Current control method for modular multilevel converter based on parameter identification

Abstract

The present application provides a modular multi-level converter current control method based on parameter identification. This solution uses the recursive least squares method to identify and dynamically update the system gain coefficient online, correct the current model in real time, solve the interference of inductance drift on the control, and combine the multi-objective function to quantify the deviation of current tracking, circulating current suppression and capacitor voltage balance. It can also achieve dynamic priority adjustment of the control target through weight distribution. According to the control strategy corresponding to the multi-objective function, the switching signal is directly generated, which avoids the delay of the traditional modulation link and improves the dynamic response speed. This target sub-module selection mechanism optimizes the capacitor voltage balance and reduces the device loss. In addition, a fault switching function is provided in some implementations to ensure the system stability under extreme working conditions. While improving the current tracking accuracy, the coordinated optimization of multiple objectives is achieved, providing efficient and reliable technical support for the flexible interconnection of new energy high-penetration power grids.

Description

Technical Field

[0001] The present application relates to the field of electrical control technology, and in particular to a current control method for a modular multi-level converter based on parameter identification. Background Art

[0002] The control performance of traditional modular multilevel converter (MMC) model-predictive current control strategies relies heavily on the accuracy of the predictive model. However, during actual system operation, parameters often change over time due to device aging and environmental influences. Therefore, the actual model should be time-varying. However, accurate models cannot be obtained based on data from real industrial scenarios, making it difficult for the control system to achieve optimal performance.

[0003] Therefore, to address the problem of parameter mismatch in MMC control systems, a modular multilevel converter current control method based on parameter identification is needed to eliminate the impact of parameter mismatch on system performance and improve the effectiveness and universality of the control method. Summary of the Invention

[0004] The purpose of this application is to solve at least one of the above-mentioned technical defects, especially the technical defect that parameter mismatch in the MMC current control process in the prior art has a significant impact on system performance, resulting in insufficient effectiveness and universality of the control method.

[0005] In a first aspect, the present application provides a modular multilevel converter current control method based on parameter identification, the method being used in a target flexible interconnected system, the target flexible interconnected system comprising two groups of modular multilevel converters MMC connected back-to-back via a DC bus and two groups of three-phase AC systems corresponding to the MMCs, each of the MMC converters adopting a three-phase six-bridge arm structure, each phase comprising an upper and lower bridge arm, each bridge arm comprising a plurality of cascaded half-bridge sub-modules and a bridge arm inductor, the upper and lower bridge arms of each phase being connected and then connected to the corresponding phase bus in the corresponding three-phase AC system through a filter inductor;

[0006] The method comprises:

[0007] collecting operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determining current target parameters according to the operating parameters;

[0008] According to the target parameters, the current model is identified by a recursive least square method to determine the system gain coefficient;

[0009] The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the method;

[0010] Establishing a target prediction model according to the system gain coefficient and the target parameter, and obtaining a target parameter prediction value for the next cycle based on the target prediction model;

[0011] A multi-objective function is established according to the error corresponding to the predicted value of the target parameter, and a corresponding switching signal is generated according to the multi-objective function to control the conduction and shutdown of each bridge arm sub-module of the MMC.

[0012] As an optional implementation manner, the target parameters include output current, circulating current and submodule capacitor voltage;

[0013] The multi-objective function includes a first objective function, a second objective function, and a third objective function, and establishing the multi-objective function according to the error corresponding to the target parameter prediction value includes:

[0014] Calculating an absolute error between the output current and a reference value as a first objective function;

[0015] calculating a deviation between the circulation and a reference circulation as a second objective function;

[0016] The deviation between the submodule capacitor voltage and the rated voltage is calculated as a third objective function.

