A model prediction-based grid-connected converter fault-tolerant control method and system

Abstract

The application provides a model prediction-based grid-connected converter fault-tolerant control system and method, and belongs to the technical field of converter control, and comprises the following steps: determining a fault device in a fault phase of a grid-connected converter based on measured grid-connected converter alternating current side current and PCC point current; after confirming that a power device fault occurs in a bridge arm of a certain phase of the grid-connected converter, a direct current side capacitor sharing type fault-tolerant control algorithm is put into operation, and the gate signals of all power devices of the fault bridge arm are blocked, and the direct current side capacitor replaces the fault bridge arm to support the controllable operation of the converter.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of converter control, and particularly relates to a model prediction-based grid-connected converter fault-tolerant control method and system. BACKGROUND

[0002] The statements in this section merely provide background information related to the application and do not necessarily constitute the prior art.

[0003] In recent years, the power electronics industry has entered a rapid development stage. With the continuous increase of the grid-connected penetration rate of new energy in China, the types and quantities of converters applied in the power grid are increasing. At the same time, the power grid has higher and higher requirements for the working efficiency, miniaturization and reliability of power electronic equipment. At present, the S i C, G a Third-generation wide-bandgap semiconductor materials such as SiC, GaN are still under development. The second-generation power electronic device insulated gate bipolar transistor (IGBT) is widely used in power conversion and transmission due to its fast switching speed, small on-state voltage drop, small switching loss, large safe operating area, pulse current impact resistance, and small driving power.

[0004] However, in practical applications, the switching frequency of power electronic converters is as high as several hundred kilohertz. In this high-frequency working state, it is easy to cause problems such as junction temperature over-limit, performance degradation, electromagnetic interference, etc. of the switching device, thereby causing switching failure and other faults. When the converter fails, it will appear in the non-normal operating state such as phase loss and voltage reduction, and even endanger the reliability and safety of the entire grid-connected system. Therefore, the fault-tolerant operation of the grid-connected converter is crucial. In summary, the converter usually works in high-voltage, strong current, strong electromagnetic interference, high-frequency switching speed, limited heat dissipation conditions, etc. Therefore, power electronic switching device abnormalities or connection line drop cause frequent power module failures, commutation failures, and even serious economic losses. Therefore, it is necessary to study the fault-tolerant control method and device of the grid-connected converter.

[0005] The mainstream device used in the bridge arm switching device of the three-level converter is the insulated gate bipolar transistor (IGBT), and its main fault types are short circuit fault and open circuit fault. The main reason for open circuit fault is overcurrent, poor contact or drive damage. The three common fault types of three-level converter power modules are: drive signal loss, IGBT open / short circuit fault, and diode open / short circuit fault.

[0006] As an indispensable part of the open or short circuit fault response mechanism of the converter, the main goal of the fault-tolerant control strategy is to ensure the safe and controllable operation of the converter after the fault until the shutdown for maintenance. At present, the fault-tolerant control strategy for the converter is mainly divided into two categories: redundancy type and non-redundancy type.

[0007] The existing document "Karimi S, Gaillard A, Poure Pand Saadate S. FPGA-Based Real-Time Power Converter Failure Diagnosis for Wind Energy Conversion Systems [J]. IEEE Transactionson Industrial Electronics, vol. 55, no. 12, pp. 4299-4308" adopts a redundancy fault-tolerant control strategy, which is mainly divided into three types of redundancy fault-tolerant control strategies at the system level, bridge arm level and device level, and configures one or more mirror devices in the converter system. When the system is running normally, the mirror device remains silent; after the system fails, the mirror device replaces the faulty device to ensure the safe and controllable operation of the converter system. However, this strategy has high implementation cost and high space occupancy rate, and its application is limited to the field of aerospace and other extremely demanding reliability fields.

[0008] The existing document "Jlassi Iand Cardoso A. Fault-Tolerant Back-to-Back Converter for Direct-Drive PMSG Wind Turbines Using Direct Torque and Power Control Techniques [J]. IEEE Transactionson Power Electronics, vol. 34, no. 11, pp. 11215-11227" applies a fault-tolerant control strategy to a two-level back-to-back converter system under a direct control framework, connects the output ends of the three-phase bridge arms of the grid side and the motor side, and realizes fault-tolerant operation of the converter system, but does not use rolling optimization model predictive control, which cannot eliminate the influence of uncertain factors such as electromagnetic interference resistance. SUMMARY

[0009] In order to overcome the shortcomings of the above-mentioned prior art, the present application provides a grid-connected converter fault-tolerant control method based on model prediction, which adopts a non-redundant fault-tolerant control strategy—DC side capacitor sharing type fault-tolerant control strategy, which is a two-way method for balancing implementation cost and diagnostic effect.

