Fault diagnosis method and system for cascaded H-bridge converter

Abstract

The invention relates to the technical field of electrical engineering fault diagnosis, and provides a fault diagnosis method and system for a cascaded H-bridge converter, and the method comprises the steps: collecting a power grid current at a transient AC side through a current sensor, extracting a DC component Igdc, comparing the DC component Igdc with a preset threshold value, judging whether the system has an open-circuit fault of a power device or not, and if yes, carrying out the fault diagnosis. Suspicious fault pipe pairs are determined; removing the direct current bus voltage balance control from the system, and recovering the direct current bus voltage balance control of other modules except the fault module after the fault module is positioned; by changing a driving signal of a power device of the fault module, the fault module is short-circuited, and a specific fault switch tube is positioned at the same time; and the system recovers stable work by using other module modulation coefficient redundancy. According to the method, after an open-circuit fault occurs, the fault device is quickly, efficiently and reliably positioned and the normal work of the system is recovered by temporarily cutting off voltage balance control.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical engineering fault diagnosis, and in particular to a fault diagnosis method and system for a cascaded H-bridge converter. Background Art

[0002] Compared to traditional two-level converters, multilevel converters offer a major advantage in their smaller output voltage steps, resulting in lower harmonic content and better electromagnetic compatibility. Cascaded H-bridge multilevel converters are widely used in photovoltaic systems, static synchronous compensators, and other fields due to their high modularity, high voltage capability, low harmonic content, and strong fault tolerance. However, the large number of power devices in a CHB significantly increases the likelihood of power device failure. Power device failures can be classified as open-circuit or short-circuit. While short-circuit failures have a short time interval from onset to significant harm, requiring rapid detection and isolation, their fault characteristics are highly visible and can be quickly and accurately detected and located using hardware integrated into the power devices. However, open-circuit failures are subtle, making them difficult to quickly and accurately identify. Among these, single power device failures have the highest probability and have been a hot topic of research.

[0003] The document "A State Estimator-Based Approach for Open-Circuit Fault Diagnosis in Single-Phase Cascaded H-Bridge Rectifiers" (IEEE Transactions on Industry Applications, 2019, 55(2): 1608-1618.), published in 2019, proposes a fault diagnosis strategy based on grid current residual to locate open-circuit fault power devices. This diagnostic scheme mathematically models the cascaded H-bridge inverter system to obtain an estimated expression for the grid current. The estimated value of the grid current is then subtracted from the measured value to obtain the grid current residual, and the current residual is then used to diagnose and locate the fault. Due to the high accuracy requirements for capacitor voltage and grid current, this scheme requires the use of high-precision voltage / current sensors, resulting in high diagnostic costs.

[0004] The 2014 paper "Research on Fault Diagnosis and Fault-Tolerant Control of Cascaded STATCOMs" (by Yang Xiaodong, China University of Mining and Technology) proposes an AI-based fault diagnosis solution. This solution uses wavelet analysis to perform multi-resolution analysis of the DC bus voltage. Using AI, it classifies the characteristic vectors derived from the wavelet decomposition for different fault types, thereby diagnosing and locating the fault. This solution requires a large training dataset and is significantly affected by parameters.

[0005] Some scholars have also proposed fault diagnosis schemes based on capacitor voltage. These schemes all locate the fault based on the rise of the DC bus voltage of the faulty module after the open circuit fault of the power device. The document "A FaultySubmodule Mathematical Model-Based Localization Strategy for Switch Open-Circuit Fault of Module Multilevel Converter" (IEEE Transactions on Power Electronics, 2023, 38(3): 3899-3916.) published in 2023 obtained the estimated value expression of the capacitor voltage through mathematical modeling and calculated the capacitor voltage residual to locate the faulty module. The rising trend of the capacitor voltage over a period of time was monitored to identify the faulty module. The document published in 2022, "Online Diagnosis and Ride-Through Operation for Cascaded H-Bridge Converter Based STATCOM With a Single Open-Circuit IGBT" (IEEE Transactions on Industrial Electronics, 2022, 69(8):7549-7559.) monitors the rising trend of capacitor voltage over a period of time to identify the faulty module. The document published in 2023, "A Novel Detection and Localization Approach of Open-Circuit Switch Fault for the Grid-Connected Modular Multilevel Converter" (IEEE Transactions on Industrial Electronics, 2023, 70(1):112-124.) proposes an improved Pauta criterion to monitor abnormal capacitor voltage data and thus locate the faulty module. All of the above methods can effectively locate the faulty module.

[0006] However, the DC voltage balancing control in the cascaded H-bridge system suppresses the fault signature of capacitor voltage rise, reducing the robustness and reliability of the fault diagnosis strategy. Furthermore, in publicly available fault diagnosis schemes, faulty power device location and fault tolerance are separated, hindering the rapid recovery of the system. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to improve the efficiency and reliability of locating open-circuit fault components of a cascaded H-bridge converter.

