Converter valve internal fault detection method and device, converter system and computer equipment

Abstract

The application provides a converter valve fault detection method and device, a converter system and computer equipment, and belongs to the technical field of power transmission. The method comprises the following steps: adopting a differential current detection method of longitudinal differential, transverse differential and time domain differential to obtain the current difference value between the AC side and the DC side of the current source type main valve, the imbalance degree of the parallel branch current and the DC side current change rate; adopting a voltage detection method of amplitude detection, waveform similarity detection and dynamic response detection to obtain the bridge arm voltage deviation of the voltage source type support valve, the bridge arm voltage sequence correlation coefficient and the bridge arm voltage change slope; based on the detected current of the current source type main valve and the voltage of the voltage source type support valve, a current-voltage correlation analysis matrix is constructed, the combination mode of the current characteristics and the voltage characteristics is analyzed by the matrix, and the fault condition of the current source type main valve and the voltage source type support valve is determined. The application realizes accurate detection of the fault of the hybrid valve structure converter.

Description

Technical Field

[0001] This invention relates to the field of power transmission technology, specifically to a method for detecting faults inside a converter valve, a device for detecting faults inside a converter valve, a converter system, and a computer device. Background Technology

[0002] With the development of high-voltage direct current (HVDC) transmission technology, flexible DC transmission systems have become a key component of modern power systems due to their advantages such as high controllability, low loss, and suitability for new energy grid connection. The fully-controllable composite converter (F3C), as a novel hybrid converter topology, combines the high overload capacity of a current source converter (CSC) with the flexible control characteristics of a voltage source converter (VSC). It offers both high controllability and low loss advantages in flexible DC transmission, and has great potential in long-distance power transmission and offshore wind power grid connection.

[0003] The F3C converter incorporates both current-source main valves and voltage-source support valves. This hybrid valve structure complicates its fault characteristics, exhibiting the following features:

[0004] 1) Fault feature coupling: When the current source valve fails, the voltage source support valve may generate reverse current interference, leading to misjudgment by traditional differential protection;

[0005] 2) Mixed signal interference: The modulation strategy of the voltage source valve will affect the fault current waveform of the current source valve, and a single criterion is not enough to accurately distinguish the fault type;

[0006] 3) Difficulty in detecting local faults: When some components inside the valve are short-circuited or open-circuited, the fault signal is weak and the sensitivity of traditional methods is insufficient.

[0007] Existing converter fault detection methods are divided into current-source converter fault detection and voltage-source converter fault detection. Fault detection for current-source converters mainly relies on overcurrent protection and differential current criteria. For example, DC-side overcurrent protection determines a fault when the DC current exceeds a threshold, but this method cannot distinguish the specific fault location within the valve and is easily affected by system transient processes. Another example is AC-side differential protection, which detects faults by comparing the current difference on both sides of the converter transformer, but its sensitivity to local short circuits within the valve is insufficient. The fault current in current-source converter valves rises rapidly, and traditional methods, relying on fixed thresholds, are difficult to adapt to different fault degrees. Furthermore, relying solely on current signals, they cannot distinguish between internal valve faults and external system disturbances, easily leading to malfunctions. Fault detection for voltage-source converters typically employs submodule capacitor voltage monitoring and bridge arm energy balance methods, such as the submodule voltage deviation method, which determines faults by comparing the voltage of each submodule capacitor with theoretical values. However, this method requires extremely high measurement accuracy, and the signal change is not significant during high-resistance faults. For example, differential current protection for bridge arms detects current imbalance between the upper and lower bridge arms, but this imbalance may be masked by the control system's regulation. Fault characteristics in voltage source converter valves are relatively slow, and traditional voltage monitoring methods have a delayed response, hindering rapid protection. Furthermore, existing methods struggle to detect localized faults within the valve, easily leading to the accumulation and expansion of faults.

[0008] However, converters with a hybrid valve structure (current-source main valve + voltage-source support valve) exhibit more complex fault characteristics. Existing fault detection methods cannot distinguish between faults in current-source valves and voltage-source valves, making it impossible to accurately locate faults and leading to delayed or misjudged protection actions. Therefore, there is an urgent need to research a fault detection method specifically for converters with a hybrid valve structure. Summary of the Invention

[0009] To address the shortcomings of existing technologies, this invention provides a method and apparatus for detecting faults within converter valves, a converter system, and computer equipment.

[0010] The first aspect of this invention provides a method for detecting internal valve faults in a converter, wherein the converter employs a hybrid valve structure consisting of a current-source main valve and a voltage-source support valve, and the method includes:

[0011] A three-dimensional differential current detection method, including longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection, is adopted to obtain the current difference between the AC side and DC side of the current source type main valve, the unbalance of the current in each parallel branch of the current source type main valve, and the rate of change of the DC side current, respectively.

[0012] A three-level voltage detection method, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, was adopted to obtain the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope of the voltage source type support valve, respectively.

[0013] Based on the current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch and the rate of change of the DC side current, the bridge arm voltage deviation, the correlation coefficient of the bridge arm voltage sequence and the slope of the bridge arm voltage change of the voltage source type support valve, a current-voltage correlation analysis matrix is ​​constructed.

[0014] The combination patterns of current and voltage characteristics are analyzed using a current-voltage correlation analysis matrix. Based on these patterns, the fault conditions of the current-source main valve and the voltage-source support valve are determined.

[0015] In this embodiment of the invention, the three-dimensional differential current detection method, including longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection, is used to obtain the current difference between the AC and DC sides of the current source type main valve, the current imbalance of each parallel branch of the current source type main valve, and the rate of change of the DC side current, respectively. This includes:

[0016] The longitudinal differential current detection method is used to obtain the current difference between the AC side and the DC side of the current source type main valve;

[0017] The unbalance of current in each parallel branch of the current source type main valve is obtained by using the transverse differential current detection method.