[0010] To achieve the above object, one or more embodiments of the present application provide the following technical solutions:

[0011] In a first aspect, a model prediction-based grid-connected converter fault-tolerant control method is disclosed, comprising:

[0012] Confirming that a power device fault occurs in a certain phase bridge arm of the converter;

[0013] After that, if the fault type is open circuit, the AC side phase voltage of the converter fault phase is connected to the neutral point N of the two capacitors through the bidirectional thyristor, the remaining normal switch devices are reconfigured, and the energy storage of the DC side capacitor is utilized;

[0014] Then, two-phase four-switch voltage vector synthesis is reconfigured to continue stable operation, and the gate signals of all power devices of the fault bridge arm are blocked, and the DC side capacitor replaces the fault bridge arm to support the controllable operation of the converter.

[0015] As a further technical solution, it further comprises: when the converter is in normal operation, the bidirectional thyristors of the converter are in an off state, the AC side voltage, current and phase angle signals of the converter are obtained by measurement, the three-phase AC voltage and current are converted into static two-phase coordinate values through Clark coordinate transformation, the relationship between the current and the voltage is calculated through a state space expression, the current value at the next sampling time under the action of all possible switch states is predicted, the cost function under the action of each switch state is evaluated, the switch state that minimizes the cost function is selected, and the switch state is applied to control the PWM modulation signal of the converter.

[0016] In the above solution, when predicting the current value at the next sampling time under the action of all possible switch states: based on all possible switch combinations (000~111) of the converter, the next sampling period current value in the α-β coordinate is calculated using a three-phase voltage balance equation. For details, see (3) in step seven in the embodiment.

[0017] As a further technical solution, the process of confirming that a power device fault occurs in a certain phase bridge arm of the converter comprises:

[0018] Measuring the AC side current of the converter and the PCC point current;

[0019] Based on the measurement data, it is determined whether the first condition is met, and if the first condition is met, the converter side power monitoring value is calculated according to the switch state;

[0020] Based on the converter side power monitoring value, it is determined whether the second condition is met, and if the second condition is met, a fault positioning process is started for verification;

[0021] If the third condition is met, the phase with the minimum absolute value of the power converter side power is the fault phase, and the power device originally conducting in the fault phase is the fault device.

[0022] As a further technical solution, if the fault type is short circuit, the fuse is actuated to convert the short circuit into an open circuit.

[0023] As a further technical solution, the start-up hypothesis verification process is used for fault positioning, and the specific steps are as follows:

[0024] If open circuit faults occur in the three-phase bridge arms respectively, the pseudo-power update expression after the open circuit faults in the three-phase bridge arms is obtained in combination with the switch state of the power converter and the direction of the three-phase current at the grid side;

[0025] If the pseudo-power update value corresponding to the open circuit fault in the X phase is less than the diagnosis threshold, it is indicated that the pseudo-power expression matches the actual connection mode of the power converter, and the hypothesis is established;

[0026] Subsequently, the power device originally conducting current in the phase bridge arm is identified as the fault device, and this positioning process is called hypothesis verification.

[0027] In a second aspect, a model prediction-based fault-tolerant control system for a grid-connected converter is disclosed, comprising:

[0028] The fault device determination module is configured to confirm that a power device fault occurs in a certain phase bridge arm of the power converter;

[0029] The fault-tolerant control module is configured to, if the fault type is open circuit, connect the AC side phase voltage of the fault phase of the power converter to the neutral points N of two capacitors through bidirectional thyristors, reconfigure the remaining normal switch devices, and utilize the energy storage of the DC side capacitor;

[0030] The two-phase four-switch voltage vector synthesis is reconfigured to achieve continuous stable operation, and the gate signals of all power devices in the fault bridge arm are blocked, and the DC side capacitor replaces the fault bridge arm to support the controllable operation of the power converter.

[0031] As a further technical solution, the power converter normal operation module is configured to, when the power converter is in normal operation, the bidirectional thyristors of the power converter are in an off state, the AC side voltage, current, and phase angle signals of the power converter are obtained through measurement, the three-phase AC voltage and current are converted into static two-phase coordinate values through Clark coordinate transformation, the relationship between the current and the voltage is calculated through a state space expression, the current value at the next sampling time under the action of all possible switch states is predicted, the cost function under the action of each switch state is evaluated, the switch state that minimizes the cost function is selected, and the switch state is applied to control the PWM modulation signal of the power converter.

[0032] The above one or more technical solutions have the following beneficial effects:

[0033] In the technical solution of the application, when the converter is in normal operation, the bidirectional thyristors TRi are all in the off state, the system controller executes the standard model predictive control algorithm, after confirming that a power device fault occurs in a phase bridge arm, the controller immediately puts into the DC side capacitor shared fault-tolerant control algorithm, and at the same time blocks the gate signals of all power devices of the fault bridge arm, and the DC side capacitor replaces the fault bridge arm to support the controllable operation of the converter.

[0034] The advantages of the additional aspects of the application will be partially given in the following description, partially become obvious from the following description, or be learned by the practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0035] The accompanying drawings, which form a part of the specification, are included to provide a further understanding of the application and are incorporated herein by reference. The illustrations are shown for the purpose of enabling those skilled in the art to implement the application and are not intended to limit the scope of the application.