[0008] The present invention solves the above technical problems through the following technical means:

[0009] The present invention provides a fault diagnosis method for a cascaded H-bridge converter, comprising:

[0010] S10. Use current sensors to collect transient AC grid current and extract its DC component. The DC component The absolute value of is compared with a preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty transistor pair is determined, and the process continues to step S20; otherwise, the process returns to step S10.

[0011] S20, removing the DC bus voltage balancing control from the system and locating the faulty module;

[0012] S30. Restore DC bus voltage balancing control of the modules other than the faulty module; short-circuit the faulty module by changing the drive signal of the power device of the faulty module and monitoring the capacitor voltage change trend of the faulty module under the drive signal; utilize the redundancy of the modulation coefficients of other modules to restore the system to stable operation;

[0013] S40 , according to the execution process of step S30 , combined with the suspected faulty switch pair determined in step S10 , locate the specific faulty switch.

[0014] Furthermore, the DC component threshold is specifically:

[0015]

[0016] Where β is the threshold coefficient, which ranges from 10% to 20%.

[0017] Furthermore, the determination of the faulty pipe pair is specifically as follows:

[0018] For the DC component If the polarity is positive, the suspected faulty tube pair is S2 and S3; if the polarity of the DC component is negative, the suspected faulty tube pair is S1 and S4.

[0019] Furthermore, the fault location module in step S20 includes the following steps:

[0020] (1) Use the DC side capacitor voltage sensor to record the capacitor voltage value;

[0021] (2) Compare the capacitor voltage of each module with a preset capacitor voltage threshold. If the capacitor voltage exceeds the threshold, it is determined that the corresponding module has an open circuit fault.

[0022] Furthermore, the DC-side capacitor voltage sensor is used to record the value of the capacitor voltage, specifically:

[0023] In the converter topology, N sensors are used to detect and record the capacitor voltage signal. The sampling frequency of the sensors is the same, and the number of sampling points in one cycle is:

[0024]

[0025] Where T is the sensor output signal period, T s is the controller cycle;

[0026] Define a data storage matrix to record the sampling values ​​of each capacitor voltage signal in the past cycle, denoted as matrix A i ; where k is any sampling time t;

[0027] A i =[V Ci (kL-1) V Ci (kL) … V Ci (k-1) V Ci (k)]

[0028] The above data storage matrix is ​​periodically slid to obtain a periodic sliding window to obtain the sampling values ​​of the three-phase current; the sampling window length L is the number of current sampling points in one cycle, L = 1 / (T s ·f0), where T s is the controller cycle, and f0 is the frequency of the output current. By storing the sampling points within one cycle of the three-phase grid current into three matrices, each new sampling deletes the earliest sampling point in the matrix and adds the new sampling point to the end of the matrix, thereby updating the matrix.

[0029] Furthermore, the capacitor voltage threshold is specifically set as follows:

[0030]

[0031] Where, is the capacitor voltage reference value, α is the capacitor voltage threshold coefficient, and its value is 1.3-1.5;

[0032] Capacitor voltage fluctuation due to power pulsation:

[0033]

[0034] Furthermore, the step S30 is to short-circuit the faulty module by changing the driving signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the driving signal, specifically:

[0035] Adjust the drive signal of the fault module so that the switch tubes S1 and S3 of the fault module remain turned on, and the switch tubes S2 and S4 remain turned off;

[0036] Monitor the capacitor voltage change trend of the faulty module. If the capacitor voltage remains unchanged or slowly decreases, it is determined that the switches S1 and S3 are normally turned on and the faulty module is short-circuited. Otherwise, adjust the drive signal of the faulty module so that the switches S2 and S4 of the faulty module remain turned on and the switches S1 and S3 remain turned off.

[0037] Furthermore, the redundancy of the modulation coefficients of other modules is utilized in step S30 to restore the system to stable operation, specifically:

[0038] Assuming the current of the system remains unchanged during normal operation, the single-phase system operation must meet the following requirements:

[0039] v a =v g +jωLi g

[0040] Where v a is the output voltage of the cascaded H-bridge, v g is the grid voltage, i g is the grid current; where v a The output voltage of each H-bridge is superimposed:

[0041]

[0042] Express v in terms of modulation i :

[0043] v i =(m di +m qi )V ref

[0044] Where m di and m qi are the modulation coefficients of the active and reactive components of the i-th H-bridge respectively; and the modulation coefficients must satisfy the constraints:

[0045]

[0046] v a It can be expressed as:

[0047]

[0048] Then we have:

[0049]

[0050] Therefore, the active and reactive power output by the cascaded H-bridge multilevel converter can be expressed as:

[0051]

[0052] Assuming the system is operating normally, i g Remain unchanged, that is:

[0053]

[0054] Under the premise that the modulation coefficient of the system is redundant, after a single module fails, the modulation coefficients of the remaining modules are increased so that the cascaded H-bridge converters can output active and reactive power that ensures stable operation of the system.