[0018] The DC-side current change rate of the current source type main valve is obtained by using the time-domain differential current detection method.

[0019] In this embodiment of the invention, the method of obtaining the current difference between the AC and DC sides of the current source type main valve using a longitudinal differential current detection method includes:

[0020] Real-time acquisition of AC three-phase current and DC current;

[0021] The AC side loop current is calculated based on the three-phase current on the AC side. The difference between the AC side loop current and the DC side current is then calculated to obtain the current difference between the AC side and the DC side of the current source type main valve.

[0022] In this embodiment of the invention, the method of using transverse differential current detection to obtain the unbalance of the current in each parallel branch of the current source type main valve includes: real-time acquisition of the current in each parallel branch, calculation of the deviation between the current in each parallel branch and the average current value, and taking the deviation between the current in each parallel branch and the average current value as the unbalance of the current in each parallel branch.

[0023] In this embodiment of the invention, the method of using time-domain differential current detection to obtain the DC-side current change rate of the current source type main valve includes: performing differential processing on the real-time collected DC-side current to calculate the DC-side current change rate.

[0024] In this embodiment of the invention, the three-level voltage detection method, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, is used to obtain the bridge arm voltage deviation of the voltage source type support valve, the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve, and the slope of the bridge arm voltage change of the voltage source type support valve, respectively. This includes:

[0025] The voltage deviation of the bridge arm of the voltage source type support valve is obtained by using a voltage amplitude detection method.

[0026] The correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve is obtained by using the voltage waveform similarity detection method.

[0027] The slope of the bridge arm voltage change of the voltage source type support valve is obtained by using a voltage dynamic response detection method.

[0028] In this embodiment of the invention, the method of obtaining the bridge arm voltage deviation of the voltage source type support valve by means of voltage amplitude detection includes: measuring the voltage of each bridge arm of the voltage source type support valve in real time, calculating the deviation of each bridge arm voltage from the theoretical voltage, and obtaining the bridge arm voltage deviation of the voltage source type support valve.

[0029] In this embodiment of the invention, the step of using a voltage waveform similarity detection method to obtain the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve includes:

[0030] Set the sliding window length, slide the window according to the sliding window length, and extract the measured voltage sequence and theoretical voltage sequence within the current window;

[0031] Calculate the correlation coefficient between the measured voltage sequence and the theoretical voltage sequence, and use it as the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve.

[0032] In this embodiment of the invention, the method of using voltage dynamic response detection to obtain the slope of the bridge arm voltage change of the voltage source type support valve includes: performing high-frequency sampling on the bridge arm voltage of the voltage source type support valve, extracting the voltage change time, obtaining the voltage value corresponding to different change times, calculating the ratio of the difference between the voltage values ​​corresponding to two different change times to the difference between the change times, and obtaining the slope of the bridge arm voltage change.

[0033] In this embodiment of the invention, the current-voltage correlation analysis matrix is ​​constructed based on the current difference between the AC and DC sides of the current-source type main valve, the current imbalance of each parallel branch, and the rate of change of the DC side current; the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope of the voltage-source type support valve; including:

[0034] The current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch, the rate of change of the DC side current, and the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve are used as fault features to form a fault feature vector including the current difference between the AC and DC sides, the unbalance of the current in each parallel branch, the rate of change of the DC side current, the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope.

[0035] Based on the fault feature vectors, arrange them by column to construct an association matrix;

[0036] Weights are assigned to each fault feature in the correlation matrix to form a current-voltage correlation analysis matrix.

[0037] In this embodiment of the invention, assigning weights to each fault feature in the correlation matrix includes:

[0038] Fault simulation data was obtained by using a converter DC transmission simulation model.

[0039] Using fault simulation data as sample data, calculate the current difference between AC and DC sides, the unbalance of current in each parallel branch, the rate of change of DC current, the deviation of bridge arm voltage, the correlation coefficient of bridge arm voltage sequence, and the proportion of abnormal bridge arm voltage change slope in multiple sample data. Use this proportion as the weight of each fault feature.

[0040] Each fault feature in the correlation matrix is ​​assigned a corresponding weight.

[0041] In this embodiment of the invention, the fault condition of the current-source type main valve and the voltage-source type support valve is determined based on a combination of current and voltage characteristics, including:

[0042] In the combined mode of current and voltage characteristics, if both current and voltage characteristics show abnormalities, and the following conditions are met:

[0043] The fault type is determined to be a composite fault of the current-source type main valve and the voltage-source type support valve; where T is the mathematical transpose symbol, F is the fault feature vector, and F T Let M be the transposed fault feature vector, M be the correlation matrix, and K be the fault feature vector. composite The preset composite fault value;

[0044] If only the current characteristics are abnormal, the fault type is determined to be a single valve fault of the current source type main valve. If only the voltage characteristics are abnormal, the fault type is determined to be a single valve fault of the voltage source type support valve.

[0045] A second aspect of the present invention provides a converter valve internal fault detection device, wherein the converter adopts a hybrid valve structure of a current source type main valve and a voltage source type support valve, and the device includes:

[0046] The current detection unit is used to obtain the current difference between the AC side and the DC side of the current source type main valve, the unbalance of the current in each parallel branch of the current source type main valve, and the rate of change of the DC side current by using a three-dimensional differential current detection method including longitudinal differential current detection, transverse differential current detection and time-domain differential current detection.

[0047] The voltage detection unit is used to obtain the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope of the voltage source type support valve by employing a three-level voltage detection method including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection.