[0036] Figure 1 The flowchart of the implementation steps of the application is shown in the following figure:

[0037] Figure 2 The hardware structure diagram of the DC side capacitor shared fault-tolerant control converter of the embodiment of the application is shown in the following figure:

[0038] Figure 3 The fault-tolerant converter structure diagram of the C phase fault of the embodiment of the application is shown in the following figure:

[0039] Figure 4 The space voltage vector diagram of the fault-tolerant converter of the embodiment of the application is shown in the following figure:

[0040] Figure 5 The space voltage vector diagram of the fault-tolerant converter of the C phase bridge arm fault of the embodiment of the application is shown in the following figure:

[0041] Figure 6 The grid-connected current curve diagram of the converter system in the normal operation state is shown in the following figure:

[0042] Figure 7 The grid-connected voltage diagram in the normal operation state is shown in the following figure:

[0043] Figure 8 The current THD diagram in the normal operation state is shown in the following figure:

[0044] Figure 9 The grid-connected current curve diagram of the converter system in the fault-tolerant operation state is shown in the following figure:

[0045] Figure 10 The grid-connected voltage diagram in the fault-tolerant operation state is shown in the following figure:

[0046] Figure 11THD diagram of current under fault-tolerant operation;

[0047] Figure 12 Figures showing DC capacitor voltages Vdc1 and Vdc2;

[0048] Figure 13 This is a flowchart of a fault diagnosis method under the predictive control framework of this invention. Detailed Implementation

[0049] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0050] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0051] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0052] A grid-connected converter is a power electronic device that connects a DC power source to an AC power grid, converting DC power into AC power synchronized with the grid frequency and phase. It is widely used to effectively connect renewable energy generation such as photovoltaic and wind power to the public power grid. It not only achieves DC-AC conversion but also ensures that the output AC power is synchronized with the grid power, thereby guaranteeing the stable operation of the power system.

[0053] Model predictive control: Model predictive control requires establishing a reasonable predictive model for the system. At each sampling time, the controller needs to be able to combine the current measurement information and the predictive model to accurately predict the dynamics of the system in a finite time domain. By solving the optimization problem in the prediction time domain at each sampling time, the optimal control sequence that makes the predicted value of the system dynamics meet the optimization objective is obtained, and the first element of the optimal control sequence is applied to the controlled object.

[0054] Example 1

[0055] See appendix Figure 2 As shown, this embodiment discloses a DC-side capacitor-shared fault-tolerant control converter, which, when operating, includes:

[0056] During normal operation of the converter, all bidirectional thyristors TRi are in the off state. The system controller executes the standard model predictive control algorithm, using the measured AC side voltage, current, and phase angle signals to convert the three-phase AC voltage and current into stationary two-phase coordinate values ​​through Clark coordinate transformation. The relationship between current and voltage is calculated through state-space expression, predicting the current value at the next sampling under all possible switching states. The cost function of each switching state is evaluated, and the switching states Sa, Sb, and Sc that minimize the cost function are selected. The new switching states are then used to control the converter's PWM modulation signal.

[0057] The specific steps involved in the above implementation include:

[0058] 1. Signal Acquisition and Coordinate Transformation:

[0059] a. Measure and obtain AC side voltage, current, and phase angle signals.

[0060] b. Use Clark transformation to convert the three-phase voltage and current to values ​​in the αβ stationary coordinate system:

[0061] [Vα, Vβ]=Clark(Va, Vb, Vc); [Iα, Iβ]=Clark(Ia, Ib, Ic).

[0062] 2. Establish a state-space model:

[0063] a. Based on the converter topology and parameters, establish the state-space equations in the αβ coordinate system:

[0064] dx / dt = Ax + Bu;

[0065] y=Cx;

[0066] Where x is a state variable, such as current, u is an input, such as switch state and DC voltage, and y is an output.

[0067] 3. Current prediction:

[0068] a. Discretize the state-space equations to obtain a discrete prediction model;

[0069] b. For all possible switching states, there are usually 8, from 000 to 111, calculate the predicted current at the next sampling time;

[0070] I(k+1)=f(I(k), V(k), Sa, Sb, Sc, Vdc).

[0071] 4. Cost function evaluation:

[0072] a. Define the cost function J, which typically includes current tracking error and other performance metrics;

[0073] J = (Iref - I(k+1))^2 + λ·other terms;

[0074] b. For each switching state, calculate the corresponding cost function value.

[0075] 5. Optimal switch state selection:

[0076] a. Compare the cost function values ​​for all switching states;

[0077] b. Select the switch state combination (Sa, Sb, Sc)opt that minimizes the cost function J.

[0078] 6. Control signal generation and application:

[0079] a. Generate the corresponding PWM modulation signal based on the selected optimal switching state;

[0080] b. Apply the PWM signal to the power switch of the converter.

[0081] 7. Repeat in a loop:

[0082] Repeat steps 1-6 in the next control cycle.

[0083] Upon confirming a power device failure in a phase arm, the controller immediately engages a DC-side capacitor-shared fault-tolerant control algorithm. This involves connecting the AC phase voltage of the faulty phase to the neutral point N of the two capacitors via bidirectional thyristors, reconfiguring the remaining normal switching devices, and utilizing the energy storage of the DC-side capacitors. This reconstructs a two-phase four-switch voltage vector synthesis control strategy to ensure continued stable system operation, thereby achieving system fault tolerance and improving reliability. Simultaneously, the gate signals of all power devices in the faulty phase arm are blocked, and the DC-side capacitors replace the faulty phase arm to support the controllable operation of the converter.