[0055] Furthermore, the step S40 is specifically as follows:

[0056] (1) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S2;

[0057] (2) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S4;

[0058] (3) When the switch tube S2 and the switch tube S4 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S3;

[0059] (4) When the switch tube S2 and the switch tube S4 remain turned on, the faulty module is short-circuited; and the suspected faulty tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S1.

[0060] The present invention also provides a fault diagnosis system for a cascaded H-bridge converter, which executes the above method when the system is running, and includes the following modules:

[0061] System fault judgment module, used to use current sensors to collect transient AC grid current and extract its DC component The DC component The absolute value of is compared with the preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty tube pair is determined, and the fault module location module is continued to execute; otherwise, the execution returns to the system fault determination module;

[0062] Fault module locating module, used to remove DC bus voltage balancing control from the system and locate the faulty module;

[0063] The fault recovery module is used to restore the DC bus voltage balance control of the modules except the faulty module. It short-circuits the faulty module by changing the drive signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the drive signal. It also uses the redundancy of the modulation coefficients of other modules to restore the system to stable operation.

[0064] The faulty device locating module is used to locate the specific faulty switch tube according to the execution process of the fault recovery module and the suspected faulty tube pair determined by the system fault judgment module.

[0065] The advantages of the present invention are:

[0066] (1) After a fault is diagnosed, the capacitor voltage balancing control is immediately removed from the system, so that the fault characteristic of the capacitor voltage rise is amplified, thereby improving the speed and robustness of the fault module diagnosis.

[0067] (2) The system modulation redundancy is utilized for fault tolerance without the need to invest in idle H-bridge modules, thus saving system operating costs.

[0068] (3) Combining the location of specific power devices with fault tolerance improves the speed of fault diagnosis.

[0069] (4) The data used for fault diagnosis and fault tolerance is the amount of data required by the control system. Therefore, there is no need to add additional voltage / current sensors, saving costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 1 is a flow chart of a fault diagnosis method for a cascaded H-bridge converter according to an embodiment of the present invention;

[0071] Figure 2 1 is a topology diagram of a single-phase cascaded H-bridge multilevel converter according to an embodiment of the present invention;

[0072] Figure 3 is a specific flow chart of step S10 in an embodiment of the present invention;

[0073] Figure 4Schematic diagram of the output voltage and fundamental wave and DC output voltage components of normal and faulty HBs in an embodiment of the present invention;

[0074] Figure 5 is a DC equivalent circuit of a power device failure in an embodiment of the present invention;

[0075] Figure 6 This is the performance analysis of the H bridge when S1 fails in the embodiment of the present invention;

[0076] Figure 7 is a control block diagram of a cascaded H-bridge multi-level converter according to an embodiment of the present invention;

[0077] Figure 8 is a schematic block diagram of post-fault removal voltage balancing control in an embodiment of the present invention;

[0078] Figure 9 is a schematic block diagram of voltage balancing control of a health recovery module in an embodiment of the present invention;

[0079] Figure 10 is the grid current waveform before and after the S1 open circuit fault in the embodiment of the present invention;

[0080] Figure 11 : is the DC component waveform of the grid current before and after the S1 open circuit fault in the embodiment of the present invention;

[0081] Figure 12 This is a comparison of capacitor voltages before and after releasing voltage balancing control under healthy operation and single-transistor failure (S1 open circuit) conditions in an embodiment of the present invention when module parameters are inconsistent;

[0082] Figure 13 is the capacitance voltage variation trend of the faulty module before and after the fault in the embodiment of the present invention;

[0083] Figure 14 is a curve showing a change in capacitor voltage of a faulty module during a fault tolerance process according to an embodiment of the present invention;

[0084] Figure 15 1 is a capacitor voltage curve of each module after the faulty module is short-circuited in an embodiment of the present invention. DETAILED DESCRIPTION

[0085] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0086] Example 1

[0087] This embodiment provides a fault diagnosis method for a cascaded H-bridge converter. The process is as follows: Figure 1 Shown, including:

[0088] S10. Use current sensors to collect transient AC grid current and extract its DC component. The DC component The absolute value of is compared with a preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty tube pair is determined, and step S20 is continued; otherwise, the process returns to step S10.

[0089] For the DC component If the polarity is positive, the suspected faulty tube pair is S2 and S3; if the polarity of the DC component is negative, the suspected faulty tube pair is S1 and S4.