[0017] As an optional implementation manner, generating corresponding switching signals according to the multi-objective function to control the on and off of each bridge arm sub-module of the MMC includes:

[0018] Determine the output voltage level and submodule conduction strategy according to the multi-objective function, and generate corresponding switching signals to control the conduction and shutdown of the MMC bridge arm submodule;

[0019] The output voltage level and the submodule conduction strategy are used to indicate the expected output effect of the MMC on the target flexible interconnection system and the expected operating state of the MMC.

[0020] As an optional implementation manner, determining the output voltage level and the submodule conduction strategy according to the multi-objective function includes:

[0021] Determining an output voltage level according to the minimum value of the first objective function, and determining the total number of submodules that need to be turned on in the upper and lower bridge arms of the MMC according to the output voltage level;

[0022] According to the minimum value of the second objective function, adjusting the increase or decrease in the number of upper and lower bridge arm submodules of the MMC based on the circulation suppression target;

[0023] sorting the proximity between the capacitor voltage of each submodule and the rated voltage according to the value of the third objective function corresponding to each submodule to obtain a sorting result;

[0024] A target submodule combination is determined according to the total number of submodules, the increase and decrease, and the sorting result.

[0025] As an optional implementation, the method further includes:

[0026] Determining influence weights of the first objective function, the second objective function, and the third objective function;

[0027] Determining a comprehensive objective function according to the impact weights;

[0028] The output voltage level and the submodule conduction strategy are determined according to the comprehensive objective function.

[0029] As an optional implementation, the recursive least squares method includes:

[0030] Construct an observation matrix in least square format based on historical current data and historical voltage data;

[0031] The weights of the historical current data and the historical voltage data in parameter updating are dynamically adjusted through the forgetting factor, and the system gain coefficient vector and covariance matrix are iteratively calculated.

[0032] As an optional implementation, the method further includes:

[0033] determining the convergence of the recursive least squares method in real time;

[0034] If the recursive least squares method does not converge to a preset operating condition within a preset time, controlling the on and off of each bridge arm submodule of the MMC by using a standby fixed parameter model;

[0035] Fault data is recorded, and the recursive least squares method is reinitialized after the fault is resolved.

[0036] In a second aspect, the present application provides a modular multilevel converter current control device based on parameter identification, the device being used in a target flexible interconnected system, the target flexible interconnected system comprising two groups of modular multilevel converters MMC connected back-to-back via a DC bus and two groups of three-phase AC systems corresponding to the MMCs, each of the MMC converters adopting a three-phase six-bridge arm structure, each phase comprising an upper and lower bridge arm, each bridge arm comprising a plurality of cascaded half-bridge sub-modules and a bridge arm inductor, the upper and lower bridge arms of each phase being connected and then connected to the corresponding phase bus in the corresponding three-phase AC system through a filter inductor;

[0037] The device comprises:

[0038] a determination module, configured to collect operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determine current target parameters based on the operating parameters;

[0039] a processing module, configured to perform parameter identification on the current model by a recursive least square method according to the target parameters, and determine a system gain coefficient;

[0040] The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the device;

[0041] The processing module is further configured to establish a target prediction model according to the system gain coefficient and the target parameter, and obtain a target parameter prediction value for the next cycle based on the target prediction model;

[0042] The processing module is further configured to establish a multi-objective function based on the error corresponding to the target parameter prediction value, and generate corresponding switching signals to control the on and off of each bridge arm sub-module of the MMC based on the multi-objective function.

[0043] In a third aspect, the present application provides a computer device comprising one or more processors and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the steps of the method described in the first aspect are performed.

[0044] In a fourth aspect, the present application provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute the steps of the method described in the first aspect.

