A Method and System for Rapid Fault Removal of UPFC Sub-Modules

Abstract

A method and system for quickly removing faults in a UPFC sub-module, including: performing intrinsic modeling on the UPFC sub-module circuit to calculate the DC bus voltage, AC port current, and IGBT junction temperature; calculating the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature; determining whether there is an abnormality based on the absolute value residuals. If there is an abnormality, obtain the intrinsic parameters of the circuit fault, match the intrinsic parameters of the circuit fault with a preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result; set a static priority, and calculate a dynamic priority in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; according to the dynamic priority, take corresponding fault removal for different fault types of different sub-modules. The present invention has better performance in terms of response speed, multi-type fault identification, and system adaptability.

Description

Technical Field

[0001] The present invention belongs to the technical field of electrical engineering, and particularly relates to a method and system for quickly removing faults in UPFC sub-modules. Background Art

[0002] As the most comprehensive device in the flexible AC transmission system, the Unified Power Flow Controller (UPFC) can independently regulate the active power, reactive power of the line and the node voltage, improving the power transmission capacity and stability of the power grid. However, its complex sub-module structure and high-frequency operation characteristics pose severe challenges to the design of control and protection strategies, especially in terms of rapid fault identification and dynamic response, where there are technical bottlenecks.

[0003] Existing UPFC projects mostly adopt complex redundant configuration protection strategies at the system level. By adding a large number of redundant devices to operate in parallel and using a "two-out-of-three" logic to improve reliability; however, such solutions usually target system-level control and protection strategies, requiring a large amount of economic investment for construction. Moreover, the static priority arbitration adopted is prone to response delays due to logical conflicts in the case of multiple concurrent faults, making it difficult to meet the requirements of complex fault scenarios; Patent CN118885879A proposed a fault prediction method for UPFC devices based on wavelet neural networks, including: defining the fault variables of the unified power flow controller; classifying the fault variables and defining the data acquisition structure for device status monitoring and fault prediction; obtaining the signal data of each node of the unified power flow controller, filtering, A / D conversion, and then classifying and encoding it; constructing a device fault prediction model using wavelet neural networks; inputting the real-time signal data into the trained device fault prediction model to obtain the device status prediction result output by the device fault prediction model; judging whether there is a device fault according to the device status prediction result; sending a fault warning signal for the predicted device fault. However, single faults still require training, with low processing efficiency; and it can only diagnose and predict, without subsequent fault handling methods, processing priority determination, etc. Moreover, the object is the overall system and valve level, not involving sub-modules. Summary of the Invention

[0004] In view of this, the present invention provides a method and system for quickly removing faults in UPFC sub-modules. By means of circuit intrinsic modeling and filtering transient interference, combined with a BP neural network to improve the accuracy of composite fault diagnosis, and introducing dynamic weight distribution to optimize the fault arbitration logic, rapid removal in multiple fault scenarios is achieved. Compared with the prior art, it has significant advantages in terms of response speed, diagnosis accuracy, and system adaptability.

[0005] To solve the above technical problems, the present invention adopts the following technical solutions:

[0006] The first aspect of the present invention proposes a method for quickly removing faults in UPFC sub-modules, which is characterized by including:

[0007] Perform intrinsic modeling of the UPFC sub-module circuit, and calculate the DC bus voltage, AC port current, and IGBT junction temperature based on the modeling and the UPFC sub-module control strategy;

[0008] Calculate the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature through the real-time collected DC bus voltage, AC port current, and IGBT junction temperature and their corresponding calculated values; all IGBTs of each UPFC sub-module are located on the same substrate, and the measured IGBT junction temperature is the substrate temperature;

[0009] Judge whether there is an abnormality according to the absolute value residuals. If there is an abnormality, obtain the intrinsic parameters of the circuit fault, match the intrinsic parameters of the circuit fault with a preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result;

[0010] Set static priorities, and calculate dynamic priorities in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; the static priorities are the priorities of different fault types of the same sub-module, and the dynamic priorities are the priorities of different fault types of different sub-modules;

[0011] According to the dynamic priorities, perform corresponding fault removal for different fault types of different sub-modules.

