sub-module open circuit fault.

[0049] 2. The detection algorithm of this invention is simple. It adopts a recursive calculation method that combines prediction and correction. At each moment, the optimal estimate of the current state can be obtained only based on the latest observation data. There is no need to store a large amount of historical observation data. It has good real-time performance, is easy to implement on a computer, and can reduce the computational burden and reduce processor requirements.

[0050] 3. Since the submodule fault diagnosis algorithm of the present invention does not involve system parameters, it is not affected by the uncertainty of system parameters and has high robustness.

[0051] 4. The submodule fault diagnosis algorithm of this invention does not require any modification to the hardware circuit, making it easy to understand and implement; since it only involves capacitor voltage, bridge arm current and switching function, the proposed method does not require additional hardware resources. Attached Figure Description

[0052] 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 recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a topology diagram of the three-phase MMC and its sub-modules;

[0054] Figure 2 This is a flowchart illustrating the implementation of the method proposed in this invention.

[0055] Figure 3 This is a schematic diagram of the structure of a computer device. Detailed Implementation

[0056] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0057] like Figure 2 As shown: A modular multilevel converter fault diagnosis method based on Kalman filters, including:

[0058] S1: Collect the required data, including the capacitor voltage u of each submodule on the bridge arm. c and circulation i diff The above data forms the basis for fault diagnosis, and the system's operating status and characteristics can be analyzed through this data.

[0059] S2: Based on the circulation i diff and capacitor voltage u cUsing the changing mathematical relationships and the Kalman filter algorithm, the optimal estimate of the circulation i at time k is calculated. diff_c (k), and obtain the circulation measurement value i at this time. diff (k), and calculate the optimal estimate of the capacitor voltage u of the submodule at time k. c_c (k), and obtain the measured value of the capacitor voltage u at this time. c (k); In the above steps, the Kalman filter algorithm is used to complete the processing of the circulating current i. diff and capacitor voltage u c The Kalman filter algorithm is computationally efficient and real-time, enabling timely system state estimation and updates. It allows for rapid response to faults, reducing the time the system spends in an unstable state during diagnosis and improving system stability and availability. The circulating current i is calculated using the Kalman filter algorithm. diff The optimal estimate of the capacitor voltage u c_c (k) can accurately capture dynamic changes in the system and infer possible faults from them.

[0060] S3: Based on the optimal estimate of circulation i diff_c (k) and the corresponding circulation measurement i diff (k), calculate the circulation error Δi diff And calculate the variance of the circulation error Δi diff_var Circulation error Δi diff This represents the difference between the estimated value and the actual measured value, i.e., the deviation of the system in estimating the circulation, while the variance of the circulation error Δi diff_var This indicates the distribution range and stability of the circulation error. We use a Kalman filter to calculate the variance Δi of the circulation error. diff_var A smaller variance indicates that the circulation error is relatively stable, and the difference between the estimated and measured values ​​is small. A larger variance indicates that there is greater uncertainty or measurement error in the system. The variance of the circulation error Δi diff_var It reflects the fluctuation of the circulating error and is an important basis for judging whether a fault has occurred.

[0061] S4: By controlling the variance of the circulation error Δi diff_var With the preset threshold Δ idiff_var_th Compare and determine the phase in which the fault occurred; calculate the variance of the circulating current error Δi. diff_var With the preset threshold Δ idiff_var_th By comparing the results, if the threshold is exceeded, it can be determined that the corresponding phase has failed. The purpose of this stage is to quickly identify the parts of the system that may have problems, so as to facilitate further diagnosis and maintenance.

[0062] S5: After determining the phase where the fault occurred, based on the optimal estimate of the capacitor voltage uc_c (k) and the corresponding capacitor voltage measurement value u c (k), the capacitor voltage error Au c and the capacitor voltage error variance Au c_var of the sub-module is calculated c_var reflects the fluctuation of the capacitor voltage of the sub-module and can be used for locating the faulty sub-module;

[0063] S6: The faulty sub-module is determined according to the capacitor voltage error variance Au c_var , and then the faulty sub-module is replaced, and the process returns to S1. Once the faulty sub-module is determined, appropriate measures such as replacing the sub-module can be taken to handle the fault;

[0064] After the fault is handled, the process returns to S1 to restart the data acquisition and processing process to monitor the running state of the system.