[0045] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0046] Based on any of the above embodiments, the corresponding method of the present application provides a topological structure of a target flexible system to implement the application scenario of the present application, and uses the recursive least squares method to dynamically identify the system parameters online, fundamentally solving the prediction error problem caused by parameter mismatch in traditional model predictive control. The traditional method relies on a fixed parameter model and cannot adapt to the parameter drift caused by device aging or environmental changes, resulting in a decrease in current tracking accuracy and insufficient circulating current suppression capability. This solution constructs a current model in the form of least squares by collecting operating parameters such as output current, output voltage and grid voltage in real time, and uses the recursive least squares method to iteratively update the system gain coefficient and dynamically correct the current prediction equation. This process can compensate for the impact of inductance value drift on the model in real time, ensuring a high degree of match between the prediction model and the actual working conditions. Compared with the traditional fixed parameter model, this method significantly improves the accuracy of current tracking and enhances the system's adaptability to dynamic working conditions. Based on multiple control objectives, a corresponding multi-objective function is established. Parameter identification results are used to generate corresponding sub-module operation strategies in real time based on the multi-objective function, generating the corresponding switching signals to control the corresponding sub-modules. The seamless integration of the parameter identification process with predictive control avoids additional modulation steps, shortens the control cycle, and thus improves dynamic response speed. By updating parameters online, the system's adaptability to complex operating conditions is enhanced, maintaining stability in the face of grid voltage fluctuations or sudden load changes, providing a key technical guarantee for the reliable operation of flexible interconnected systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0048] Figure 1 A schematic diagram of an application scenario of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0049] Figure 2 A schematic flow chart of a current control method for a modular multi-level converter based on parameter identification according to an embodiment of the present application;

[0050] Figure 3 This is an overall control block diagram corresponding to a modular multi-level converter current control method based on parameter identification provided in one embodiment of the present application;

[0051] Figure 4 A schematic diagram showing the effect of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0052] Figure 5 A schematic diagram showing the effect of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0053] Figure 6 A schematic diagram showing the effect of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0054] Figure 7 A schematic diagram showing the effect of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0055] Figure 8 A schematic diagram showing the effect of a modular multi-level converter current control method based on parameter identification provided by one embodiment of the present application;

[0056] Figure 9 This is a diagram of the internal structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0058] Finite set model predictive control (FPMC) is widely used in electrical equipment control research due to its robustness, fast dynamic response, and multi-objective integrated control capabilities. Controllability relies on the accuracy of the predictive model, but in actual operation, parameters often vary over time due to device aging and environmental influences. However, accurate models cannot be obtained without excessive computational burden, making it difficult to achieve optimal control system performance.

[0059] In order to solve the problem of parameter mismatch in model predictive control, model-free control is a feasible solution. For example, in some feasible implementations, based on a robust model predictive current control, an extended state observer is used to estimate the disturbance value caused by parameter mismatch, and the disturbance is compensated in the model, but the model still requires inductance parameters for calculation. Another feasible implementation method uses a signal tracker to reduce the estimation residual of the traditional observer and improve the system's estimation accuracy of the disturbance. In summary, according to the technical route of this type of solution, this application proposes a medium-voltage distribution network MMC model predictive current control strategy based on parameter identification for the problem of parameter mismatch in the MMC control system. The current prediction model is combined with the recursive least squares algorithm to improve the model accuracy and eliminate the impact of parameter mismatch on system performance.

[0060] like Figure 1 As shown, Figure 1 This is a schematic diagram of an application scenario corresponding to the modular multi-level converter current control method based on parameter identification provided by an embodiment of the present application, which is used to illustrate the corresponding target flexible interconnection system. The target flexible interconnection system includes two groups of modular multi-level converters MMC connected back-to-back through DC buses and two groups of three-phase AC systems corresponding to the MMCs. Each of the MMC converters adopts a three-phase six-bridge arm structure, each phase includes an upper and lower bridge arm, each bridge arm includes multiple cascaded half-bridge sub-modules and bridge arm inductors, and the upper and lower bridge arms of each phase are connected and then connected to the corresponding phase bus in the corresponding three-phase AC system through a filter inductor.

[0061] Specifically, in actual application scenarios, Figure 1 The target flexible interconnection system corresponding to the structural diagram shown may include two 10 kV three-phase AC lines, two groups of MMC converters and a DC bus, and the two MMC converters are connected in a back-to-back manner.