[0084] Reconfiguration steps:

[0085] a. Activate the bidirectional thyristor: Connect the AC side of the faulty phase to the neutral point N of the DC side capacitor through the bidirectional thyristor;

[0086] b. Reconfigure the remaining switches: reconfigure the four switches of the remaining two healthy phases into a two-phase four-switch structure;

[0087] c. Update the control model: Modify the system's mathematical model to reflect the new circuit topology.

[0088] Refactoring steps:

[0089] a. Redesign the voltage vector synthesis strategy: Based on the two-phase four-switch structure, calculate a new voltage vector diagram; design a new switching sequence and PWM strategy;

[0090] b. Modify the control algorithm: Update the system model in Model Predictive Control (MPC) or other control algorithms; adjust the control objectives and constraints;

[0091] c. Design the current controller: Modify the current reference value generation method and adjust the controller parameters to adapt to the new system dynamics.

[0092] Implement new control strategies:

[0093] a. Update the PWM signal generator: Implement new switching modes and PWM strategies;

[0094] b. Adjust protection strategies: Update overcurrent, overvoltage, and other protection thresholds;

[0095] c. Implement current balance control: Ensure that the current of the remaining two phases is balanced to avoid system imbalance.

[0096] The control signals for the bidirectional thyristor are shown in Table 1.

[0097] Table 1. Bidirectional Thyristor Control Signals

[0098]

[0099] To better illustrate the technical solution of this application, the fault-tolerant conversion structure design of the grid-connected converter is as follows: Figure 2 As shown, the DC-side capacitor-shared fault-tolerant control strategy is applicable not only to two-level and three-level converters, but also to back-to-back structures and individual rectifier or inverter structures. Figure 3 Two capacitors of identical capacity and model are connected in parallel on the DC side. F1-F6 represent fast-acting fuses, which will convert a short circuit into an open circuit when a bridge arm fails. TR1-TR3 are three bidirectional thyristors, which act as connecting switches to connect the converter output terminal to the midpoint N of the two capacitors on the DC side.

[0100] See appendix Figure 1 As shown, the fault-tolerant control method for grid-connected converters based on model prediction includes:

[0101] Step 1: Measure the arm current and determine if it is within the normal range. If yes, execute standard model predictive control; otherwise, proceed to step 2.

[0102] Step Two: Steps for Model-Driven Open-Circuit Diagnosis within the Predictive Control Framework:

[0103] Model-driven open-circuit fault diagnosis technology is mainly based on the mechanistic model of the converter. It reveals the intrinsic mechanism of open-circuit faults by analyzing the evolution trajectory of various state parameters under normal and fault conditions, thus forming a fault diagnosis method. This type of method mainly uses two vectors—voltage and current—to identify open-circuit faults. Voltage detection methods for open-circuit faults are fast and accurate; however, their drawback in practical applications is the requirement for additional hardware detection equipment. Therefore, this invention presents an open-circuit fault detection method—an open-circuit diagnosis method based on Tellegen's theorem—which mainly utilizes the converter's output current. It does not rely on additional hardware equipment or voltage sensors; it only requires integrating diagnostic functions into the existing control algorithm. This hardware-free technical solution is currently the best approach, balancing cost and diagnostic effectiveness.

[0104] Space vector modulation of fault-tolerant grid-connected converters is used to control the converter. In this step, the desired reference voltage vector (Vref) is input, and the on / off control signal of the converter bridge arm switch is output.

[0105] Under normal operating conditions, the relationship between the three-phase voltage and the switch state function is as follows:

[0106]

[0107] S a S b S c These represent the states of the three-phase switches a, b, and c, respectively. When S X =1 indicates that the upper bridge arm switch is on and the lower bridge arm switch is off; S X =0 indicates that the lower bridge arm switch is on and the upper bridge arm switch is off; uan, ubn, and ucn are the three-phase output voltages of the inverter, representing the voltages of phase a, phase b, and phase c relative to the neutral point n, respectively; Vdc is the DC bus voltage, which is the input DC voltage of the inverter.

[0108] In SVPWM, the switching state of each phase arm of the inverter can be represented by a switching function. By combining different switching functions, eight basic space voltage vectors can be obtained, which are then used to synthesize the desired reference voltage vector. In the above equation, S... a S b S c These represent the states of the three-phase switches a, b, and c, respectively. When S X =1 indicates that the upper bridge arm switch is on and the lower bridge arm switch is off; S X =0 indicates that the lower bridge arm switch is on and the upper bridge arm switch is off.

[0109] When a power switch in the converter fails, a series capacitor replaces the faulty bridge arm, reconstructing a three-phase four-switch structure from a three-phase six-switch configuration. This requires rapid and accurate reconfiguration of the vectors. The fault-tolerant converter structure is as follows:Figure 2 As shown, unlike the grid-connected converter under normal operating conditions, it has only two bridge arms and four switching devices. To analyze the switching state of the fault-tolerant converter, we take a fault in the C-phase bridge arm as an example. When a fault occurs in the C-phase, point C is directly connected to the midpoint N of the two capacitors. When point N is balanced, the voltage is 0.