[0012] Preferably, according to the intrinsic modeling of the UPFC sub-module circuit and the UPFC sub-module control strategy, establish mathematical formulas for the DC bus voltage, AC port current, and IGBT junction temperature:

[0013]

[0014] In the above formula, is the calculated value of the DC bus voltage at t time, is the calculated value of the AC port current at t time, is the calculated value of the IGBT junction temperature at t time, S bridge is the switching function of the switching tubes of the bridge arm; I dc is the sum of the DC currents input from the DC side to the three-phase bridge arm; ω 1 is the angular frequency of the AC side voltage; I S and are respectively the amplitude and phase of the AC side current of the converter in the UPFC; MIs / 4, are respectively the amplitude and phase of the circulating current second harmonic; M is the modulation ratio, is the switching function of the switch tube of the sub-module, C is the DC bus capacitor, is the DC bus voltage, is the number of sub-modules in the converter bridge arm of the UPFC, is the equivalent series resistance of the DC bus capacitor, T a ( t ) is the IGBT substrate temperature measured at the moment of t , P thIGBT is the IGBT thermal loss, Z thIGBT is the IGBT thermal resistance network.

[0015] Preferably, the switching function of the switch tube of the bridge arm S bridge and the switching function of the switch tube of the sub-module are obtained through the control strategy adopted by the UPFC sub-module. The control strategy adopted by the UPFC sub-module includes the sub-module voltage equalization control strategy and the double-frequency circulating current suppression strategy; among them, the double-frequency circulating current suppression strategy outputs a modulation wave through carrier phase shift modulation.

[0016] Preferably, judging whether there is an abnormality according to the absolute value residual is specifically:

[0017] Record the absolute value residuals of each sampling point of the collected DC bus voltage, AC port current, and IGBT junction temperature with a set sampling window, and calculate the mean and standard deviation of all absolute value residuals of each sampling window of the DC bus voltage, AC port current, or IGBT junction temperature;

[0018] If the absolute value of the difference between the absolute value residual of the DC bus voltage, AC port current, or IGBT junction temperature at a sampling point and the mean of the DC bus voltage measurement value, AC port current measurement value, or IGBT junction temperature measurement value of its corresponding sampling window is greater than 3 times the standard deviation of the DC bus voltage measurement value, AC port current measurement value, or IGBT junction temperature measurement value of the corresponding sampling window, it is determined that the UPFC sub-module is abnormal.

[0019] Preferably, the circuit fault eigen-parameters include the DC bus voltage measurement value, DC bus voltage deviation rate, AC port current measurement value, AC port current change rate di arm / dt , IGBT junction temperature measurement value, and IGBT junction temperature gradient dTj / dt ;

[0020] Among them, the DC bus voltage deviation rate DVR Uc、 AC port current change rate di arm / dt、 IGBT junction temperature gradient dT j / dt The calculation formula is:

[0021]

[0022] In the above formula, u c,m ( t ) is the measured value of the DC bus voltage at time t , is the calculated value of the DC bus voltage at time t , U c,rated is the rated value of the DC bus voltage, I arm,m ( t ) and I arm,m ( t -1) are respectively the measured values of the AC port current at time t and t -1, T j,m ( t ) and T j,m ( t -1) are respectively the measured values of the IGBT junction temperature at time t and t -1, is the set sampling window.

[0023] Preferably, the matching of the circuit fault eigen-parameters with the preset abnormal feature library and the multi-level fault screening and identification according to the matching results are specifically as follows:

[0024] First, perform the first-level fault screening, record the circuit fault eigen-parameters within a sampling window after determining the abnormality, form a vector with the DC bus voltage, DC bus voltage deviation rate, and AC port current among them, and calculate the cosine similarity between it and the first abnormal reference vectors corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that there is a corresponding fault; if all cosine similarities are less than or equal to the first threshold and greater than the second threshold, use the second-level fault screening; if all cosine similarities are less than or equal to the second threshold, use the third-level fault screening;

[0025] The second - layer fault screening is to widen the sampling window, form a vector with all the intrinsic parameters of circuit faults, and calculate the cosine similarity between it and the second abnormal reference vectors corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that the corresponding fault exists;

[0026] The third - layer fault screening is to input the intrinsic parameters of circuit faults of adjacent modules into a pre - trained BP neural network and output the probability of each fault. If the probability exceeds the set third threshold, it is considered that the corresponding fault exists;

[0027] The first abnormal reference vectors corresponding to the various faults are vectors composed of the historical data of the DC bus voltage, the DC bus voltage deviation rate, and the AC port current within a set sampling window after various faults occur; the second abnormal reference vectors corresponding to the various faults are vectors composed of the historical data of all the intrinsic parameters of circuit faults within the widened sampling window after various faults occur obtained through historical data.

[0028] Preferably, the static priorities of capacitor over - voltage fault, IGBT short - circuit fault, and IGBT over - heat fault are set to 80, 70, and 60 respectively.