[0062] The two MMC converters are both three-phase six-bridge-arm structures. Taking the MMC converter on the AC line 1 side as an example, each phase consists of an upper and lower bridge arm, each bridge arm consists of X cascaded half-bridge sub-modules and a bridge arm inductor Lmj, where m is the phase connected to the AC bus, m is a, b, or c, and j represents the upper and lower bridge arms, j=1 is the upper bridge arm, and j=2 is the lower bridge arm; each phase of the MMC converter is connected to the AC bus through a filter inductor Lsm; the upper endpoints of the upper bridge arm of each phase are connected together, called the P node; the lower endpoints of the lower bridge arm of each phase are connected together, called the N node; the back-to-back MMC converter is composed of the DC sides of the two MMC converters connected through a DC bus;

[0063] The grid connection points of the AC busbar 1 of the medium-voltage distribution network flexible interconnection system are A1, B1, and C1, which are connected to the three-phase bridge arm connection points a1, b1, and c1 of the back-to-back MMC converter. The grid connection points of the AC busbar 2 are A2, B2, and C2, which are connected to the three-phase bridge arm connection points a2, b2, and c2 of the back-to-back MMC converter.

[0064] In this application, relevant examples based on actual application scenarios are provided. Figure 1 The target flexible interconnection system shown is expanded.

[0065] The technical concept corresponding to the present application is to provide a topological structure of a target flexible system to implement the application scenario of the present application, and to identify the system parameters online dynamically through the recursive least squares method, which fundamentally solves the prediction error problem caused by parameter mismatch in traditional model predictive control. The traditional method relies on a fixed parameter model and cannot adapt to the parameter drift caused by device aging or environmental changes, resulting in a decrease in current tracking accuracy and insufficient circulating current suppression capability. This solution constructs a current model in the form of least squares by collecting operating parameters such as output current, output voltage and grid voltage in real time, and uses the recursive least squares method to iteratively update the system gain coefficient and dynamically correct the current prediction equation. This process can compensate for the impact of inductance value drift on the model in real time, ensuring a high degree of match between the prediction model and the actual working conditions. Compared with the traditional fixed parameter model, this method significantly improves the accuracy of current tracking and enhances the system's adaptability to dynamic working conditions. Based on multiple control objectives, a corresponding multi-objective function is established. Parameter identification results are used to generate corresponding sub-module operation strategies in real time based on the multi-objective function, generating the corresponding switching signals to control the corresponding sub-modules. The seamless integration of the parameter identification process with predictive control avoids additional modulation steps, shortens the control cycle, and thus improves dynamic response speed. By updating parameters online, the system's adaptability to complex operating conditions is enhanced, maintaining stability in the face of grid voltage fluctuations or sudden load changes, providing a key technical guarantee for the reliable operation of flexible interconnected systems.

[0066] The method provided in this application is described in detail below based on corresponding implementation methods in some actual application scenarios.

[0067] See also Figure 2 , Figure 2 A flow chart of a modular multi-level converter current control method based on parameter identification according to an embodiment of the present application is provided. The method is applied to Figure 1 The target flexible interconnection system shown;

[0068] like Figure 2 As shown, the method includes:

[0069] S101, collecting operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determining current target parameters based on the operating parameters;

[0070] S102, performing parameter identification on the current model by recursive least squares method according to the target parameters to determine the system gain coefficient;

[0071] The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the method;

[0072] The present application solves the parameter mismatch problem through parameter identification. In the present application, parameter identification is used to indicate an implementation method for determining the actual operating state of the current system based on the actual operating conditions. In some implementation methods, the current model can be rewritten into a least squares form, so that the relevant process of parameter identification can be performed through the recursive least squares method to determine the system gain coefficient of the least squares form model, which can directly indicate the actual operating conditions associated with the operating state of each device.