[0048] The fault analysis unit is used to construct a current-voltage correlation analysis matrix based on the current difference between the AC and DC sides of the current source type main valve, the current imbalance of each parallel branch, and the rate of change of the DC side current, as well as the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve. The current-voltage correlation analysis matrix is ​​used to analyze the combination pattern of current and voltage characteristics, and the fault status of the current source type main valve and the voltage source type support valve is determined based on the combination pattern of current and voltage characteristics.

[0049] A third aspect of the present invention provides a converter system, the converter system comprising: a converter valve and an in-valve fault detection device, the in-valve fault detection device being used to perform the above-described converter valve in-valve fault detection method.

[0050] A fourth aspect of the present invention provides a computer device, comprising: a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above-described converter valve internal fault detection method.

[0051] The converter valve internal fault detection method of this invention designs differentiated fault location strategies for current-source type main valves and voltage-source type support valves. By measuring the AC and DC currents of the current-source type main valve, multiple differential protection criteria are constructed. A three-dimensional differential current detection method (longitudinal differential, lateral differential, and time-domain differential) is used to obtain the current difference between the AC and DC sides of the current-source type main valve, the imbalance of the parallel branch current, and the current change rate, which are used to accurately locate faults within the main valve. A fault detection strategy is designed based on the deviation between the theoretical and measured values ​​of the bridge arm voltage of the voltage-source type support valve. A three-level voltage detection method (amplitude detection, waveform similarity detection, and dynamic response detection) is used to obtain the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage-source type support valve, which are used to locate support valve faults. By decoupling the fault characteristics of the current-source main valve and the voltage-source support valve, and employing a differentiated localization strategy, the problems of signal cross-interference and difficulty in fault localization caused by the hybrid valve structure (current-source main valve + voltage-source support valve) are effectively solved, significantly improving the accuracy of fault detection. Simultaneously, based on the detected various current and voltage characteristics, a current-voltage correlation analysis matrix is ​​constructed. This matrix is ​​used to analyze the combination patterns of current and voltage characteristics, collaboratively determining the fault status of the current-source main valve and the voltage-source support valve, thus achieving accurate detection of faults in hybrid valve structure converters.

[0052] Other features and advantages of the technical solution of the present invention will be described in detail in the following detailed embodiments section. Attached Figure Description

[0053] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0054] Figure 1 This is a flowchart of the converter valve internal fault detection method provided in the embodiments of the present invention;

[0055] Figure 2 This is an architecture diagram of the converter valve internal fault detection method provided in an embodiment of the present invention;

[0056] Figure 3 This is a block diagram of the converter valve internal fault detection device provided in an embodiment of the present invention. Detailed Implementation

[0057] To make the technical solutions and advantages of the embodiments of the present invention clearer, the exemplary embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not an exhaustive list of all embodiments. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0058] Meaning:

[0059] Longitudinal differential current detection: The fault range is determined by comparing the amplitude and phase differences of the current at each end of the protected equipment.

[0060] Lateral differential current detection: Fault location is achieved by comparing the current differences between different branches of the same phase.

[0061] Time-domain differential current detection: Internal faults are determined by comparing the differences in current waveforms of different branches or sides of the same device in real time.

[0062] Voltage amplitude detection is a key technology in power systems and electronic measurements. It extracts signal peak values ​​and monitors them in real time to achieve fault diagnosis.

[0063] Voltage waveform similarity detection is a key technology for power system fault diagnosis, equipment condition monitoring, and power quality analysis. Its core lies in quantifying the characteristic differences of different voltage waveforms.

[0064] Voltage dynamic response detection is a key technology for evaluating the voltage stability of power electronic devices (such as switching power supplies and DVR dynamic voltage regulators) under load changes or grid disturbances. Its core lies in quantifying the transient characteristics and recovery capability of voltage waveforms.

[0065] Figure 1 This is a flowchart of a converter valve internal fault detection method provided in an embodiment of the present invention. Figure 1 As shown, the converter valve internal fault detection method provided in this embodiment includes the following steps:

[0066] S100 employs a three-dimensional differential current detection method, including longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection, to obtain the current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch of the current source type main valve, and the rate of change of the DC side current, respectively.

[0067] S200 employs a three-level voltage detection method, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, to obtain the bridge arm voltage deviation of the voltage source type support valve, the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve, and the slope of the bridge arm voltage change of the voltage source type support valve, respectively.

[0068] S300, based on the current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch and the rate of change of the DC side current, the bridge arm voltage deviation, the correlation coefficient of the bridge arm voltage sequence and the slope of the bridge arm voltage change of the voltage source type support valve, a current-voltage correlation analysis matrix is ​​constructed.

[0069] The S400 uses a current-voltage correlation analysis matrix to analyze the combination patterns of current and voltage characteristics, and determines the fault status of the current source type main valve and the voltage source type support valve based on the combination patterns of current and voltage characteristics.

[0070] like Figure 2 As shown in the embodiment of the present invention, the converter valve internal fault detection method, for a hybrid valve structure of current-source main valve + voltage-source support valve, adopts a hierarchical detection and collaborative judgment approach. By real-time acquisition of key parameters such as AC side current, DC side current, and support valve arm voltage, a complete fault feature extraction and diagnosis system is constructed, including three functional modules: data acquisition, calculation and analysis, and decision-making. Each module adopts a parallel processing architecture to ensure the real-time performance and accuracy of fault detection. By decoupling the fault characteristics of the two types of valves (current-source valves rely on current differential, and voltage-source valves rely on voltage deviation), precise location is achieved. Simultaneously, multiple criteria are used to improve the detection sensitivity of hidden faults such as local short circuits within the valve.

[0071] For current-source type main valves, a multi-dimensional differential current detection method is adopted, including three detection dimensions: longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection. Longitudinal differential current detection identifies valve group-level faults by comparing the current difference between the AC side and the DC side; transverse differential current detection locates branch-level faults by analyzing the current imbalance between parallel branches; and time-domain differential current detection captures early fault characteristics by monitoring the rate of change of current.