[0029] Preferably, the calculation of the dynamic priority , specifically:

[0030]

[0031] In the above formula, P static is the set static priority, α , β and γ are the weight coefficients of the DC bus voltage, IGBT junction temperature, and AC port current respectively, U c,m ( t ) is the measured value of the DC bus voltage at time t , U c,rated is the rated value of the DC bus voltage, is t the measured value of the IGBT junction temperature at time is the maximum value of the IGBT junction temperature, is t the measured value of the AC port current at time, is the maximum value of the AC port current.

[0032] Preferably, according to the dynamic priority, corresponding fault removal is taken for different fault types, specifically:

[0033] When a single fault occurs, if it is an IGBT short - circuit fault, the drive of the sub - module is blocked and the bypass switch is triggered to put the redundant module into operation; if it is a capacitor over - voltage fault, the faulty module is removed and the redundant unit is put into operation, and the bridge - arm level output stability is maintained through the modulation control algorithm; if it is an IGBT over - heat fault, the redundant unit is put into operation or the output of the sub - module is reduced.

[0034] If multiple faults occur, an excision sequence is generated from high to low according to the dynamic priority, the faulty sub - modules are isolated in sequence and the redundant modules are put into use. When the redundant modules are insufficient, the converter is triggered to stop operating emergently.

[0035] The second aspect of the present invention provides a fast fault excision system for a UPFC sub - module using the method described in the first aspect of the present invention, including a circuit physical intrinsic modeling module, an abnormal state identification module, and a fault processing module, characterized in that:

[0036] The circuit physical intrinsic modeling module: is used for performing the circuit physical intrinsic modeling of the UPFC sub - module, and calculating the DC bus voltage, the AC port current, and the IGBT junction temperature.

[0037] The abnormal state identification module: is used for calculating the absolute - value residuals of the DC bus voltage, the AC port current, and the IGBT junction temperature through the real - time collected DC bus voltage, AC port current, IGBT junction temperature and their corresponding calculated values; all the IGBTs of each UPFC sub - module are located on the same substrate, and the measured IGBT junction temperature is the substrate temperature; it is judged whether there is an abnormality according to the absolute - value residuals. If there is an abnormality, the circuit fault intrinsic parameters are obtained, the circuit fault intrinsic parameters are matched with a preset abnormal feature library, and multi - level fault screening and identification are carried out according to the matching result.

[0038] The fault processing module: is used for setting the static priority and calculating the dynamic priority in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; the static priority is the priority of different fault types of the same sub - module, and the dynamic priority is the priority of different fault types of different sub - modules; according to the dynamic priority, corresponding fault excision is taken for different fault types of different sub - modules.

[0039] The beneficial effects of the present invention compared with the prior art are as follows: Based on the control strategy and circuit physical intrinsic modeling, the DC bus voltage, AC port current, and IGBT junction temperature are calculated; according to the residual between the calculated value and the collected value, it is judged whether there is an abnormality. If there is an abnormality, the circuit fault intrinsic parameters are obtained, and the circuit fault intrinsic parameters are matched with the preset abnormal feature library, and multi-level fault screening and identification are performed according to the matching result; it can quickly and accurately locate the fault type, reduce misjudgment and missed judgment. At the same time, through the multi-level screening mechanism, the comprehensiveness and reliability of fault diagnosis are effectively improved, providing a more targeted basis for fault handling. Set the static priority, and calculate the dynamic priority in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature, and perform corresponding fault removal for different fault types of different sub-modules to ensure the priority protection of key sub-modules, thereby maximizing the system stability; it can effectively avoid over-cutting or under-cutting, and improve the accuracy and efficiency of fault handling. The present invention has better performance than the traditional method in terms of response speed, multi-type fault identification, and system adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the UPFC sub-module system, control strategy, and fault fast removal system of the present invention;

[0041] Figure 2 It is a schematic diagram of the voltage equalization control strategy of the UPFC sub-module in the embodiment of the present invention;

[0042] Figure 3 It is a schematic diagram of the double-frequency circulating current suppression strategy in the embodiment of the present invention;

[0043] Figure 4 It is a BP network topology diagram;

[0044] Figure 5 It is a flowchart of the corresponding fault removal for different fault types of the present invention;

[0045] Figure 6 It is a waveform diagram of the fault occurrence and removal using the fault fast removal method disclosed in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0047] Embodiment 1 of the present invention proposes a method for quickly removing UPFC sub-module faults, including:

[0048] Perform circuit intrinsic modeling of the UPFC sub-module, and calculate the DC bus voltage, AC port current, and IGBT junction temperature;

[0049] Calculate the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature and their corresponding calculated values by collecting the DC bus voltage, AC port current, and IGBT junction temperature in real time; all IGBTs of each UPFC sub-module are located on the same substrate, and the measured IGBT junction temperature is the temperature of the substrate;

[0050] Judge whether there is an abnormality according to the absolute value residuals. If there is an abnormality, obtain the intrinsic parameters of the circuit fault, match the intrinsic parameters of the circuit fault with the preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result;

[0051] Set a static priority, and calculate a dynamic priority in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; the static priority is the priority of different fault types of the same sub-module, and the dynamic priority is the priority of different fault types of different sub-modules;

[0052] According to the dynamic priority, perform corresponding fault removal on different fault types of different sub-modules.

[0053] It should be noted that as Figure 1 shown, the UPFC sub-module includes a first switch tube S1, a second switch tube S2, a third switch tube S3, a fourth switch tube S4, a positive AC port AC_P, a negative AC port AC_N, a positive DC port DC_P, a negative DC port DC_N, a DC bus capacitor C, a voltage sensor MV, a current sensor MC, a temperature sensor MT, a sensor Ethernet interface Eth. The first switch tube and the second switch tube are connected in series to form a first branch, the third switch tube and the fourth switch tube are connected in series to form a second branch, the first branch and the second branch are connected in parallel, the positive AC port is connected to the midpoint of the first branch, the negative AC port is connected to the midpoint of the second branch, the positive DC port is connected to the positive end of the first branch, the positive end of the second branch, and the positive end of the DC bus capacitor, and the negative DC port is connected to the negative end of the first branch, the negative end of the second branch, and the negative end of the DC bus capacitor. The current sensor MC measures the arm current input at the positive AC port i arm , and the voltage sensor MV measures the DC bus voltage u c , and the temperature sensor MT is connected to the thermistor of the switching device to measure the IGBT junction temperature T j , and the sampling signals of the three sensors transmit the three sensing signals to the main control unit through Ethernet;

[0054] According to the intrinsic modeling of the UPFC sub-module circuit and the control strategy of the UPFC sub-module, establish mathematical formulas for the DC bus voltage, AC port current, and IGBT junction temperature:

[0055]

[0056] In the above formula, is t the calculated value of the DC bus voltage at the moment, is t the calculated value of the AC port current at the moment, is t the calculated value of the IGBT junction temperature at the moment, S bridge is the switching function of the switching device of the bridge arm; I dc is the sum of the DC currents input from the DC side to the three-phase bridge arm; ω 1 is the angular frequency of the AC side voltage; I S and are respectively the amplitude and phase of the AC side current of the converter in the UPFC; MI s / 4, are respectively the amplitude and phase of the second harmonic of the circulating current; M is the modulation ratio, is the switching function of the switching device of the sub-module, C is the DC bus capacitor, is the DC bus voltage, is the number of sub-modules of the converter bridge arm in the UPFC, is the equivalent series resistance of the DC bus capacitor, T a ( t ) is the IGBT substrate temperature measured at the moment of t , P thIGBT is the IGBT thermal loss, Z thIGBT is the IGBT thermal resistance network;

[0057] Obtained according to the actual topological parameters of the UPFC , , C, measure I dc , I S , ω 1, , , obtained according to the control strategy of the UPFC sub-module S bridge , , M ; Input these values into the above formula to obtain the calculated values of the DC bus voltage, AC port current and IGBT junction temperature.

[0058] The switching function of the switching device of the bridge arm Sbridge Switching function of the sub-module Obtained through the control strategy adopted by the UPFC sub-module, such as Figure 1 As shown, the control strategy of the UPFC sub-module includes the sub-module voltage equalization control strategy and the second-harmonic circulating current suppression strategy; among them, the second-harmonic circulating current suppression strategy outputs a modulation wave through carrier phase-shift modulation.

[0059] Specifically, the sub-module voltage equalization control strategy and the second-harmonic circulating current suppression strategy are respectively as Figure 2 and Figure 3 As shown. In the sub-module voltage equalization control strategy, the driving signals of each sub-module output by the carrier phase-shift modulation unit are redistributed through the capacitor voltage sorting algorithm. If the capacitor voltages of each module are balanced, normal driving is output without redistributing the driving signals; while the second-harmonic circulating current suppression strategy decomposes the arm circulating current i zj Suppressed by a quasi-PR controller and outputs a modulation wave compensation voltage.