[0110] Figure 3 The reconstructed space voltage vector for the normal and fault operation states of the grid-connected converter is generated by... Become indivual.

[0111] Step 3: Fault diagnosis steps under the predictive control framework:

[0112] Tellegen's theorem, based on Kirchhoff's current law, states that the total power absorbed by all branches of a collector circuit is zero. During normal converter operation, the actual connection configuration remains consistent with the switching state, and the quasi-power Pg remains zero. However, an open-circuit fault in a power device disrupts this consistency. The quasi-power Pg calculated based on the switching state will no longer be zero; this characteristic can be used for open-circuit fault diagnosis. This open-circuit fault detection method exhibits strong robustness unaffected by operating conditions. The core of the open-circuit fault diagnosis algorithm based on Tellegen's theorem is real-time monitoring of the absolute value of the quasi-power of the converter system. This value is close to zero under normal operating conditions, but spikes after an open-circuit fault in the power device. The specific process is as follows: Figure 13 As shown:

[0113] (1) Initialize data.

[0114] (2) Measure the AC side current of the converter With PCC point current Calculate the difference between these two currents, Δig= -ig.

[0115] (3) Judgment condition: |Δ [k] |>thr1 and| [k] |<0.1thr1, where thr1 is set to 0.15, indicates that the current at the PCC point is very small, which initially confirms that there may be an open circuit fault;

[0116] (4) If the conditions are met, proceed to the next step; if not, return to step 2 to continue the measurement.

[0117] (5) Calculate the converter-side power monitoring value P based on the switch state Si. m P m It is a theoretical power value calculated based on the switching state Si and the measured current.

[0118] (6) Judgment conditions: | |>thr0(m∈{0,1,2,...,7}) Here, the diagnostic threshold thr0 is set to 1.00.

[0119] The apparent absolute value of the power of a converter system, when it is significantly greater than zero, means that there is a power imbalance in the system, which is usually an indication of a fault. Pm is the power value calculated based on the measured voltage, current and switching status, and |Pm| represents the absolute value of the calculated power value.

[0120] (7) If the condition is met, proceed to the next step; if not, return to step (2) to continue the measurement.

[0121] (8) Based on Table 4, initiate the hypothesis verification process to locate the fault.

[0122] (9) Judgment condition: Min| | <thr0(x∈{a,b,c}), It is the power value, where x represents the phase, which can be a, b, or c, corresponding to the three phases of a three-phase system. The superscript q indicates that this is a specific type of power calculation or measurement method, and here it corresponds to the approximate power update value for an open-circuit fault in phase X.

[0123] (10) If the condition is met, proceed to the next step; if not, return to step (2) to continue the measurement.

[0124] (11) Converter-side power | The phase with the smallest absolute value is the faulty phase.

[0125] (12) The power device that was originally conducting in the faulty phase is the faulty device.

[0126] Upon detecting a converter malfunction, the fault location process is immediately initiated. At this point, the actual connection configuration of the converter is no longer compatible with its switching state Si due to the open-circuit fault in the power devices. Assuming the open-circuit fault occurs in each of the three phase arms, the process is combined with the current converter switching state Si and the grid-side three-phase current. The direction can be found in Table 4, which provides the approximate power update expressions after each of the three phase bridge arms experiences an open-circuit fault. If the approximate power update value corresponds to an open-circuit fault in phase X. If the value is less than the diagnostic threshold thr0, it indicates that the approximate power expression is incorrect. The assumption was valid as it matched the actual connection method of the converter. Subsequently, the power device that was originally conducting current in that phase bridge arm was identified as a faulty device; this location process is called hypothesis verification.

[0127] Table 4. Power approximation of converter system Update expression after open-circuit fault

[0128]

[0129] Step 4: Based on the identified faulty components, construct a fault-tolerant model for the DC-side capacitor sharing of the grid-connected converter under a predictive control framework:

[0130]

[0131] Assuming that Sc does not participate in vector synthesis during a C-phase fault, then:

[0132] Furthermore, due to:

[0133] The relationship between the phase voltages of the three terminals a, b, and c relative to point n and the AC line voltage of the grid-connected converter can be obtained as follows:

[0134]

[0135] Depending on the different changes in the switching state Si, the line voltage U ac and U bc With S a and S b The relationship of change is shown in Table 2.

[0136] Table 2U ac and U bc With S a and S b Change Relationship

[0137]

[0138] The relationships in Table 2 can be summarized as follows:

[0139]

[0140] Point N coincides with the fault point C, therefore: , ,but

[0141]

[0142] Assumption Simplified, we get:

[0143]

[0144] After Clark coordinate transformation, the above equation is converted to the α-β coordinate system:

[0145] .