[0072] A theoretical voltage dynamic calculation model was established for voltage source type support valves. Taking into account factors such as modulation strategy, DC voltage level and submodule status, a three-level voltage detection system including amplitude detection, waveform similarity detection and dynamic response detection was implemented to achieve comprehensive monitoring of support valve faults.

[0073] This embodiment employs a hybrid valve collaborative protection mechanism, establishing a current-voltage correlation analysis matrix. By analyzing the combination patterns of current and voltage characteristics, it can accurately distinguish between single valve faults and compound faults, achieving collaborative judgment of current-source main valve faults and voltage-source support valve faults. A graded action strategy is adopted, implementing rapid blocking for severe faults, precise isolation for local faults, and automatic initiation of redundant system switching, maximizing system continuity while ensuring safety.

[0074] In step S100 above, a longitudinal differential current detection method is used to obtain the current difference between the AC side and the DC side of the current source type main valve, which is used to identify valve group level faults; a transverse differential current detection method is used to obtain the current imbalance of each parallel branch of the current source type main valve, which is used to locate branch level faults; and a time-domain differential current detection method is used to obtain the DC side current change rate of the current source type main valve, which is used to detect early faults.

[0075] In a specific embodiment, the process of the longitudinal differential detection method is as follows:

[0076] (1) Real-time acquisition of three-phase current on the AC side i a ( t ), i b ( t ), i c ( t and DC side current i dc ( t );

[0077] (2) Based on the three-phase current on the AC side i a ( t ), i b ( t ), i c ( t Calculate the AC side loop current. i xy ( t (x and y represent a, b, and c, depending on the commutation circuit), calculate the AC side circuit current. i xy ( t ) and DC side current i dc ( t The difference Δ I long :

[0078] ;

[0079] (3) In the subsequent fault analysis process, determine the AC side loop current. i xy ( t ) and DC side current i dc ( t The difference Δ I long Is it greater than the dynamic threshold? Klong I n ( I n Rated current, K long (This is an adjustable coefficient, usually taken as 1.2~1.5). If Δ I long > K long I n If so, it is determined that there is a valve group-level fault in the current source type main valve.

[0080] In a specific embodiment, the process of the transverse differential current detection method is as follows:

[0081] (1) Real-time acquisition of current in each parallel branch i branch-1 , i branch-2 ,..., i branch-n ;

[0082] (2) Calculate the deviation Δ between the current of each parallel branch and the average current value. I branch :

[0083] ;

[0084] Where n represents the number of parallel branches, i branch_j , i branch_k These represent the currents of the j-th and k-th parallel branches, respectively; the deviation Δ of the current in each parallel branch from the average current value. I branch This indicates the degree of imbalance of current in parallel branches;

[0085] (3) In the subsequent fault analysis process, determine the deviation Δ between the current of any parallel branch and the average current value. I branch Is it greater than the imbalance deviation? K trans I avg ( I avg The average current, K trans (This is the imbalance coefficient, usually taken as 0.2~0.3). If any Δ I branch > K trans I avg If so, then the branch is determined to be faulty.

[0086] In a specific embodiment, the process of the time-domain differential current detection method is as follows:

[0087] (1) The DC side current acquired in real time i dc ( t Perform differentiation to calculate the rate of change of DC current:

[0088] ;

[0089] (2) In the subsequent fault analysis process, determine the rate of change of DC side current. Is it greater than the rate of change threshold? K rate (Set according to the system transient response characteristics), if If so, it is determined to be an early failure.

[0090] In addition, a dynamic threshold adjustment algorithm was designed for longitudinal differential detection, lateral differential detection and time-domain differential detection. This algorithm can dynamically adjust the threshold used to judge faults during subsequent fault analysis, automatically adapt to different operating conditions, and significantly improve detection sensitivity.

[0091] In step S200 above, the voltage amplitude detection method is used to obtain the bridge arm voltage deviation of the voltage source type support valve; the voltage waveform similarity detection method is used to obtain the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve; and the voltage dynamic response detection method is used to obtain the slope of the bridge arm voltage change of the voltage source type support valve.

[0092] In a specific embodiment, the process of obtaining the bridge arm voltage deviation of the voltage source type support valve using the amplitude detection method is as follows: real-time measurement of the voltage of each bridge arm of the voltage source type support valve. U arm Calculate the voltage of each bridge arm. U arm With theoretical voltage U ref deviation Δ U :

[0093] ;

[0094] In the subsequent fault analysis process, if it is determined that Δ U > K amp U ref ( K amp This is the amplitude tolerance coefficient. U ref If the voltage is the theoretical voltage, then it is determined to be an amplitude anomaly.

[0095] In a specific embodiment, the process of obtaining the correlation coefficient of the bridge arm voltage sequence using the waveform similarity detection method is as follows:

[0096] (1) Set the sliding window length, slide the window according to the sliding window length, and extract the measured voltage sequence within the current window. U meas and theoretical voltage sequence U theory ;

[0097] (2) Calculate the measured voltage sequence U meas With theoretical voltage sequence U theory correlation coefficient ρ :

[0098] ;

[0099] in, Represents the measured voltage sequence within the same sliding window. U meas standard deviation Represents the theoretical voltage sequence within the same sliding window. U theory Standard deviation, correlation coefficient ρ That is, the correlation coefficient of the bridge arm voltage sequence;

[0100] (3) In the subsequent fault analysis process, if (K) corr If the similarity threshold is set to 0.8 to 0.9, then waveform mismatch is determined, indicating a fault in the voltage source type support valve.

[0101] The waveform similarity detection algorithm employs a sliding window correlation analysis method, which can effectively distinguish between normal modulation fluctuations and true fault signals. For detected fault information, a voltage deviation sorting algorithm can accurately locate the specific fault submodule.