[0073] S103: establishing a target prediction model according to the system gain coefficient and the target parameter, and obtaining a target parameter prediction value for the next cycle based on the target prediction model;

[0074] S104 , establishing a multi-objective function according to the error corresponding to the target parameter prediction value, and generating corresponding switching signals to control the on and off of each bridge arm sub-module of the MMC according to the multi-objective function.

[0075] This embodiment provides a target-flexible system topology for the application scenarios of this application. Using recursive least squares, the system parameters are dynamically identified online, fundamentally resolving the prediction error problem caused by parameter mismatch in traditional model predictive control. Traditional methods rely on fixed-parameter models and are unable to adapt to parameter drift caused by device aging or environmental changes, resulting in reduced current tracking accuracy and insufficient circulating current suppression. This solution constructs a least-squares current model by collecting real-time operating parameters such as output current, output voltage, and grid voltage. The system gain coefficients are then iteratively updated using recursive least squares to dynamically correct the current prediction equation. This process compensates for the impact of inductance drift on the model in real time, ensuring a close match between the prediction model and actual operating conditions. Compared to traditional fixed-parameter models, this method significantly improves current tracking accuracy and enhances the system's adaptability to dynamic operating conditions. Based on multiple control objectives, a corresponding multi-objective function is established. Based on the parameter identification results, corresponding sub-module operation strategies are generated in real time to generate corresponding switching signals to control the corresponding sub-modules. The seamless integration of the parameter identification process with predictive control avoids additional modulation steps, shortens the control cycle, and thus improves dynamic response speed. By updating parameters online, the system's adaptability to complex working conditions is improved, and it can remain stable in the face of grid voltage fluctuations or sudden load changes, providing key technical guarantees for the reliable operation of the flexible interconnected system.

[0076] As an optional implementation manner, the target parameters include output current, circulating current and submodule capacitor voltage;

[0077] The multi-objective function includes a first objective function, a second objective function, and a third objective function, and establishing the multi-objective function according to the error corresponding to the target parameter prediction value includes:

[0078] Calculating an absolute error between the output current and a reference value as a first objective function;

[0079] calculating a deviation between the circulation and a reference circulation as a second objective function;

[0080] The deviation between the submodule capacitor voltage and the rated voltage is calculated as a third objective function.

[0081] This implementation achieves comprehensive optimization of MMC control by constructing multiple objective functions (output current error, circulating current deviation, and capacitor voltage deviation). Traditional approaches often focus solely on current tracking or circulating current suppression, failing to address multi-objective collaborative optimization, leading to capacitor-voltage imbalances in submodules or excessive circulating current amplitudes. This solution defines independent objective functions for output current, circulating current, and capacitor voltage, quantifying the deviations from each control objective. The output current error (the first objective function) directly reflects current tracking accuracy, the circulating current deviation (the second objective function) suppresses unbalanced current between bridge arms and reduces device losses, and the capacitor voltage deviation (the third objective function) ensures voltage balance in submodules and avoids overvoltage risks. By jointly optimizing multiple objective functions, the system can dynamically adjust control priorities under different operating conditions. For example, prioritizing current error reduction during sudden load changes, while prioritizing circulating current suppression and capacitor voltage balance during steady-state operation. This multi-objective design not only enhances control flexibility but also reduces interference between subsystems through the synergistic effect of the objective functions, thereby improving overall system stability and efficiency.

[0082] As an optional implementation manner, generating corresponding switching signals according to the multi-objective function to control the on and off of each bridge arm sub-module of the MMC includes:

[0083] Determine the output voltage level and submodule conduction strategy according to the multi-objective function, and generate corresponding switching signals to control the conduction and shutdown of the MMC bridge arm submodule;

[0084] The output voltage level and the submodule conduction strategy are used to indicate the expected output effect of the MMC on the target flexible interconnection system and the expected operating state of the MMC.