[0090] It should be noted that the topology of the single-phase cascade H-bridge multi-level converter is as follows Figure 2 As shown in the figure, by changing the drive signals of each power device—specifically, changing the drive signals that rapidly switch between high and low levels during normal operation to a continuous low level—an open-circuit fault in the power device is simulated. As a result, the output voltage of the cascaded H-bridge converter undergoes a transient change, which in turn causes the grid current to generate a DC component related to the open-circuit fault in the power device.

[0091] The specific process of S10 is as follows Figure 3 As shown, this embodiment introduces an open-circuit fault into a power device during the normal operation of the cascaded H-bridge system, and simulates the open-circuit fault by changing the drive signal of the power device. When the drive signal of the faulty power device is set to a low level, its open-circuit fault is simulated, causing a DC bias in the AC side current, and generating a DC component that can be identified and extracted by a low-precision current sensor within a safety margin. Since the damage caused by the open-circuit fault is hidden to the system, the system will not collapse immediately. However, if the open-circuit fault is left unchecked, it may increase the damage to the system, and increase the current or voltage stress on other devices, thereby causing a secondary fault or even a collapse of the system. Therefore, fault diagnosis needs to ensure rapidity, and requirements are made for the selection of thresholds.

[0092] A detailed analysis of the output voltage components after a failure of different power devices of HB (H-Bridge) is as follows: Figure 4 As shown. Figure 4As can be seen from the figure, the output voltage of a normal HB is filtered through a low-pass filter (LPF) to obtain the fundamental component, while the output voltage of a faulty HB is filtered to obtain the superposition of the fundamental component and the DC component. In other words, the output voltage of a faulty HB can be viewed as a reduced-order AC voltage source with a DC voltage disturbance. It is worth noting that the polarity of the DC voltage disturbance of the faulty module depends on the location of the faulty power device. Figure 4 As shown in FIG, when an open-circuit fault occurs in S1 or S4, the DC voltage disturbance is negative; when an open-circuit fault occurs in S2 or S3, the DC voltage disturbance is positive.

[0093] When it comes to the DC voltage component generated by the faulty H-bridge, a circuit analysis is performed. The simplified DC equivalent circuit of the system is as follows Figure 5 As shown, the grid voltage U g , filter inductor L f The healthy H-bridge is considered as a DC short circuit, and the faulty HB is equivalent to a DC voltage source. It can be seen that the abnormal DC component in the grid current comes from the DC voltage of the faulty module, and the output DC current polarity of the phase where the faulty HB is located is always opposite to that of other phases. In addition, Figure 5 As shown in the bottom table, the DC current polarity is determined only by the position of the fault switch.

[0094] Generally, the threshold value of the DC component of the grid current should be selected as:

[0095]

[0096] Where β is the threshold coefficient, which ranges from 10% to 20%.

[0097] S20: Remove the DC bus voltage balancing control from the system and locate the fault module; the fault location module includes the following steps:

[0098] (1) Use the DC side capacitor voltage sensor to record the capacitor voltage value; specifically:

[0099] In the converter topology, N sensors are used to detect and record the capacitor voltage signal. The sampling frequency of the sensors is the same, and the number of sampling points in one cycle is:

[0100]

[0101] Where T is the sensor output signal period, T s is the controller cycle;

[0102] Define a data storage matrix to record the sampling values ​​of each capacitor voltage signal in the past cycle, denoted as matrix A i ; where k is any sampling time t;

[0103] A i =[V Ci (kL-1)V Ci (kL)…V Ci (k-1)V Ci (k)]

[0104] The above data storage matrix is ​​periodically slid to obtain a periodic sliding window to obtain the sampling values ​​of the three-phase current; the sampling window length L is the number of current sampling points in one cycle, L = 1 / (T s ·f0), where T s is the controller cycle, and f0 is the frequency of the output current. By storing the sampling points within one cycle of the three-phase grid current into three matrices, each new sampling deletes the earliest sampling point in the matrix and adds the new sampling point to the end of the matrix, thereby updating the matrix.

[0105] (2) Compare the capacitor voltage of each module with a pre-set capacitor voltage threshold. If the capacitor voltage threshold is exceeded, it is determined that the corresponding module has an open circuit fault. The capacitor voltage threshold is specifically set as follows:

[0106]

[0107] Where, is the capacitor voltage reference value, α is the capacitor voltage threshold coefficient, and its value is 1.3-1.5;

[0108] Capacitor voltage fluctuation due to power pulsation:

[0109]

[0110] It should be noted that if Figure 6 As shown in Figure 1, the performance analysis of the H bridge when S1 fails. λ represents the direction of current. When λ = 1, i g >0, λ=0 when i g <0 (current flowing out of the H-bridge is in the positive direction). [g1, g2, g3, g4] represents the switching sequence of the power devices. Figure 6 The upper and lower sub-figures are the current flow paths when S1 is normal and faulty respectively. Figure 6 In (a), the current is in the positive direction. When S1 has an open circuit fault, the current does not flow through the capacitor, that is, the capacitor no longer discharges. Figure 6 In (b), when λ=1 and S1 fails, the capacitor switches from the freewheeling state to the charging state. When the current is in the negative direction, the fault tube of S1 has no effect on HB, as shown in Figure 6(c) shows that the H-bridge output performance is affected only when λ = 1. In short, when S1 fails, the capacitor voltage will rise, and this rise only lasts for half the current cycle. The same analysis shows that any power device failure will cause the capacitor voltage to rise.