[0146] Step 5: Fault phase proportional DC bias current compensation for model predictive fault-tolerant control:

[0147] like Figure 2 As shown, since the faulty phase is directly connected to the neutral point lead of the capacitor, the neutral point voltage of the series capacitor on the DC side is unbalanced. The voltages of the upper and lower capacitors fluctuate sinusoidally and are in completely opposite phases. When the DC side neutral point voltage is unbalanced, proportional DC bias current compensation is required for the faulty phase to eliminate the capacitor voltage deviation. The relationship between the switch state and the DC capacitor current is as follows:

[0148]

[0149] Capacitor current , It can be represented as:

[0150]

[0151] For capacitor voltage and The fault phase current can be obtained by differentiating the difference. :

[0152]

[0153] The integral is:

[0154] In the above formula and This is the initial value, which is generally considered to be 0.

[0155] Capacitor voltage deviation is extracted using a low-pass filter:

[0156]

[0157] In this step, fault-proportional DC bias current compensation is primarily used to improve system performance under fault conditions. This step typically occurs after reconfiguration and refactoring, and the impact of DC-side voltage imbalance needs to be considered as part of the control strategy.

[0158] The inputs include: fault phase information, indicating which phase has failed; reference values ​​for the three-phase current during normal operation; current system status, including the current of the remaining healthy phases; DC bus voltage; and system parameters such as filter inductance and resistance.

[0159] Output: Corrected two-phase current reference values; DC bias compensation current value. Updated control objective for model predictive control.

[0160] Step 6: Establish the objective function and iterate through the optimal switching state Si to obtain the basic space voltage vector relationship for fault-tolerant operation based on neutral point voltage balance, as shown in Table 3.

[0161] Table 3. Basic Space Voltage Vector Table for Fault-Tolerant Operation Based on Neutral Point Voltage Balance

[0162]

[0163] Based on Table 3, the following can be drawn: Figure 4 The space voltage vector distribution diagram shown has a rhombus-shaped modulation region, divided into four right-angled sectors by four basic space voltage vectors (vectors V0, V1, V2, and V3 are assigned to sectors I, II, III, and IV, respectively). These four non-zero vectors include two short vectors of equal magnitude and two long vectors of equal magnitude and opposite directions. The rhombus-shaped region formed by connecting the vertices of these vectors and symmetrically distributed about the origin represents the maximum modulation range of SVPWM. The inscribed circle of this rhombus represents the trajectory of the space voltage vector under maximum linear modulation. The reference voltage satisfies:

[0164]

[0165] From the above equation, it can be seen that to achieve neutral point balance, the maximum amplitude of the three-phase AC voltage in the linear modulation region under fault-tolerant conditions is [value missing]. Compared with SVPWM modulation under normal operating conditions, the DC bus voltage utilization is reduced by half, and it is only suitable for short-term fault operation.

[0166] This embodiment employs model predictive control, implementing a rolling time-domain strategy to solve the robust MPC optimization problem in each control cycle, applying only the optimal control input at the current moment. The predictive model considering uncertainties is typically constructed in the following form: x(k+1) = Ax(k) + Bu(k) + w(k), where w(k) represents the uncertainty and disturbance terms.

[0167] Step 7: Predict fault-tolerant current control using the grid-connected converter model, and output control commands to the converter switching IGBTs:

[0168] (1) Establish the state equations in the stationary three-phase coordinate system:

[0169] ;

[0170] (2) Perform Clark transform: ;

[0171] (3) Discretized form of load current prediction:

[0172] ;

[0173] (4) Minimize the cost function:

[0174] ;

[0175] (5) Delay compensation: ;

[0176]

[0177] When the sampling frequency is much higher than the grid frequency, the grid voltage is assumed to satisfy: .

[0178] In the above formula, i α (k+2), i β (k+2): Current component in the α-β coordinate system at time k+2. Ts: Sampling period. L: System inductance. u α (k), u β (k): u at time k an u bn u cn Voltage components of three-phase voltage in the α-β coordinate system. α (k+1), e β (k+1): At time k+1, the voltage components of the three-phase AC voltages a, b, and c on the grid side in the α-β coordinate system; R: System resistance; L: Inductance; i a i b i c Three-phase currents a, b, and c; u an u bn u cn The voltages at the neutral point of the three phases a, b, and c; e a e b e c a: Three-phase AC voltage; R: Resistance; i: Voltage on the three-phase AC side. α i β Current components in the α-β coordinate system; u α u β :u an u bn u cn Voltage components of three-phase voltage in the α-β coordinate system; e α e β a, b, c: Voltage components of the three-phase AC voltage on the grid side in the α-β coordinate system; k: Current sampling time; k+1: Next sampling time; Ts: Sampling period; u α (k), u β (k): α-β voltage at time k; e α (k), e β (k): α-β back electromotive force at time k; iα (k), i β (k): α-β current at time k; g i : Cost function; i α (k+1), i β (k+1): Reference value of α-β current at the next moment; i α (k+1), i β (k+1): The predicted α-β current value at the next moment.

[0179] Simulation verification experiment:

[0180] The main circuit includes the following components: DC power supply, voltmeter, IGBT converter, filter circuit, voltage and current measuring elements, and three-phase AC power grid.

[0181] The control circuit includes the following components: PLL phase-locked loop, Clark transformation module, vector selection module, data conversion module, and predictive control function module.