[0102] In a specific embodiment, the method for obtaining the slope of the bridge arm voltage change using the dynamic response detection method is as follows: The bridge arm voltage of the voltage source type support valve is sampled at high frequency to extract the moment of voltage abrupt change and obtain... t 1. t Voltage at time 2 U(t 1 )、U (t 2 ) Calculate two different mutation times t 1. t Voltage value corresponding to 2 U(t 1 ), U(t) 2 ) The difference between the time of mutation and the time of mutation (t 2- t The ratio of 1) gives the slope S of the bridge arm voltage change:

[0103] ;

[0104] In the subsequent fault analysis process, it is determined whether the voltage change slope S exceeds the preset slope threshold. S max If the slope S of the voltage change exceeds the preset threshold S max If so, it is determined that the voltage source type support valve has a transient fault.

[0105] The aforementioned dynamic response detection method solves the problem of missed detection of high-resistance faults by frequently sampling the bridge arm voltage of the voltage source type support valve, extracting the moment of voltage change, and introducing transient response analysis.

[0106] In step S300 above, the current difference between the AC and DC sides of the current source type main valve, the imbalance of the parallel branch current, the rate of change of the DC side current, and the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve are used as fault features to form a fault feature vector F:

[0107] ;

[0108] Where, Δ I long Δ is the current difference between the AC and DC sides. I branch For the unbalance of current in parallel branches, Let Δ be the rate of change of current. U For bridge arm voltage deviation, ρ is the correlation coefficient of the bridge arm voltage sequence, and S is the slope of the bridge arm voltage change;

[0109] Based on the fault feature vector F, arrange them by column to construct the correlation matrix M (6*6);

[0110] Weights are assigned to each fault feature in the correlation matrix M (6*6) to form a current-voltage correlation analysis matrix.

[0111] In a specific embodiment, weights are assigned to each fault feature in the correlation matrix. The specific method is as follows: fault simulation is performed using a converter DC transmission simulation model to obtain fault simulation data; the fault simulation data is used as sample data, and the current difference between the AC side and the DC side, the unbalance of the current in each parallel branch, the rate of change of the DC side current, the bridge arm voltage deviation, the correlation coefficient of the bridge arm voltage sequence, and the proportion of abnormal bridge arm voltage change slope are calculated in multiple sample data. This proportion is used as the weight of each fault feature, and corresponding weights are assigned to each fault feature in the correlation matrix.

[0112] In step S400 above, the combination pattern of current characteristics and voltage characteristics is analyzed using the current-voltage correlation analysis matrix. In the combination pattern of current characteristics and voltage characteristics, if both current characteristics and voltage characteristics show abnormalities, and the following conditions are met:

[0113] If the fault type is determined to be a composite fault of the current source type main valve and the voltage source type support valve; where T is the mathematical transpose symbol, F is the fault feature vector, and F T Let M be the transposed fault feature vector, M be the correlation matrix, and K be the fault feature vector. composite The preset composite fault value;

[0114] If only the current characteristics are abnormal, the fault type is determined to be a single valve fault of the current source type main valve; if only the voltage characteristics are abnormal, the fault type is determined to be a single valve fault of the voltage source type support valve.

[0115] The converter valve internal fault detection method of this invention designs differentiated fault location strategies for current-source type main valves and voltage-source type support valves. By measuring the AC and DC currents of the current-source type main valve, multiple differential protection criteria are constructed. A three-dimensional differential current detection method (longitudinal differential, lateral differential, and time-domain differential) is used to obtain the current difference between the AC and DC sides of the current-source type main valve, the imbalance of the parallel branch current, and the current change rate, which are used to accurately locate faults within the main valve. A fault detection strategy is designed based on the deviation between the theoretical and measured values ​​of the bridge arm voltage of the voltage-source type support valve. A three-level voltage detection method (amplitude detection, waveform similarity detection, and dynamic response detection) is used to obtain the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage-source type support valve, which are used to locate support valve faults. By decoupling the fault characteristics of the current-source main valve and the voltage-source support valve, and employing a differentiated localization strategy, the problems of signal cross-interference and difficulty in fault localization caused by the hybrid valve structure (current-source main valve + voltage-source support valve) are effectively solved, significantly improving the accuracy of fault detection. Simultaneously, based on various detected current and voltage characteristics, a current-voltage correlation analysis matrix is ​​constructed. This matrix is ​​used to analyze the combination patterns of current and voltage characteristics, collaboratively determining the fault conditions of the current-source main valve and the voltage-source support valve. This enables accurate differentiation between single valve faults and compound faults, achieving accurate detection of faults in hybrid valve structure converters.

[0116] Figure 3 This is a block diagram of a converter valve internal fault detection device provided in an embodiment of the present invention. Figure 3As shown, the converter valve internal fault detection device provided in this embodiment includes: a current detection unit, a voltage detection unit, and a fault analysis unit. The current detection unit employs a three-dimensional differential current detection method, including longitudinal differential current detection, lateral differential current detection, and time-domain differential current detection, to obtain the current difference between the AC and DC sides of the current-source type main valve, the current imbalance of each parallel branch of the current-source type main valve, and the DC side current change rate, respectively. The voltage detection unit employs a three-level voltage detection method, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, to obtain the bridge arm voltage deviation of the voltage-source type support valve, the correlation coefficient of the bridge arm voltage sequence of the voltage-source type support valve, and the slope of the bridge arm voltage change of the voltage-source type support valve, respectively. The fault analysis unit is used to construct a current-voltage correlation analysis matrix based on the current difference between the AC and DC sides of the current source type main valve, the current imbalance of each parallel branch, and the rate of change of the DC side current, as well as the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve. The current-voltage correlation analysis matrix is ​​used to analyze the combination pattern of current characteristics and voltage characteristics, and the fault conditions of the current source type main valve and the voltage source type support valve are determined based on the combination pattern of current characteristics and voltage characteristics.