[0060] Judging whether there is an abnormality according to the absolute residual specifically as follows:

[0061] Record the absolute residuals of each sampling point of the collected DC bus voltage, AC port current, and IGBT junction temperature with a set sampling window, and calculate the mean and standard deviation of all the absolute residuals of each sampling window of the DC bus voltage, AC port current, or IGBT junction temperature;

[0062] If the absolute value of the difference between the absolute residual of the DC bus voltage, AC port current, or IGBT junction temperature at a sampling point and the mean of the measured values of the DC bus voltage, AC port current, or IGBT junction temperature of its corresponding sampling window is greater than 3 times the standard deviation of the measured values of the DC bus voltage, AC port current, or IGBT junction temperature of the corresponding sampling window, it is judged that the UPFC sub-module is abnormal.

[0063] The calculation formula of the absolute residual is:

[0064]

[0065] In the above formula, r Uc,t and U c,m ( t ) are respectively t The DC bus voltage residual and the measured value at time r Iarm,t and I arm,m ( t ) are respectively tThe moment arm current residual and the measured value, r Tj,t and T j,m ( t ) are respectively t the IGBT junction temperature residual and the measured value at the moment;

[0066] Through the residual r i Calculate the mean value μ and the standard deviation σ The formula for is:

[0067]

[0068] In the above formula, n 1 is the total number of residuals calculated and recorded within the sampling window, is the i th residual within the sampling window;

[0069] The number of sub-modules in the converter arm of this embodiment is 4, and the DC bus voltage U c,rated = 200V, the rated value of the arm current I arm,rated = 12A. t = π The measured DC bus voltage at / 2 U c,m ( π / 2) = 208V, the arm current I arm,m ( π / 2) = 14.8A. At this time, the calculated values of the intrinsic model under normal conditions are 198V and 11.8A respectively; the sliding window selects n 1 = 50. At this time, the standard deviations of the calculated arm current data residuals are respectively σ Iarm = 0.5, μ Iarm = 0.2A. The calculation for judging abnormality is:

[0070]

[0071] Therefore, there is an abnormality;

[0072] The intrinsic parameters of the circuit fault include the measured value of the DC bus voltage, the deviation rate of the DC bus voltage, the measured value of the AC port current, the change rate of the AC port current di arm / dt , the measured value of the IGBT junction temperature and the IGBT junction temperature gradient dT j / dt ;

[0073] Among them, the DC bus voltage deviation rate DVR Uc、 AC port current change rate di arm / dt、 IGBT junction temperature gradient dT j / dt The calculation formula is as follows:

[0074]

[0075] In the above formula, u c,m ( t ) is the measured value of the DC bus voltage at time t , is the calculated value of the DC bus voltage at time t , U c,rated is the rated value of the DC bus voltage, I arm,m ( t ) and I arm,m ( t -1) are respectively the measured values of the AC port current at time t and t -1, T j,m ( t ) and T j,m ( t -1) are respectively the measured values of the IGBT junction temperature at time t and t -1, is the set sampling window.

[0076] The matching of the circuit fault eigenparameters with the preset abnormal feature library and the multi-level fault screening and identification according to the matching results are specifically as follows:

[0077] First, perform the first-level fault screening. Record the circuit fault eigenparameters within a sampling window after determining an abnormality. Form a vector from the DC bus voltage, DC bus voltage deviation rate, and AC port current among them, and calculate the cosine similarity between it and the first abnormal reference vectors corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that there is a corresponding fault; if all cosine similarities are less than or equal to the first threshold and greater than the second threshold, use the second-level fault screening; if all cosine similarities are less than or equal to the second threshold, use the third-level fault screening;

[0078] The second - layer fault screening is to widen the sampling window, form a vector with all the intrinsic parameters of circuit faults, and calculate the cosine similarity between it and the second abnormal reference vectors corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that the corresponding fault exists;

[0079] The third - layer fault screening is to input the intrinsic parameters of circuit faults of adjacent modules into the pre - trained BP neural network as shown in Figure 4 and output the probability of each fault. If the probability exceeds the set third threshold, it is considered that the corresponding fault exists;

[0080] The first abnormal reference vectors corresponding to the various faults are vectors composed of the historical data of the DC bus voltage, DC bus voltage deviation rate, and AC port current within a set sampling window after various faults occur; the second abnormal reference vectors corresponding to the various faults are vectors composed of the historical data of all the intrinsic parameters of circuit faults within the widened sampling window after various faults occur obtained through historical data.