[0065] In this embodiment, the circulating current i diff and the capacitor voltage u c change the mathematical relationship formula respectively as:

[0066]

[0067]

[0068] wherein, i diff (t) is the circulating current value at time t, i diff (t-1) is the circulating current value at time t-1, U dc (t) is the DC side voltage value at time t, U u (t) and U l (t) are the sum of the upper and lower bridge arm sub-module capacitor voltages at time t respectively; u c (t) is the capacitor voltage value at time t, u c (t-1) is the capacitor voltage value at time t-1, i arm (t) is the bridge arm current value at time t, s(t) is the switching function value at time t, and L is the inductance value of the bridge arm, with the unit of the parameter being H; C is the capacitance value of the sub-module, with the unit of the parameter being F.

[0069] The mathematical relationship between the circulating current i diff and the capacitor voltage u c change provides a basis for the subsequent Kalman filter algorithm, and at the same time, can accurately describe the dynamic behavior of the system, providing a theoretical basis for fault detection and location.

[0070] As a preferred embodiment of the above embodiment, according to the circulating current i diff and the capacitor voltage u cThe mathematical relationship and the Kalman filter algorithm are used to calculate the optimal estimation value i of the circulating current at time k diff_c (k) and the optimal estimation value u of the capacitor voltage of the sub-module c_c (k) includes:

[0071] The state prediction method of the Kalman filter is used to calculate the predicted state estimation value i of the circulating current at time k according to the optimal estimation value i diff_c (k-1) and the optimal estimation value u of the capacitor voltage of the sub-module c_c (k-1) through the circulating current i diff and the capacitor voltage u c The mathematical relationship formula is used to calculate the predicted state estimation value i of the circulating current at time k diff_p (k) and the predicted state estimation value u of the capacitor voltage of the sub-module c_p (k);

[0072] The prior error covariance matrix P c (k-1) at time k is calculated according to the optimal state estimation error covariance matrix P p (k-1);

[0073] The Kalman gain K(k) at time k is calculated through the prior error covariance matrix P p (k) of the predicted state;

[0074] The state is updated using the Kalman gain K(k), the circulating current measurement value i diff (k) and the capacitor voltage measurement value u c (k) to obtain the optimal estimation value i of the circulating current at time k diff_c (k) and the optimal estimation value u of the capacitor voltage of the sub-module c_c (k);

[0075] The prior error covariance matrix P p (k) of the predicted state is corrected and updated according to the Kalman gain K(k) to obtain the self-covariance matrix P c (k) of the optimal state estimation error.

[0076] By calculating the prior error covariance matrix P p (k) of the predicted state, the uncertainty of the predicted state can be quantified, which helps to more accurately weigh the observation value and the predicted value during state updating, improves the accuracy of state estimation, and also provides important information for subsequent Kalman gain K(k) calculation and state updating. The Kalman gain K(k) is a key parameter in the Kalman filter algorithm, which ensures the robustness and accuracy of the algorithm in dealing with uncertainty and noise. By reasonably calculating the Kalman gain, it can ensure that the observation information is fully utilized during state updating, while avoiding excessive dependence on inaccurate observation values.