[0117] In a specific embodiment, the current detection unit employs a longitudinal differential detection method to obtain the current difference between the AC side and the DC side of the current source type main valve, which is used to identify valve group level faults; it employs a transverse differential detection method to obtain the current imbalance of each parallel branch of the current source type main valve, which is used to locate branch level faults; and it employs a time-domain differential detection method to obtain the DC side current change rate of the current source type main valve, which is used to detect early faults.

[0118] The process of the longitudinal differential detection method is as follows:

[0119] (1) Real-time acquisition of three-phase current on the AC side i a ( t ), i b ( t ), i c ( t and DC side current i dc ( t );

[0120] (2) Based on the three-phase current on the AC side i a ( t ), i b ( t ), i c (t Calculate the AC side loop current. i xy ( t (x and y represent a, b, and c, depending on the commutation circuit), calculate the AC side circuit current. i xy ( t ) and DC side current i dc ( t The difference Δ I long :

[0121] ;

[0122] (3) In the subsequent fault analysis process, the fault analysis unit determines the AC side loop current. i xy ( t ) and DC side current i dc ( t The difference Δ I long Is it greater than the dynamic threshold? K long I n ( I n Rated current, K long (This is an adjustable coefficient, usually taken as 1.2~1.5). If Δ I long > K long I n If so, it is determined that there is a valve group-level fault in the current source type main valve.

[0123] The process of the lateral differential detection method is as follows:

[0124] (1) Real-time acquisition of current in each parallel branch i branch1 , i branch2 ,..., i branchn ;

[0125] (2) Calculate the deviation Δ between the current of each parallel branch and the average current value. I branch :

[0126] ;

[0127] Where n represents the number of parallel branches, ibranch_j , i branch_k These represent the currents of the j-th and k-th parallel branches, respectively; the deviation Δ of the current in each parallel branch from the average current value. I branch This indicates the degree of imbalance of current in parallel branches;

[0128] (3) In the subsequent fault analysis process, the fault analysis unit judges the deviation Δ between the current of any parallel branch and the average current value. I branch Is it greater than the imbalance deviation? K trans I avg ( I avg The average current, K trans (This is the imbalance coefficient, usually taken as 0.2~0.3). If any Δ I branch > K trans I avg If so, then the branch is determined to be faulty.

[0129] The process of the time-domain differential detection method is as follows:

[0130] (1) The DC side current acquired in real time i dc ( t Perform differentiation to calculate the rate of change of DC current:

[0131] ;

[0132] (2) In the subsequent fault analysis process, the fault analysis unit judges the rate of change of current. Is it greater than the rate of change threshold? K rate (Set according to the system transient response characteristics), if If so, it is determined to be an early failure.

[0133] In a specific embodiment, the voltage detection unit uses an amplitude detection method to obtain the bridge arm voltage deviation of the voltage source type support valve; uses a waveform similarity detection method to obtain the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve; and uses a dynamic response detection method to obtain the slope of the bridge arm voltage change of the voltage source type support valve.

[0134] The amplitude detection method involves: real-time measurement of the voltage of each arm of the voltage source type support valve. U arm Calculate the voltage of each bridge arm. U armWith theoretical voltage U ref deviation Δ U :

[0135] ;

[0136] Subsequently, the fault analysis unit performs fault analysis. If it is determined that Δ U > K amp U ref ( K amp This is the amplitude tolerance coefficient. U ref If the voltage is the theoretical voltage, then it is determined to be an amplitude anomaly.

[0137] The waveform similarity detection process is as follows:

[0138] (1) Set the sliding window length, slide the window according to the sliding window length, and extract the measured voltage sequence within the current window. U meas and theoretical voltage sequence U theory ;

[0139] (2) Calculate the measured voltage sequence U meas With theoretical voltage sequence U theory correlation coefficient ρ :

[0140] ;

[0141] in, Represents the measured voltage sequence within the same sliding window. U meas standard deviation Represents the theoretical voltage sequence within the same sliding window. U theory Standard deviation, correlation coefficient ρ That is, the correlation coefficient of the bridge arm voltage sequence;

[0142] (3) Subsequently, the fault analysis unit performs fault analysis. If (K) corr If the similarity threshold is set to 0.8 to 0.9, then waveform mismatch is determined, indicating a fault in the voltage source type support valve.

[0143] The dynamic response detection method specifically involves: performing high-frequency sampling of the bridge arm voltage of the voltage source type support valve, extracting the moment of voltage abrupt change, and obtaining... t 1. t Voltage at time 2 U(t1 ), U(t) 2 ) Calculate two different mutation times t 1. t Voltage value corresponding to 2 U(t 1 ), U(t) 2 ) The difference between the time of mutation and the time of mutation ( t 2- t The ratio of 1) gives the slope S of the bridge arm voltage change:

[0144] ;

[0145] Subsequently, the fault analysis unit performs fault analysis to determine whether the voltage change slope S exceeds a preset slope threshold. S max If the slope S of the voltage change exceeds the preset threshold S max If so, it is determined that the voltage source type support valve has a transient fault.

[0146] In a specific embodiment, the process of the fault analysis unit constructing the current-voltage correlation analysis matrix is ​​as follows: the current difference between the AC and DC sides of the current source type main valve, the unbalance of the parallel branch current, the rate of change of current, and the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve are used as fault features to form a fault feature vector F:

[0147] ;

[0148] Where, Δ I long Δ is the current difference between the AC and DC sides. I branch For the unbalance of current in parallel branches, Let Δ be the rate of change of current. U For bridge arm voltage deviation, ρ is the correlation coefficient of the bridge arm voltage sequence, and S is the slope of the bridge arm voltage change;

[0149] Based on the fault feature vector F, arrange them by column to construct the correlation matrix M (6*6);

[0150] Weights are assigned to each fault feature in the correlation matrix M (6*6) to form a current-voltage correlation analysis matrix.