[0182] Figures 6-8 It refers to the grid-connected current, voltage, and current distortion rate of the converter system under normal operating conditions. Figures 9-11 This refers to the grid-connected current, voltage, and current distortion rate of the converter system under fault-tolerant operation when a phase C open-circuit fault occurs. In the simulation experiment, the initial value of the current reference signal is set to 3A, and it steps to 8A at 0.06s. Figure 9 It can be seen that under fault-tolerant conditions, the three-phase current waveforms are still sine waves with a phase difference of 120°, and the assigned values ​​change according to the set reference values, proving that the fault-tolerant current control under the model predictive control framework is accurate.

[0183] Figure 6 and Figure 9 The current distortion rates of the inverter during normal operation and fault-tolerant operation are 2.2% and 3.15%, respectively, indicating that the current distortion rate remains very low even during a C-phase fault. Figure 12 The curves show the voltages of the two capacitors on the DC side, Vdc1 and Vdc2. The DC power supply is 400V, so Vdc / 2 = 200V. It can be seen that the amplitudes are both over 200±7V. After current bias compensation, the fluctuations are small and the voltages are basically balanced.

[0184] The current distortion rate is calculated as follows: based on Figure 6 and Figure 9 The current waveform was analyzed using the FFT analysis tool in MATLAB's Signal Processing Toolbox, yielding harmonic analysis results. An FFT transformation was performed on the current waveform to obtain the fundamental frequency and the amplitudes of each harmonic. The fundamental frequency is 50Hz. Figure 6 and Figure 9The display shows a fundamental amplitude of 7.922, with the horizontal axis representing the harmonic order (up to 100th) and the vertical axis representing the percentage of each harmonic relative to the fundamental. MATLAB automatically calculates the root mean square value of all harmonic components, divides this value by the fundamental amplitude, and multiplies it by 100% to obtain the final THD value, i.e., the current distortion rate.

[0185] Example 2

[0186] The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described method.

[0187] Example 3

[0188] The purpose of this embodiment is to provide a computer-readable storage medium.

[0189] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the above method.

[0190] Example 4

[0191] The purpose of this embodiment is to provide a model-predictive fault-tolerant control system for grid-connected converters, including:

[0192] The fault device identification module is configured to: identify the faulty device in the faulty phase of the grid-connected converter based on the measured AC side current and PCC point current of the grid-connected converter.

[0193] The fault-tolerant control module is configured to: after confirming that a power device fault has occurred in a certain phase arm of the grid-connected converter, activate the DC-side capacitor-shared fault-tolerant control algorithm, and simultaneously block the gate signals of all power devices in the faulty arm. The DC-side capacitor replaces the faulty arm to support the controllable operation of the converter.

[0194] In terms of specific hardware implementation, two capacitors of identical capacity and model are connected in parallel on the DC side. Each switching device is connected in series with a fast-acting fuse. When a bridge arm fails, the fuse breaks, converting the short-circuit fault into an open-circuit fault. A bidirectional thyristor is connected in series between the midpoint and point N of each bridge arm, serving as a connecting switch to connect the converter output to the midpoint N of the two capacitors on the DC side. During normal operation of the converter, all bidirectional thyristors TRi are in the off state. The system controller executes the standard model predictive control algorithm. After confirming a power device failure in a certain phase bridge arm, the controller immediately switches to the DC bus shared fault-tolerant control algorithm, simultaneously blocking the gate signals of all power devices in the faulty bridge arm. The DC bus replaces the faulty bridge arm to support the controllable operation of the converter.

[0195] Model-driven open-circuit diagnostic method under predictive control framework: The open-circuit fault detection method mainly utilizes the converter's output current, without relying on additional hardware devices or voltage sensors. It only requires integrating diagnostic functions into the existing control algorithm. Based on Tellegen's theorem, this open-circuit fault diagnosis algorithm, a hardware-free solution, currently balances cost and diagnostic effectiveness. This diagnostic method and fast-acting fuses serve as backups, meaning only one of the software or hardware needs to be activated. While fuses cannot automatically reset after tripping, requiring hardware repair, the software detection method can automatically restore the normal three-phase six-switch circuit after fault recovery.

[0196] Fault-tolerant operation of grid-connected converter space voltage vector modulation: After a fault is detected, the fault-tolerant operation mode is activated to re-find the space voltage vector synthesis relationship of the grid-connected converter.

[0197] When starting the fault-tolerant operation mode, step six is ​​specifically adopted, and the spatial voltage vector synthesis relationship of the grid-connected converter is obtained by looking up Table 3.

[0198] Modeling DC-side discrete capacitor voltage for model predictive fault-tolerant control: This method is for cases where the faulty phase is directly connected to the capacitor neutral point lead, and the neutral point voltage of the DC-side series capacitor is unbalanced, with the voltages of the upper and lower capacitors fluctuating sinusoidally and having completely opposite phases. When the DC-side neutral point voltage is unbalanced, proportional DC bias current compensation is required for the faulty phase to eliminate the capacitor voltage deviation.