[0081] The first threshold of this embodiment is 0.9, and the second threshold is 0.7;

[0082] A vector is formed by the DC bus voltage, DC bus voltage deviation rate, and AC port current within a sampling window after abnormality in this embodiment, and the cosine similarity between it and the first abnormal reference vectors corresponding to various faults in the preset abnormal feature library is calculated. Among them, the cosine similarity with the first abnormal reference vector of the IGBT short - circuit fault exceeds the first threshold of 0.9. The calculation process is as follows:

[0083]

[0084] In the formula, S is the cosine similarity, , are the values of the vectors composed of the DC bus voltage, DC bus voltage deviation rate, and AC port current, and the first abnormal reference vector of the IGBT short - circuit fault at the i th sampling point within the sampling window respectively;

[0085] S = 0.92>0.9, directly matching as an IGBT short - circuit fault, without triggering the second - and third - layer fault screenings;

[0086] The static priorities of capacitor over - voltage fault, IGBT short - circuit fault, and IGBT over - heat fault are set to 80, 70, and 60 respectively.

[0087] The calculation of the dynamic priority , specifically:

[0088]

[0089] In the above formula, P static is the set static priority, α , β and γ are the weight coefficients of the DC bus voltage, IGBT junction temperature, and AC port current respectively, U c,m ( t ) is t the measured value of the DC bus voltage at time U c,rated is the rated value of the DC bus voltage, is t the measured value of the IGBT junction temperature at time is the maximum value of the IGBT junction temperature, is t the measured value of the AC port current at time, is the maximum value of the AC port current.

[0090] In this embodiment, α , β and γ are respectively set to 0.5, 0.2, and 0.3.

[0091] As Figure 5 shown, according to the dynamic priority, corresponding fault excision is performed for different fault types, specifically:

[0092] When a single fault occurs, if it is an IGBT short - circuit fault, the sub - module drive is blocked and the bypass switch is triggered to put the redundant module into operation; if it is a capacitor over - voltage fault, the faulty module is removed and the redundant unit is put into operation, and the modulation control algorithm is used to maintain the stability of the bridge - arm level output; if it is an IGBT over - heat fault, the redundant unit is put into operation or the output of the sub - module is reduced;

[0093] If multiple faults occur, an excision sequence is generated from high to low according to the dynamic priority, the faulty sub - modules are isolated in sequence and the redundant modules are put into use. When the redundant modules are insufficient, the converter is triggered to shut down emergently.

[0094] It should be noted that when excising a fault, the excision is performed by sending an excision instruction to the module that controls the UPFC. The excision instruction uses a 16 - bit instruction code and is transmitted through a high - speed optical fiber.

[0095] The instruction code for IGBT short - circuit fault excision in this embodiment adopts the following form: 0x1A3F. The first 4 bits 0x1 indicate IGBT short - circuit, the module address is 0xA3, and the operation code 0xF is for blocking + bypass.

[0096] Figure 6Table 1 and Figure 0 respectively show the waveforms corresponding to the faults of the UPFC sub-module and the comparison of the fault response time of the embodiments of the present invention with different methods under different fault types. Figure 6 It can be seen that when the switch tube open-circuit fault occurs at 3 s, the DC bus voltage surges, and then the capacitor voltage overvoltage fault occurs. At this time, the drive of the sub-module needs to be blocked and the bypass switch needs to be triggered. As can be seen from Table 1, the method provided by the present invention is superior to the hardware method in terms of response time, and has better dynamic and complex fault handling capabilities compared with the static fault arbitration method.

[0097] Table 1 Comparison of Fault Response Times

[0098]

[0099] As Figure 1 shown, Embodiment 2 of the present invention provides a fast fault removal system for UPFC sub-modules using the method described in Embodiment 1 of the present invention, including a circuit physical intrinsic modeling module, an abnormal state identification module, and a fault processing module, characterized in that:

[0100] The circuit physical intrinsic modeling module: is used to perform circuit intrinsic modeling of the UPFC sub-module, and calculate the DC bus voltage, AC port current, and IGBT junction temperature;

[0101] The abnormal state identification module: is used to calculate the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature through the real-time collected DC bus voltage, AC port current, and IGBT junction temperature and their corresponding calculated values; all IGBTs of each UPFC sub-module are located on the same substrate, and the measured IGBT junction temperature is the substrate temperature; judge whether there is an abnormality according to the absolute value residuals. If there is an abnormality, obtain the circuit fault intrinsic parameters, match the circuit fault intrinsic parameters with a preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result;

[0102] The fault processing module: is used to set static priorities, and calculate dynamic priorities in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; the static priorities are the priorities of different fault types of the same sub-module, and the dynamic priorities are the priorities of different fault types of different sub-modules; according to the dynamic priorities, corresponding fault removal is taken for different fault types of different sub-modules.