[0077] In the embodiment, at time k, the circulating current prediction state estimation value i diff_p (k), the capacitor voltage prediction state estimation value u c_p (k) of the submodule, and the prior error covariance matrix P p (k) of the prediction state, the calculation formulas are respectively:

[0078]

[0079]

[0080] P p (k)=A·P c (k-1)·A T +Q

[0081] wherein i diff_p (k) is the circulating current prediction state estimation value at time k, i diff_c (k-1) is the optimal estimation value of the circulating current at time k-1, U dc (k) is the DC side voltage value at time k, U u (k) and U l (k) are the sum of the capacitor voltages of the upper and lower bridge arm submodules at time k respectively; u c_p (k) is the capacitor voltage prediction state estimation value of the submodule at time k, u c_c (k-1) is the optimal estimation value of the capacitor voltage at time k-1, and C is the capacitor value of the submodule, i arm (k) is the bridge arm current at time k, s(k) is the switching function value at time k, ΔT is the sampling interval, and Q is the variance of the process noise. Both A and B take the value of 1. The circulating current prediction state estimation value i diff_p (k) and the prior error covariance matrix P p (k) of the prediction state provide an estimation of the system state and an evaluation of the credibility of the estimation, thereby providing a basis and reference for the subsequent state updating process.

[0082] As a preferred embodiment of the above, according to the updating step of the Kalman filter, the optimal estimation value i diff_c (k) of the circulating current at time k and the optimal estimation value u c_c (k) of the capacitor voltage of the submodule, and the Kalman gain K(k) and the autocovariance P c (k) of the optimal state estimation error, the calculation formulas are respectively:

[0083] i diff_c (k)=i diff_p (k)+K(k)·(i diff (k)-i diff_p (k))

[0084] u c_c (k) = u c_p (k) + K(k) · (u c (k) - u c_p (k))

[0085] K(k) = P p (k) / (P p (k) + R)

[0086] P c (k) = [1 - K(k)] · P p (k)

[0087] Where, i diff_c (k) is the optimal estimation of the circulating current at time k, i diff (k) is the measured value of the circulating current at time k, u c_c (k) is the optimal estimation of the capacitor voltage of the sub-module at time k, u c (k) is the measured value of the capacitor voltage at time k, and R is the variance of the measurement noise.

[0088] The circulating current error variance Δi diff_var is calculated according to the optimal estimation i diff_c (k) of the circulating current at time k and the measured value i diff (k) of the circulating current, and the calculation formula is as follows:

[0089]

[0090] Where, Δi diff_var is the circulating current error variance, Δi diff_i is the circulating current error, Δi diff_avg is the average value of Δi diff_i , n can be 20, and the sampling interval is 0.5 ms.

[0091] The capacitor voltage error variance Δu c_c is calculated according to the optimal estimation u c (k) of the capacitor voltage of the sub-module at time k and the measured value u c_var (k) of the capacitor voltage, and the calculation formula is as follows:

[0092]

[0093] Where, Δu c_var is the capacitor voltage error variance, Δu c_i is the capacitor voltage error, and Δu c_avg is the average value of Δu c_i .

[0094] In the specific implementation process, the operation of judging the fault phase is as follows: the circulating current error variance Δidiff_var is compared with a preset threshold value Δi diff_var_th , when the phase current error variance Δi diff_var is greater than the threshold value Δi diff_var_th and lasts for a time ΔT1, it is considered that the phase has a fault. By comparing the loop current error variance with the preset threshold value, it can be determined in real time whether a fault occurs, which helps to find the fault in time and take corresponding measures to prevent the fault from expanding and causing greater impact on the system.

[0095] wherein the loop current error variance Δi diff_var is greater than the threshold value Δi diff_var_th and lasts for a time ΔT1 in the range of [5ms, 10ms], and the time ΔT1 can be specifically 5ms.

[0096] As a preferred embodiment of the above embodiment, the operation of determining the fault sub-module is: integrating the capacitor voltage error variance Δu c_var of each sub-module of the fault phase, the integration time being ΔT2, ΔT2 being in the range of [5ms, 10ms], and the integration value being specifically 5ms, and the sub-module with the largest integration value being determined as the fault sub-module, and the sub-module fault monitoring process being restarted from S1 after replacing the fault sub-module. By integrating and comparing the capacitor voltage error variances of each sub-module of the fault phase, the fault sub-module can be accurately determined, which helps to quickly locate the fault source, improves the maintenance efficiency, and reduces the maintenance cost.