[0111] The control block diagram of the single-phase cascade H bridge in this example is as follows Figure 7 As shown in the figure, it consists of three controllers: a cluster voltage controller, a current controller, and a DC bus voltage balancing controller. The cluster voltage controller generates the reference value of the d-axis component of the grid current, which regulates the active power output by the inverter. The grid current sample is delayed by 1 / 4 cycle to generate the orthogonal component required for coordinate system transformation, thereby obtaining the q-axis component of the sampled current. The q-axis component of the reference current controls the reactive power. The reference grid current is then regulated in the dq coordinate system using the PI controller and passed through the current controller to obtain the modulation voltage reference value. To ensure that the capacitor voltages of each H-bridge are equal, the modulation voltage of each submodule is obtained by feeding it into the DC bus voltage balancing controller.

[0112] The individual capacitor voltage controllers in this controller balance the capacitor voltages by controlling the active power flowing into each H-bridge capacitor. Therefore, this controller introduces an additional term into the AC voltage reference of each H-bridge to control the active power distribution between the H-bridges. Furthermore, if the actions of this controller are to not interfere with the grid current regulated by different controllers, the following conditions must be met to decouple the individual capacitor voltage controllers from the rest of the control system:

[0113]

[0114] Violations of the above equation will change the reference voltage generated by the current controller, which will have a direct impact on the current, and will cause disturbances in the output of the cluster voltage controller, which will have an indirect impact on the current.

[0115] In order to amplify the fault characteristic of the capacitor voltage rise after the fault, it is necessary to switch the control mode of the cascade H bridge, as shown in the schematic diagram. Figure 8 As shown. That is, the modulation voltage of each module is equal to the reference voltage output by the current controller, without injecting the adjustment amount of the capacitor voltage balance control output. As mentioned in the above step S10, the grid current i g After the fault, a significant DC bias will appear. However, after temporarily releasing the capacitor voltage balancing control, i g There will be almost no obvious changes, that is, temporarily releasing the capacitor voltage balancing control will not aggravate the fault level of the system and affect its health status.

[0116] Due to the open-circuit fault of the power device in the module, the capacitor voltage of the module transiently rises. However, due to the inhibitory effect of the DC bus voltage balancing control, the fault characteristics are not obvious. Therefore, by using the control switching, the DC bus voltage balancing control is removed from the system, so that the fault characteristics of the module capacitor voltage rise are amplified. Furthermore, by recording the transient capacitor voltage changes and comparing them with the pre-set capacitor voltage threshold, the faulty module is determined. The method of this embodiment amplifies the fault characteristics, thereby avoiding the use of high-precision voltage sensors and adding no additional hardware, thus saving monitoring costs.

[0117] S30, restore the DC bus voltage balancing control of the modules except the faulty module, such as Figure 9 As shown in the figure, by changing the driving signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the driving signal, the faulty module is short-circuited; the redundancy of the modulation coefficients of other modules is used to restore the system to stable operation. The specific steps are:

[0118] Adjust the drive signal of the fault module so that the switch tubes S1 and S3 of the fault module remain turned on, and the switch tubes S2 and S4 remain turned off;

[0119] Monitor the capacitor voltage change trend of the faulty module. If the capacitor voltage remains unchanged or slowly decreases, it is determined that the switches S1 and S3 are normally turned on and the faulty module is short-circuited. Otherwise, adjust the drive signal of the faulty module so that the switches S2 and S4 of the faulty module remain turned on and the switches S1 and S3 remain turned off.

[0120] When the faulty module is short-circuited, the modulation coefficient redundancy of other modules is needed to restore the system to stable operation. The specific steps are:

[0121] Assuming the current of the system remains unchanged during normal operation, the single-phase system operation must meet the following requirements:

[0122] v a =v g +jωLi g

[0123] Where v a is the output voltage of the cascaded H-bridge, v g is the grid voltage, i g is the grid current; where v a The output voltage of each H-bridge is superimposed:

[0124]

[0125] Express v in terms of modulation i :

[0126] v i =(m di +m qi )V ref

[0127] Where m di and m qi are the modulation coefficients of the active and reactive components of the i-th H-bridge respectively; and the modulation coefficients must satisfy the constraints:

[0128]

[0129] v a It can be expressed as:

[0130]

[0131] Then we have:

[0132]

[0133] Therefore, the active and reactive power output by the cascaded H-bridge multilevel converter can be expressed as:

[0134]

[0135] Assuming the system is operating normally, i g Remain unchanged, that is:

[0136]

[0137] Under the premise that the modulation coefficient of the system is redundant, after a single module fails, the modulation coefficients of the remaining modules are increased so that the cascaded H-bridge converters can output active and reactive power that ensures stable operation of the system.