[0199] Grid-connected converter model predictive current control: Based on the fault-tolerant phase, the state equation is established. The real-time status of the three-phase current and voltage on the grid side is obtained using a phase-locked loop (PLL). The grid-side three-phase current / voltage is converted into the current / voltage in the stationary two-phase coordinate αβ through Clark coordinate transformation. The optimal output current is calculated through a cost function based on the filter impedance and inductive reactance, and the PWM modulation signal of the converter is selected to control the power output of the converter.

[0200] Example 5

[0201] The purpose of this embodiment is to provide a computer program product containing instructions that, when run on a computer, cause the computer to perform the methods and functions involved in any of the above embodiments.

[0202] The steps and methods involved in the apparatus of the above embodiments correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0203] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0204] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A fault-tolerant control method for grid-connected converters based on model prediction, characterized in that, include: During normal operation of the converter, all bidirectional thyristors of the converter are in the off state. The AC side voltage, current and phase angle signals of the converter are obtained by measurement. The three-phase AC voltage and current are converted into stationary two-phase coordinate values ​​through Clark coordinate transformation. The relationship between current and voltage is calculated through state space expression. The current value at the next sampling under all possible switching states is predicted. The cost function of each switching state is evaluated based on the current tracking error. The switching state that minimizes the cost function is selected and the PWM modulation signal of the converter is controlled by this switching state. A power device fault was confirmed in a phase arm of the converter. The process of confirming that a power device fault has occurred in a certain phase arm of the converter is as follows: measuring the AC side current of the converter and the current at the PCC point. The difference between the AC side current of the converter and the current at the PCC point is used to determine whether the first condition is met. If the first condition is met, the power monitoring value on the converter side is calculated based on the switch status. Based on the converter-side power monitoring value, determine whether the second condition is met. If the second condition is met, start the hypothesis verification process to locate the fault. Then determine whether the third condition is met. If the third condition is met, the phase with the smallest absolute value of the converter-side power update is the faulty phase, and the power device that was originally conducting in the faulty phase is the faulty device. Afterwards, if the fault type is open circuit, the AC phase voltage of the faulty phase of the converter is connected to the neutral point N of the two capacitors through a bidirectional thyristor, the remaining normal switching devices are reconfigured and the energy storage of the DC side capacitor is utilized. Then, the two-phase four-switch voltage vector synthesis is reconstructed to achieve continued stable operation. At the same time, the gate signals of all power devices in the faulty bridge arm are blocked, the DC-side capacitor replaces the faulty bridge arm to support the controllable operation of the converter, and the faulty phase is compensated by proportional DC bias current to eliminate capacitor voltage deviation.

2. The fault-tolerant control method for grid-connected converters based on model prediction as described in claim 1, characterized in that, If a power device fault is confirmed in a phase arm of the converter, and the fault type is a short circuit, the fuse will trip to convert the short circuit into an open circuit.

3. The fault-tolerant control method for grid-connected converters based on model prediction as described in claim 1, characterized in that, The hypothesis verification process is used to locate faults. The specific steps are as follows: Assuming that open-circuit faults occur on each of the three-phase bridge arms, and combining the current state of the converter switch and the direction of the three-phase current on the grid side, find the approximate power update expression after each of the three-phase bridge arms has an open-circuit fault. If the apparent power update value corresponding to the open circuit fault of phase X is less than the diagnostic threshold, it means that the apparent power update expression matches the actual connection method of the converter, and the assumption is valid. Subsequently, the power device that was originally conducting current in that phase bridge arm was identified as a faulty device, and this location process is called hypothesis verification.

4. A model-predictive-based fault-tolerant control system for grid-connected converters, implemented using the model-predictive-based fault-tolerant control method for grid-connected converters as described in claim 1, characterized in that... include: The fault device identification module is configured to: confirm that a power device fault has occurred in a certain phase arm of the converter. The fault-tolerant control module is configured to: if the fault type is open circuit, connect the AC phase voltage of the faulty phase of the converter to the neutral point N of the two capacitors through a bidirectional thyristor, reconfigure the remaining normal switching devices and utilize the energy storage of the DC side capacitor. Reconstructing the two-phase four-switch voltage vector synthesis enables continued stable operation. At the same time, the gate signals of all power devices in the faulty bridge arm are blocked. The DC-side capacitor replaces the faulty bridge arm to support the controllable operation of the converter. The faulty phase is compensated by proportional DC bias current to eliminate capacitor voltage deviation.

5. The fault-tolerant control system for a grid-connected converter based on model prediction as described in claim 4, characterized in that, it further... include: The converter normal operation module is configured as follows: During normal operation of the converter, all bidirectional thyristors of the converter are in the off state. The AC side voltage, current and phase angle signals of the converter are obtained by measurement. The three-phase AC voltage and current are converted into stationary two-phase coordinate values ​​through Clark coordinate transformation. The relationship between current and voltage is calculated through state-space expression. The current value at the next sampling time under all possible switching states is predicted. The cost function of each switching state is evaluated based on the current tracking error. The switching state that minimizes the cost function is selected and the PWM modulation signal of the converter is controlled by this switching state.

6. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-3.

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method described in any one of claims 1-3.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it performs the steps of the method described in any one of claims 1-3 above.

Images