[0097] In the above embodiment, the preset threshold value Δi diff_var_th is in the range of [1kA, 2kA], and can be specifically 1.8kA.

[0098] Further, the processor executes the computer program to implement the method of any one of claims 1-12.

[0099] Further, the computer program is executed by the processor to implement the method of any one of claims 1-12.

[0100] The application is especially suitable for MMC systems with a large number of sub-modules, and compared with the traditional sub-module fault diagnosis method, the application can significantly reduce the calculation amount of the diagnosis algorithm. The proposed method analyzes the loop current and the capacitor voltage to perform fault diagnosis. When a sub-module has a fault, the loop current of the fault phase will change obviously, and the capacitor voltage of the fault sub-module will be different from that of the normal sub-module. Therefore, in the proposed method, the loop current and the capacitor voltage are monitored. Since only the capacitor voltage, the bridge arm current and the switching function are involved, the proposed method does not require additional hardware resources. Since the system parameters are not involved, the method is not affected by parameter uncertainty, and has high robustness.

[0101] Please refer to Figure 3The structural schematic diagram of the computer device provided by the embodiment of the application is shown. The computer device 400 provided by the embodiment of the application comprises a processor 410 and a memory 420, the memory 420 stores a computer program executable by the processor 410, and the computer program is executed by the processor 410 to perform the method as above.

[0102] The embodiment of the application further provides a storage medium 430, the storage medium 430 stores a computer program, and the computer program is executed by the processor 410 to perform the method as above.

[0103] The storage medium 430 can be implemented by any type of volatile or nonvolatile storage device or combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0104] In the description of the application, the terms "first", "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include one or more of the features. The meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0105] In the application, unless otherwise specifically defined and limited, the terms "mounting", "connection", "connection", "fixing" and the like should be understood in a broad sense, for example, it can be fixed connection, or detachable connection, or integral; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through an intermediate medium; it can be the internal communication of two elements or the interaction relationship between two elements. For ordinary skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0106] In the description of the specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any suitable manner in any one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the specification and the features of the different embodiments or examples, without contradiction.

[0107] Any process or method descriptions or descriptions of the flow diagrams in the flow charts described herein, or otherwise described in the specification, can be understood as representing the steps of any one or more of the methods or processes described herein, and the scope of the preferred embodiments of the present application includes additional implementation in which the steps are performed in a different order, including an order that is substantially simultaneous, or in reverse order, depending upon the functionality involved. The descriptions of the flow diagrams and / or steps described herein, or otherwise described in the specification, can be understood as representing any one or more of the steps of any one or more of the methods or processes described herein, and the scope of the preferred embodiments of the present application includes additional implementation in which the steps are performed in a different order, including an order that is substantially simultaneous, or in reverse order, depending upon the functionality involved.

[0108] The logic and / or steps represented in the flow charts described herein, or otherwise described in the specification, for example, can be considered as a list of executable instructions for implementing the logic function, and can be specifically embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus or device, such as a computer-based system, a system including a processor, or other system that can fetch the instructions from the instruction execution system, apparatus or device and execute the instructions. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate or transport a program for use by or in connection with an instruction execution system, apparatus or device, or in conjunction with these instruction execution systems, apparatus or devices. More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connections having one or more wires (electronic devices), portable computer disks (magnetic devices), random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memories), fiber optic devices, and portable compact disc read-only memories (CDROMs). In addition, a computer-readable medium can even be paper or other suitable medium on which the program can be printed, as the program can be electronically obtained, for example, by optical scanning of the paper or other medium, followed by electronic conversion into a useable form, and then stored in computer memory.