[0138] S40: Based on the execution process of step S30 and the suspected faulty switch pair determined in step S10, locate the specific faulty switch. The specific steps are:

[0139] (1) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S2;

[0140] (2) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S4;

[0141] (3) When the switch tube S2 and the switch tube S4 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S3;

[0142] (4) When the switch tube S2 and the switch tube S4 remain conductive, the fault module is short-circuited; and the suspected faulty tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S1;

[0143] This embodiment also describes the proposed method in detail using a single-phase cascaded H-bridge multilevel converter operating in a STATCOM (Static Synchronous Compensator) mode, and provides Matlab simulation results to verify the effectiveness of the proposed method.

[0144] Simulation parameters:

[0145] A single-phase cascaded H-bridge multilevel converter system with four H-bridge modules (N), a 20V DC bus voltage, a 50V RMS grid voltage, a 50Hz grid voltage frequency, a 4mH filter inductor, a 1.7mF rated DC capacitor, a 4A grid current, and a power factor of 0. Modulation uses single-frequency carrier phase-shift modulation with a 10kHz carrier frequency. The equivalent switching frequency of the cascaded H-bridge is 20kHz.

[0146] Step (1) issues a power device open circuit fault instruction, collects the grid current, and determines the occurrence of a fault based on the DC component of the grid current:

[0147] A power device open-circuit fault command is issued, which alters the drive signal for power device S1 in the first module. Specifically, the drive signal, which rapidly switches between high and low levels during normal operation, is changed to a continuously low level, simulating an open-circuit fault in the power device. An AC-side current sensor is used to record the grid current and extract the DC component. This DC component is compared with a preset threshold. If it exceeds the threshold, an open-circuit fault in the power device is determined. The suspected faulty transistor pair is identified based on the polarity of the DC component of the grid current. Specifically, if the DC component is positive, the suspected faulty transistor pair is S2 and S3; if the DC component is negative, the suspected faulty transistor pair is S1 and S4.

[0148] like Figure 10 and Figure 11 The following figures show the waveforms of the grid current and its DC component before and after the open-circuit fault of power device S1 in the first module. It can be seen that an open-circuit fault occurred in S1 at t = 0.5s. After the fault, the grid current exhibited a negative DC bias, and its DC component exceeded the threshold, indicating an open-circuit fault. Furthermore, the negative polarity of this DC component indicates that S1 and S4 are the suspected faulty transistor pair.

[0149] Step (2) modifies the parameter value of the capacitor so that the parameters of each module are inconsistent, and verifies whether it affects the fault characteristics.

[0150] Specifically, the capacitance value of module 2 is changed to 1.6mF, the capacitance value of module 3 is changed to 1.8mF, and the capacitance value of module 4 is changed to 1.5mF. Therefore, the capacitance voltage of each module is different, that is, the parameters of each module are inconsistent. Figure 12 As shown, when the module parameters are inconsistent, Figure 12 (a) and Figure 12 (b) Comparison of capacitor voltage before and after releasing voltage balancing control under healthy operation and power device single tube fault (S1 open circuit) conditions. Figure 12 As shown in (a), due to parameter inconsistency, the capacitor voltages of the healthy CHB only deviate after voltage balancing control is released at t = 0.5s, without a significant increase. This is because when the parameters are inconsistent, the active power carried by each SB varies. However, at t = 0.5s, an open-circuit fault occurs simultaneously with the release of voltage balancing control, and the capacitor voltage of the faulty module increases significantly. This shows that only module parameter inconsistency exists; releasing voltage balancing control will not trigger a false diagnosis.

[0151] Step (3) Switch the control mode of the H-bridge, remove the DC capacitor voltage balancing control from the system, and amplify the fault characteristics to locate the faulty module:

[0152] First, the control mode of the cascaded H-bridge is switched to remove the DC bus voltage balancing control from the system. Specifically, the component of the modulation voltage of the four modules output by the capacitor voltage balancing control is removed.

[0153] Then, the DC-side capacitor voltage sensor is used to record the capacitor voltage. The sampling frequency is set to 10kHz, and a sliding window is used for real-time sampling to obtain a sampling matrix.