[0103] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or terminal device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising said element.

[0104] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for quickly removing UPFC sub-module faults, characterized in that, Including: Conduct intrinsic modeling of the UPFC sub-module circuit to calculate the DC bus voltage, AC port current, and IGBT junction temperature; Based on the DC bus voltage, AC port current, and IGBT junction temperature collected in real time and their corresponding calculated values, calculate the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature; the absolute value residual is the absolute value of the difference between the calculated value and the measured value; all IGBTs of each UPFC sub-module are located on the same substrate, and the measured IGBT junction temperature is the substrate temperature; Judge whether there is an abnormality in the UPFC sub-module according to the absolute value residual. If there is an abnormality, obtain the intrinsic parameters of the circuit fault, match the intrinsic parameters of the circuit fault with the preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result; Set the static priority, and calculate the dynamic priority in combination with the collected DC bus voltage, AC port current, or IGBT junction temperature; the static priority is the priority of different fault types of the same sub-module, and the dynamic priority is the priority of different fault types of different sub-modules; When a single fault occurs, if it is an IGBT short-circuit fault, block the sub-module drive and trigger the bypass switch to put the redundant module into operation; if it is a capacitor overvoltage fault, remove the faulty module and put the redundant unit into operation, and maintain the stability of the bridge arm level output through the modulation control algorithm; if it is an IGBT overheating fault, operate the redundant unit or reduce the output of the sub-module; If multiple faults occur, generate a cut-off sequence from high to low according to the dynamic priority, isolate the faulty sub-modules in sequence and put the redundant modules into use, and trigger the emergency shutdown of the converter when the redundant modules are insufficient.

2. The method for quickly removing faults of a UPFC sub-module according to claim 1, characterized in that: Through intrinsic modeling of the UPFC sub-module circuit, establish mathematical formulas for the DC bus voltage, AC port current, and IGBT junction temperature: In the above formula, is t the calculated value of the DC bus voltage at time is t the calculated value of the AC port current at time is t the calculated value of the IGBT junction temperature at time S bridge is the switching function of the switching device in the bridge arm; I dc is the sum of the DC currents input from the DC side to the three-phase bridge arms; ω 1 is the angular frequency of the AC side voltage; I S and are respectively the amplitude and phase of the AC side current of the converter in the UPFC; MI s / 4, are respectively the amplitude and phase of the second harmonic of the circulating current; M is the modulation ratio, is the switching function of the switching device of the sub-module, C is the DC bus capacitor, is the DC bus voltage, is the number of sub-modules in the converter bridge arm in the UPFC, is the equivalent series resistance of the DC bus capacitor, T a ( t ) is the IGBT substrate temperature measured at time t , P thIGBT is the IGBT thermal loss, Z thIGBT is the IGBT thermal resistance network.

3. The method for quickly removing faults of a UPFC sub-module according to claim 2, characterized in that: Switching function of the bridge arm S bridge and the switching function of the sub-module Obtained through the control strategy adopted by the UPFC sub-module. The control strategy adopted by the UPFC sub-module includes the sub-module voltage equalization control strategy and the second-harmonic circulating current suppression strategy; among them, the second-harmonic circulating current suppression strategy outputs a modulation wave through carrier phase-shift modulation.

4. The method for quickly removing faults of a UPFC sub-module according to claim 1, characterized in that: The judgment of whether there is an abnormality according to the absolute value residual is specifically: Use the set sampling window to record the absolute value residuals of each sampling point of the collected DC bus voltage, AC port current, and IGBT junction temperature, and calculate the mean and standard deviation of all absolute value residuals of each sampling window of the DC bus voltage, AC port current, or IGBT junction temperature; If the absolute value of the difference between the absolute value residual of the DC bus voltage, AC port current, or IGBT junction temperature of a sampling point and the mean value of the measured DC bus voltage, AC port current, or IGBT junction temperature of its corresponding sampling window is greater than 3 times the standard deviation of the measured DC bus voltage, AC port current, or IGBT junction temperature of the corresponding sampling window, it is judged that the UPFC sub-module is abnormal.