[0109] It should be understood that portions of the present application can be implemented with hardware, software, firmware or a combination thereof. In the above embodiments, a plurality of steps or methods can be implemented with software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, or their combinations can be used: discrete logic circuitry having logic gates for implementing logic functions on data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0110] Those skilled in the art of the present technology can understand that all or part of the steps carried out by the above-mentioned embodiments can be completed by a program instructing the relevant hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

[0111] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it should be understood that the above-mentioned embodiments are exemplary and cannot be understood as limiting the present application, and those skilled in the art can make changes, modifications, replacements and variations to the above-mentioned embodiments within the scope of the present application.

Claims

1. Modular multilevel converter fault diagnosis method based on Kalman filter, characterized in that, Comprising: S1 : Collect the required data, including the capacitor voltage u of each submodule on the bridge arm c and circulating current i diff ; S2: the circulating current i is estimated according to the circulating current i diff and the capacitor voltage u c and the capacitor voltage u diff_c (k), and the measured value of the circulating current i diff (k) is obtained, and the optimal estimated value of the capacitor voltage u c_c (k) of the sub-module at time k is calculated, and the measured value of the capacitor voltage u c (k) is obtained; S3: based on the ring current optimal estimate value i diff_c (k) and the corresponding ring current measurement value i diff (k), calculate the ring current error Δi diff and obtain the ring current error variance Δi diff_var ; S4: judging the fault occurrence phase by comparing the loop flow error variance Δi diff_var with a preset threshold Δi diff_var_th ​ S5: After determining the faulty phase, the optimal estimation of the capacitor voltage u c_c (k) and the corresponding capacitor voltage measurement value u c (k) to calculate the capacitor voltage error Δu c and to calculate the capacitor voltage error variance Δu c_var of the sub-modules; S6: determining the capacitance voltage error variance Δu c_var determining the fault sub-module, then replacing the fault sub-module, and returning to S1; The operation of the determining fault submodule is: calculating the capacitor voltage error variance Δu of each submodule of the fault phase c_var Integrating, the integration time is ΔT2, and the submodule with the largest integration value is determined as the fault submodule.

2. The Kalman filter based modular multilevel converter fault diagnostic method according to claim 1, characterized in that, The circulation i diff And the capacitor voltage u c The mathematical relationship formula of change is respectively: ； ； wherein, i diff (t) is the circulating current value at time t, i diff (t-1) is the circulating current value at time t-1, U dc (t) is the DC side voltage value at time t, U u (t) and U l (t) are the sum of the upper and lower bridge arm submodule capacitor voltages at time t, respectively; u c (t) is the capacitor voltage value at time t, u c (t-1) is the capacitor voltage value at time t-1, i arm (t) is the bridge arm current value at time t, s(t) is the switching function value at time t, L is the bridge arm inductance value, the parameter unit is H; C is the submodule capacitor capacity value, the parameter unit is F.

3. The Kalman filter based modular multilevel converter fault diagnostic method according to claim 2, characterized in that, According to the circulating current i diff and the capacitor voltage u c The mathematical relationship and Kalman filter algorithm are used to calculate the optimal estimation value of the circulating current i diff_c (k) and the optimal estimation value of the capacitor voltage u c_c (k) of the sub-module at time k, comprising: The state prediction method using the Kalman filter calculates the predicted state estimation value i(k) of the circulating current and the predicted state estimation value u(k) of the capacitor voltage of the sub-module at time k according to the optimal estimation value i(k-1) of the circulating current at time k-1 diff_c (k-1) and the optimal estimation value u(k-1) of the capacitor voltage of the sub-module at time k-1 through a mathematical relationship formula of the change c_c (k-1) of the circulating current i diff (k-1) and the capacitor voltage u c (k-1) of the sub-module, calculates the predicted state estimation value i(k) of the circulating current and the predicted state estimation value u(k) of the capacitor voltage of the sub-module at time k according to the mathematical relationship formula of the change diff_p (k) of the circulating current and the predicted state estimation value u(k) of the capacitor voltage of the sub-module at time k c_p (k) According to the optimal state estimation error covariance matrix P at time k-1 c (k-1), the prior error covariance matrix P of the predicted state at time k is calculated p (k); by the a priori error covariance matrix P of the predicted state p (k) calculating a Kalman gain K(k) at time k; using the Kalman gain K(k), the loop current measurement i diff (k) and the capacitor voltage measurement u c (k) to perform a state update to obtain an optimal loop current estimate i diff_c (k) and an optimal capacitor voltage estimate u c_c (k) of the sub-module. The prior error covariance matrix P(k) of the predicted state is corrected and updated according to the Kalman gain K(k) p (k) to obtain the optimal state estimation error self-covariance matrix P c (k).