[0154] Finally, the capacitor voltage is compared with the pre-set capacitor voltage threshold. If the capacitor voltage of the module exceeds the threshold, it is determined that the module has failed. The above sampling matrix is ​​calculated to obtain the average capacitor voltage of the four modules. According to the capacitor voltage threshold formula, α = 1.3 is selected to obtain

[0155] like Figure 13 Figure 1 shows the capacitor voltage variation trend of the faulty module before and after the fault. After the fault is diagnosed, the H-bridge control mode is switched, and the capacitor voltage of the faulty module rises rapidly. When it exceeds the threshold, the faulty module is located.

[0156] Step (4) uses system modulation redundancy to restore the system to healthy operation and locate the faulty power device in the process:

[0157] After locating the faulty module, the DC bus voltage balancing control is restored for all modules except the faulty module. By changing the drive signal for the faulty module's power device, the faulty module is short-circuited. The system is restored to normal operation by leveraging the redundant modulation coefficients of other modules. Simultaneously, the capacitor voltage trend of the faulty module is monitored, and the specific faulty power device is located based on the suspected faulty transistor pair.

[0158] like Figure 14 As shown in Figure 1, the change in the capacitor voltage of the faulty module during the fault tolerance process is given. After locating the faulty module, the drive signal of the faulty module is immediately changed to [1, 0, 1, 0]. That is, after turning on S1 and S3, the capacitor voltage of the faulty module still shows an upward trend, which means that the fault tolerance has failed. Combined with the suspected faulty transistor pair S1 and S4 determined in step (1), the faulty power device S1 can be located.

[0159] Then, the driving signal of the faulty module is changed to [0, 1, 0, 1], which makes S2 and S4 conductive, causing the faulty module to short-circuit, and thus restoring the system to health. Figure 15 Figure 2 shows the capacitor voltage curves of each module after the faulty module is short-circuited. It can be seen that after the faulty module is short-circuited, the capacitor voltage of the faulty module slowly decreases; the capacitor voltages of the remaining modules fluctuate around the capacitor voltage reference value, and the system returns to healthy operation.

[0160] Example 2

[0161] It should be further explained that, based on the same inventive concept, the present invention also provides a fault diagnosis system for a cascaded H-bridge converter. When the system is running, the method described in Example 1 is executed, including the following modules:

[0162] System fault judgment module, used to use current sensors to collect transient AC grid current and extract its DC component The DC component The absolute value of is compared with the preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty tube pair is determined, and the fault module location module is continued to execute; otherwise, the execution returns to the system fault determination module;

[0163] Fault module locating module, used to remove DC bus voltage balancing control from the system and locate the faulty module;

[0164] The fault recovery module is used to restore the DC bus voltage balance control of the modules except the faulty module. It short-circuits the faulty module by changing the drive signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the drive signal. It also uses the redundancy of the modulation coefficients of other modules to restore the system to stable operation.

[0165] The faulty device locating module is used to locate the specific faulty switch tube according to the execution process of the fault recovery module and the suspected faulty tube pair determined by the system fault judgment module.

[0166] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A fault diagnosis method for a cascaded H-bridge converter, characterized in that: include: S10. Use current sensors to collect transient AC grid current and extract its DC component. The DC component The absolute value of is compared with a preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty transistor pair is determined, and the process continues to S20; otherwise, the process returns to S10. S20, removing the DC bus voltage balancing control from the system and locating the faulty module; S30. Restore DC bus voltage balancing control of the modules other than the faulty module; short-circuit the faulty module by changing the drive signal of the power device of the faulty module and monitoring the capacitor voltage change trend of the faulty module under the drive signal; utilize the redundancy of the modulation coefficients of other modules to restore the system to stable operation; S40 , according to the execution process of S30 and in combination with the suspected faulty switch tube pair determined in S10 , locate the specific faulty switch tube.

2. A fault diagnosis method for a cascaded H-bridge converter according to claim 1, characterized in that: The DC component threshold is specifically: Where β is the threshold coefficient, which ranges from 10% to 20%.

3. The fault diagnosis method for a cascaded H-bridge converter according to claim 1, characterized in that: The specific steps of determining the suspected faulty pipe pair are as follows: For the DC component If the polarity is positive, the suspected faulty tube pair is S2 and S3; If the polarity of the DC component is negative, the suspected faulty tube pair is S1 and S4.

4. The fault diagnosis method for a cascaded H-bridge converter according to claim 1, characterized in that: The fault location module in step S20 includes the following steps: (1) Use the DC side capacitor voltage sensor to record the capacitor voltage value; (2) Compare the capacitor voltage of each module with a preset capacitor voltage threshold. If the capacitor voltage exceeds the threshold, it is determined that the corresponding module has an open circuit fault.