[0151] Specifically, weights are assigned to each fault feature in the correlation matrix. The method is as follows: fault simulation is performed using a converter DC transmission simulation model to obtain fault simulation data; the fault simulation data is used as sample data, and the current difference between the AC side and the DC side, the unbalance of the current in each parallel branch, the rate of change of the DC side current, the arm voltage deviation, the correlation coefficient of the arm voltage sequence, and the proportion of abnormal arm voltage change slope are calculated in multiple sample data. This proportion is used as the weight of each fault feature, and corresponding weights are assigned to each fault feature in the correlation matrix.

[0152] In a specific embodiment, the fault analysis unit uses a current-voltage correlation analysis matrix to analyze the combination pattern of current characteristics and voltage characteristics. In the combination pattern of current characteristics and voltage characteristics, if it is determined that both current characteristics and voltage characteristics are abnormal at the same time, and the following conditions are met:

[0153] If the fault type is determined to be a composite fault of the current source type main valve and the voltage source type support valve; where T is the mathematical transpose symbol, F is the fault feature vector, and F T Let M be the transposed fault feature vector, M be the correlation matrix, and K be the fault feature vector. composite The preset composite fault value;

[0154] If it is determined that only the current characteristics are abnormal, the fault type is determined to be a single valve fault of the current source type main valve; if only the voltage characteristics are abnormal, the fault type is determined to be a single valve fault of the voltage source type support valve.

[0155] The converter valve internal fault detection device in this embodiment has significant advantages over existing technologies: First, by decoupling the fault characteristics of the current-source main valve and the voltage-source support valve, a differentiated location strategy is adopted, effectively solving the signal cross-interference problem caused by the hybrid valve structure (current-source main valve + voltage-source support valve), and greatly improving the accuracy of fault detection. Second, a multi-criteria collaborative detection mechanism is designed. For the current-source valve, multiple differential protection on the AC / DC side is constructed, significantly improving the sensitivity to hidden faults such as local short circuits; for the voltage-source valve, a dynamic slope detection method is introduced to capture the voltage change characteristics of the bridge arm. In addition, the solution makes full use of the real-time data of the existing measurement system, and achieves synchronous optimization of fault location speed and protection action reliability by parallel calculation of differential current and voltage deviation. This technology can be adapted to the hybrid valve characteristics of the F3C converter without modifying the hardware architecture, and provides a highly reliable fault protection solution for new energy application scenarios.

[0156] This invention also provides a converter system, which includes a converter valve and an in-valve fault detection device, wherein the in-valve fault detection device is used to perform the above-described converter valve in-valve fault detection method.

[0157] This invention also provides a computer device, including: a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the above-described converter valve internal fault detection method.

[0158] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0159] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0160] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0162] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention. Furthermore, it should be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. As long as such combination does not violate the spirit of the embodiments of the present invention, it should also be considered as the content disclosed in the embodiments of the present invention.

Claims

1. A method for detecting internal faults in a converter valve, wherein the converter adopts a hybrid valve structure of a current-source type main valve and a voltage-source type support valve, characterized in that, The method includes: A three-dimensional differential current detection method, including longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection, is adopted to obtain the current difference between the AC side and DC side of the current source type main valve, the unbalance of the current in each parallel branch of the current source type main valve, and the rate of change of the DC side current, respectively. A three-level voltage detection method, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, was adopted to obtain the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope of the voltage source type support valve, respectively. Based on the current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch and the rate of change of the DC side current, the bridge arm voltage deviation, the correlation coefficient of the bridge arm voltage sequence and the slope of the bridge arm voltage change of the voltage source type support valve, a current-voltage correlation analysis matrix is ​​constructed. The combination patterns of current and voltage characteristics are analyzed using a current-voltage correlation analysis matrix. Based on these patterns, the fault conditions of the current-source main valve and the voltage-source support valve are determined.

2. The converter valve internal fault detection method according to claim 1, characterized in that, The method employs a three-dimensional differential current detection approach, including longitudinal differential current detection, transverse differential current detection, and time-domain differential current detection, to obtain the current difference between the AC and DC sides of the current-source type main valve, the current imbalance of each parallel branch of the current-source type main valve, and the rate of change of the DC side current, including: The longitudinal differential current detection method is used to obtain the current difference between the AC side and the DC side of the current source type main valve; The unbalance of current in each parallel branch of the current source type main valve is obtained by using the transverse differential current detection method. The DC-side current change rate of the current source type main valve is obtained by using the time-domain differential current detection method.

3. The converter valve internal fault detection method according to claim 2, characterized in that, The method of obtaining the current difference between the AC and DC sides of the current source type main valve using the longitudinal differential current detection method includes: Real-time acquisition of AC three-phase current and DC current; The AC side loop current is calculated based on the three-phase current on the AC side. The difference between the AC side loop current and the DC side current is then calculated to obtain the current difference between the AC side and the DC side of the current source type main valve.

4. The converter valve internal fault detection method according to claim 2, characterized in that, The method of using transverse differential current detection to obtain the current imbalance of each parallel branch of the current source type main valve includes: The current of each parallel branch is collected in real time, and the deviation between the current of each parallel branch and the average current value is calculated. The deviation between the current of each parallel branch and the average current value is used as the unbalance of the current of each parallel branch.

5. The converter valve internal fault detection method according to claim 2, characterized in that, The method of using time-domain differential current detection to obtain the rate of change of DC current on the current source type main valve includes: The DC-side current, acquired in real time, is differentiated to calculate the rate of change of the DC-side current.