4. The Kalman filter based modular multilevel converter fault diagnostic method according to claim 3, characterized in that, The circulating flow predicted state estimation value i diff_p The capacitor voltage predicted state estimation value u c_p (k) of the sub-module, and the prior error covariance matrix P p (k) of the predicted state, and the calculation formula is respectively: ； ； ； where i diff_p (k) is the circulating current prediction state estimation value at time k, i diff_c (k-1) is the optimal estimation value of circulating current at time k-1, U dc (k) is the DC side voltage value at time k, U u (k) and U l (k) are the sum of upper and lower bridge arm submodule capacitor voltages at time k, respectively; u c_p (k) is the capacitor voltage prediction state estimation value of the submodule at time k, u c_c (k-1) is the optimal estimation value of capacitor voltage at time k-1, C is the submodule capacitor value, i arm (k) is the bridge arm current at time k, s(k) is the switching function value at time k, ΔT is the sampling interval, Q is the variance of process noise, and A and B are both 1.

5. The Kalman filter based modular multilevel converter fault diagnostic method according to claim 4, characterized in that, According to the updating step of the Kalman filter, the optimal estimation value i diff_c (k) of the circulating current at the time k c_c (k) of the capacitor voltage of the sub-module, and the Kalman gain K(k) and the self-covariance P c (k) of the optimal state estimation error, the calculation formulas are respectively: ； ； ； ； where i diff_c (k) is the optimal estimate of the circulation at time k, i diff (k) is the measurement of the circulation at time k, u c_c (k) is the optimal estimate of the capacitor voltage of the submodule at time k, u c (k) is the measurement of the capacitor voltage at time k, R is the variance of the measurement noise.

6. The Kalman filter based modular multilevel converter fault diagnostic method of claim 1, wherein, The circulation error variance Δi diff_var The calculation formula is: ； where Δi diff_var is the circulation error variance, Δi diff_i is the circulation error, Δi diff_avg is the average of Δi diff_i , and n is the number of sampling points. 7.The Kalman filter based modular multilevel converter fault diagnosis method of claim 1, wherein, The capacitance voltage error variance Δu c_var The calculation formula is: ； where Δu c_var is the variance of the capacitance voltage error, Δu c_i is the capacitance voltage error, Δu c_avg is the average of Δu c_i , and n is the number of sampling points. 8.The Kalman filter based modular multilevel converter fault diagnosis method of claim 1, wherein, The operation of judging the phase where the fault occurs is to compare the loop current error variance Δi of any phase with the preset threshold value Δi diff_var When the loop current error variance Δi of the phase is greater than the threshold value Δi diff_var_th and lasts for a time ΔT1, the phase is considered to have a fault. diff_var When the loop current error variance Δi of the phase is greater than the threshold value Δi diff_var_th and lasts for a time ΔT1, the phase is considered to have a fault.

9. The Kalman filter based modular multilevel converter fault diagnostic method according to claim 8, characterized in that, said duration ΔΤ1 is [5ms, 10ms].

10. The Kalman filter based modular multilevel converter fault diagnostic method of claim 1, wherein, said integration time ΔΤ2 is [5ms, 10ms]. 11.The Kalman filter based modular multilevel converter fault diagnosis method of claim 1, wherein, the pre-set threshold value Δi diff_var_th is [1kA, 2kA].

12. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-11.

13. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the method of any one of claims 1-11.

Images