5. The method for quickly removing faults of a UPFC sub-module according to claim 1, characterized in that: The intrinsic parameters of the circuit fault include the measured value of the DC bus voltage, the deviation rate of the DC bus voltage, the measured value of the AC port current, the change rate of the AC port current, the measured value of the IGBT junction temperature, and the IGBT junction temperature gradient; Among them, the DC bus voltage deviation rate DVR Uc、 The AC port current change rate di arm / dt、 The IGBT junction temperature gradient dT j / dt The calculation formula is as follows: In the above formula, u c,m ( t ) is t the measured value of the DC bus voltage at time is t the calculated value of the DC bus voltage at time U c,rated is the rated value of the DC bus voltage, I arm,m ( t ) and I arm,m ( t - 1) are respectively t and t the measured values of the AC port current at time T j,m ( t ) and T j,m ( t - 1) are respectively t and t the measured values of the IGBT junction temperature at time is the set sampling window.

6. A method for quickly removing faults in a UPFC sub-module according to claim 5, characterized in that: The method of matching the intrinsic parameters of the circuit fault with a preset abnormal feature library and performing multi-level fault screening and identification according to the matching result is as follows: First, perform the first-level fault screening. Record the intrinsic parameters of the circuit fault within a sampling window after determining the abnormality. Form a vector from the DC bus voltage, the deviation rate of the DC bus voltage, and the AC port current among them, and calculate the cosine similarity between it and the first abnormal reference vector corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that the corresponding fault exists; if all cosine similarities are less than or equal to the first threshold and greater than the second threshold, use the second-level fault screening; if all cosine similarities are less than or equal to the second threshold, use the third-level fault screening; The second-level fault screening is to widen the sampling window, form a vector from all the intrinsic parameters of the circuit fault, and calculate the cosine similarity between it and the second abnormal reference vector corresponding to various faults in the preset abnormal feature library. If there is a cosine similarity greater than the set first threshold, it is considered that the corresponding fault exists; The third-level fault screening is to input the intrinsic parameters of the circuit fault of adjacent modules into a pre-trained BP neural network and output the probability of each fault; if the probability exceeds the set third threshold, it is considered that the corresponding fault exists; The first abnormal reference vector corresponding to various faults is a vector composed of the historical data of the DC bus voltage, the deviation rate of the DC bus voltage, and the AC port current within a set sampling window after various faults occur; the second abnormal reference vector corresponding to various faults is a vector composed of the historical data of all the intrinsic parameters of the circuit fault within the widened sampling window after various faults occur obtained through historical data.

7. A method for quickly removing faults in a UPFC sub-module according to claim 1, characterized in that: The static priorities of the capacitor overvoltage fault, the IGBT short-circuit fault, and the IGBT overheating fault are set to 80, 70, and 60 respectively.

8. A method for quickly removing faults in a UPFC sub-module according to claim 7, characterized in that: The calculation of the dynamic priority , specifically: In the above formula, P static is the set static priority, α , β and γ are the weight coefficients of the DC bus voltage, IGBT junction temperature, and AC port current respectively, U c,m ( t ) is the measured value of the DC bus voltage at t moment, U c,rated is the rated value of the DC bus voltage, is t the measured value of the IGBT junction temperature at is the maximum value of the IGBT junction temperature, is t the measured value of the AC port current at is the maximum value of the AC port current.

9. A UPFC sub-module fault quick removal system using the method according to any one of claims 1-8, including a circuit physical intrinsic modeling module, an abnormal state identification module, and a fault processing module, characterized in that: The circuit physical intrinsic modeling module: is used for performing circuit intrinsic modeling of the UPFC sub-module and calculating the DC bus voltage, the AC port current, and the IGBT junction temperature; Abnormal state identification module: used to calculate the absolute value residuals of the DC bus voltage, AC port current, and IGBT junction temperature through the real-time collected DC bus voltage, AC port current, IGBT junction temperature and their corresponding calculated values; all IGBTs of each UPFC sub-module are located on the same substrate, and the measured IGBT junction temperature is the substrate temperature; judge whether there is an abnormality according to the absolute value residuals. If there is an abnormality, obtain the intrinsic parameters of the circuit fault, match the intrinsic parameters of the circuit fault with the preset abnormal feature library, and perform multi-level fault screening and identification according to the matching result; Fault processing module: used to set static priorities and calculate dynamic priorities in combination with the collected DC bus voltage, AC port current or IGBT junction temperature; the static priorities are the priorities of different fault types of the same sub-module, and the dynamic priorities are the priorities of different fault types of different sub-modules; according to the dynamic priorities, corresponding fault removal is taken for different fault types of different sub-modules.

Images