6. The converter valve internal fault detection method according to claim 1, characterized in that, The method employs a three-level voltage detection approach, including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection, to obtain the arm voltage deviation, the arm voltage sequence correlation coefficient, and the arm voltage change slope of the voltage source type support valve, respectively. The voltage deviation of the bridge arm of the voltage source type support valve is obtained by using a voltage amplitude detection method. The correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve is obtained by using the voltage waveform similarity detection method. The slope of the bridge arm voltage change of the voltage source type support valve is obtained by using a voltage dynamic response detection method.

7. The converter valve internal fault detection method according to claim 6, characterized in that, The method of obtaining the bridge arm voltage deviation of the voltage source type support valve using voltage amplitude detection includes: The voltage of each arm of the voltage source type support valve is measured in real time, and the deviation between each arm voltage and the theoretical voltage is calculated to obtain the arm voltage deviation of the voltage source type support valve.

8. The converter valve internal fault detection method according to claim 6, characterized in that, The method of using voltage waveform similarity detection to obtain the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve includes: Set the sliding window length, slide the window according to the sliding window length, and extract the measured voltage sequence and theoretical voltage sequence within the current window; Calculate the correlation coefficient between the measured voltage sequence and the theoretical voltage sequence, and use it as the correlation coefficient of the bridge arm voltage sequence of the voltage source type support valve.

9. The converter valve internal fault detection method according to claim 6, characterized in that, The method of obtaining the slope of the bridge arm voltage change of the voltage source type support valve by using a voltage dynamic response detection method includes: The bridge arm voltage of the voltage source type support valve is sampled at high frequency to extract the voltage change time. The voltage value corresponding to different change times is obtained. The ratio of the difference between the voltage values ​​corresponding to two different change times to the difference between the change times is calculated to obtain the slope of the bridge arm voltage change.

10. The converter valve internal fault detection method according to claim 1, characterized in that, The current-voltage correlation analysis matrix is ​​constructed based on the current difference between the AC and DC sides of the current-source type main valve, the current imbalance of each parallel branch, and the rate of change of the DC side current; and the arm voltage deviation, arm voltage sequence correlation coefficient, and arm voltage change slope of the voltage-source type support valve. This matrix includes: The current difference between the AC and DC sides of the current source type main valve, the unbalance of the current in each parallel branch, the rate of change of the DC side current, and the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve are used as fault features to form a fault feature vector including the current difference between the AC and DC sides, the unbalance of the current in each parallel branch, the rate of change of the DC side current, the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope. Based on the fault feature vectors, arrange them by column to construct an association matrix; Weights are assigned to each fault feature in the correlation matrix to form a current-voltage correlation analysis matrix.

11. The converter valve internal fault detection method according to claim 10, characterized in that, Assigning weights to each fault feature in the correlation matrix includes: Fault simulation data was obtained by using a converter DC transmission simulation model. Using fault simulation data as sample data, calculate the current difference between AC and DC sides, the unbalance of current in each parallel branch, the rate of change of DC current, the deviation of bridge arm voltage, the correlation coefficient of bridge arm voltage sequence, and the proportion of abnormal bridge arm voltage change slope in multiple sample data. Use this proportion as the weight of each fault feature. Each fault feature in the correlation matrix is ​​assigned a corresponding weight.

12. The converter valve internal fault detection method according to claim 1, characterized in that, The method for determining the fault status of the current-source type main valve and the voltage-source type support valve based on a combination of current and voltage characteristics includes: In the combined mode of current and voltage characteristics, if both current and voltage characteristics show abnormalities, and the following conditions are met: The fault type is determined to be a composite fault of the current-source type main valve and the voltage-source type support valve; where T is the mathematical transpose symbol, F is the fault feature vector, and F T Let M be the transposed fault feature vector, M be the correlation matrix, and K be the fault feature vector. composite The preset composite fault value; If only the current characteristics are abnormal, the fault type is determined to be a single valve fault of the current source type main valve. If only the voltage characteristics are abnormal, the fault type is determined to be a single valve fault of the voltage source type support valve.

13. A fault detection device for a converter valve, wherein the converter adopts a hybrid valve structure of a current source type main valve and a voltage source type support valve, characterized in that, The device includes: The current detection unit is used to obtain the current difference between the AC side and the DC side of the current source type main valve, the unbalance of the current in each parallel branch of the current source type main valve, and the rate of change of the DC side current by using a three-dimensional differential current detection method including longitudinal differential current detection, transverse differential current detection and time-domain differential current detection. The voltage detection unit is used to obtain the bridge arm voltage deviation, the bridge arm voltage sequence correlation coefficient, and the bridge arm voltage change slope of the voltage source type support valve by employing a three-level voltage detection method including voltage amplitude detection, voltage waveform similarity detection, and voltage dynamic response detection. The fault analysis unit is used to construct a current-voltage correlation analysis matrix based on the current difference between the AC and DC sides of the current source type main valve, the current imbalance of each parallel branch, and the rate of change of the DC side current, as well as the bridge arm voltage deviation, bridge arm voltage sequence correlation coefficient, and bridge arm voltage change slope of the voltage source type support valve. The current-voltage correlation analysis matrix is ​​used to analyze the combination pattern of current and voltage characteristics, and the fault status of the current source type main valve and the voltage source type support valve is determined based on the combination pattern of current and voltage characteristics.

14. A converter system, characterized in that, The converter system includes a converter valve and an in-valve fault detection device, wherein the in-valve fault detection device is used to perform the converter valve in-valve fault detection method according to any one of claims 1-12.

15. A computer device, characterized in that, include: Memory, which stores computer programs; A processor for executing the computer program to implement the converter valve fault detection method according to any one of claims 1-12.

Images