5. The fault diagnosis method for a cascaded H-bridge converter according to claim 4, characterized in that: The DC side capacitor voltage sensor is used to record the value of the capacitor voltage, specifically: In the converter topology, N sensors are used to detect and record the capacitor voltage signal. The sampling frequency of the sensors is the same, and the number of sampling points in one cycle is: Where T is the sensor output signal period, T s is the controller cycle; Define a data storage matrix to record the sampling values ​​of each capacitor voltage signal in the past cycle, denoted as matrix A i ; where k is any sampling time t; A i =[V Ci (kL-1) V Ci (kL) … V Ci (k-1) V Ci (k)] The above data storage matrix is ​​periodically slid to obtain a periodic sliding window to obtain the sampling values ​​of the three-phase current; the sampling window length L is the number of current sampling points in one cycle, L = 1 / (T s ·f0), where T s is the controller cycle, f0 is the frequency of the output current; by storing the sampling points within one cycle of the three-phase grid current into three matrices, each new sampling will delete the earliest sampling point in the matrix and add the new sampling point to the end of the matrix, thereby updating the matrix.

6. A fault diagnosis method for a cascaded H-bridge converter according to claim 4, characterized in that: The capacitor voltage threshold is specifically set as follows: Where, is the capacitor voltage reference value, α is the capacitor voltage threshold coefficient, which is 1.3-1.5; ΔV C is the inherent fluctuation of capacitor voltage due to power pulsation, which can be expressed as:

7. The fault diagnosis method for a cascaded H-bridge converter according to claim 1, characterized in that: Step S30 is described as changing the driving signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the driving signal to short-circuit the faulty module, specifically: Adjust the drive signal of the fault module so that the switch tube S1 and the switch tube S3 of the fault module remain turned on, and the switch tube S2 and the switch S4 remain turned off; Monitor the capacitor voltage change trend of the faulty module. If the capacitor voltage remains unchanged or slowly decreases, it is determined that the switch tubes S1 and S3 are normally turned on and the faulty module is short-circuited. Otherwise, adjust the drive signal of the faulty module so that the switch tubes S2 and S4 of the faulty module remain turned on, and the switch tubes S1 and S3 remain turned off.

8. The fault diagnosis method for a cascaded H-bridge converter according to claim 1, characterized in that: The step S30 utilizes the redundancy of the modulation coefficients of other modules to restore the system to stable operation, specifically: Assuming the current of the system remains unchanged during normal operation, the single-phase system operation must meet the following requirements: v a =v g +jωLi g Where v a is the output voltage of the cascaded H-bridge, v g is the grid voltage, i g is the grid current; where v a The output voltage of each H-bridge is superimposed: Express v in terms of modulation i : in i =(m di +m qi )V ref Where m di and m qi are the modulation coefficients of the active and reactive components of the i-th H-bridge respectively; And the modulation coefficient must meet the constraints: v a It can be expressed as: Then we have: Therefore, the active and reactive power output by the cascaded H-bridge multilevel converter can be expressed as: Assuming the system is operating normally, i g Remain unchanged, that is: Under the premise that the modulation coefficient of the system is redundant, after a single module fails, the modulation coefficients of the remaining modules are increased so that the cascaded H-bridge converters can output active and reactive power that ensures stable operation of the system.

9. A fault diagnosis method for a cascaded H-bridge converter according to claim 3 or 7, characterized in that: The step S40 is specifically as follows: (1) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S2; (2) When the switch tube S1 and the switch tube S3 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S4; (3) When the switch tube S2 and the switch tube S4 remain turned on, the fault module is short-circuited; and the suspected faulty switch tube pair determined in step S10 is S2 and S3, then the faulty switch tube is S3; (4) When the switch tube S2 and the switch tube S4 remain turned on, the faulty module is short-circuited; and the suspected faulty tube pair determined in step S10 is S1 and S4, then the faulty switch tube is S1.

10. A fault diagnosis system for a cascaded H-bridge converter, characterized in that: When the system is running, the method according to any one of claims 1 to 9 is executed, including the following modules: System fault judgment module, used to use current sensors to collect transient AC grid current and extract its DC component The DC component The absolute value of is compared with the preset DC component threshold. If it exceeds the DC component threshold, it is determined that an open circuit fault of the power device has occurred in the system, the suspected faulty tube pair is determined, and the fault module location module is continued to execute; otherwise, the execution returns to the system fault determination module; Fault module locating module, used to remove DC bus voltage balancing control from the system and locate the faulty module; The fault recovery module is used to restore the DC bus voltage balance control of the modules except the faulty module. It short-circuits the faulty module by changing the drive signal of the power device of the faulty module and monitoring the change trend of the capacitor voltage of the faulty module under the drive signal. It also uses the redundancy of the modulation coefficients of other modules to restore the system to stable operation. The faulty device locating module is used to locate the specific faulty switch tube according to the execution process of the fault recovery module and the suspected faulty tube pair determined by the system fault judgment module.

Images