[0085] This implementation directly generates switching signals through a multi-objective function, eliminating the complex modulation steps in traditional methods and significantly improving control efficiency. Traditional MMC control requires indirect generation of switching signals through methods such as pulse width modulation or space vector modulation, resulting in control delays and computational overhead. This implementation directly determines the output voltage level and submodule conduction strategy based on the minimum value of the multi-objective function, and generates corresponding switching signals to directly control the corresponding submodules. In specific application scenarios, the output voltage level can be dynamically adjusted by the minimum value of the first objective function to determine the total number of submodules that need to be turned on in the upper and lower bridge arms. Circulating current suppression further optimizes the increase or decrease in the number of submodules through the second objective function. The capacitor voltage sorting mechanism, the third objective function, selects the submodule combination closest to the rated voltage. This integrated control strategy not only reduces delays in intermediate links but also shortens the control cycle to a high-frequency range, significantly improving dynamic response speed. Furthermore, directly optimizing the switching signal avoids the accumulation of modulation errors and improves control accuracy. This is particularly suitable for the rapid regulation requirements of grids with a high proportion of renewable energy connected.

[0086] As an optional implementation manner, determining the output voltage level and the submodule conduction strategy according to the multi-objective function includes:

[0087] Determining an output voltage level according to the minimum value of the first objective function, and determining the total number of submodules that need to be turned on in the upper and lower bridge arms of the MMC according to the output voltage level;

[0088] According to the minimum value of the second objective function, adjusting the increase or decrease in the number of upper and lower bridge arm submodules of the MMC based on the circulation suppression target;

[0089] sorting the proximity between the capacitor voltage of each submodule and the rated voltage according to the value of the third objective function corresponding to each submodule to obtain a sorting result;

[0090] A target submodule combination is determined according to the total number of submodules, the increase and decrease, and the sorting result.

[0091] This embodiment specifically proposes a target submodule selection mechanism based on a multi-objective function, which determines the output voltage level through the minimum value of the first objective function, determines the total number of submodules that need to be turned on in the upper and lower bridge arms, and further optimizes the increase or decrease in the number of submodules through the minimum value of the second objective function to achieve circulation suppression. The capacitor voltage sorting mechanism, i.e., the third objective function, sorts the degree of proximity between the capacitor voltage of each submodule and the rated voltage, and dynamically selects the optimal submodule combination based on the total number of submodules that need to be turned on in the upper and lower bridge arms and the circulation suppression requirements. This mechanism gives priority to submodules with voltages close to the rated value, and adjusts the increase or decrease in the number of upper and lower bridge arm submodules according to the circulation suppression target. This dynamic selection strategy not only optimizes the capacitor charge and discharge balance, but also reduces the impact of voltage fluctuations on device life. In addition, the sorting mechanism reduces the submodule switching frequency, reduces switching losses, and thus improves the overall efficiency and reliability of the system.

[0092] As an optional implementation, the method further includes:

[0093] Determining influence weights of the first objective function, the second objective function, and the third objective function;

[0094] Determining a comprehensive objective function according to the impact weights;

[0095] The output voltage level and the submodule conduction strategy are determined according to the comprehensive objective function.

[0096] This embodiment determines the overall comprehensive objective function based on each objective function through weight distribution, thereby realizing dynamic priority adjustment and comprehensive regulation of multi-objective control. According to the importance of current tracking, circulating current suppression and capacitor voltage balancing, the influence weights of the first, second and third objective functions are dynamically allocated to construct a comprehensive objective function. For example, when the load suddenly changes, the weight of current tracking is increased to quickly respond to instructions, while the weight of circulating current suppression is increased during steady-state operation to reduce losses. This dynamic weight mechanism not only balances the optimization requirements of different objectives, but also avoids conflicts between objectives through weight adjustment. In addition, the design of the comprehensive objective function simplifies the optimization process, reduces the computational complexity, and enables the system to quickly complete multi-objective decision-making within a high-frequency control cycle, thereby enhancing the real-time and adaptability of the control.

[0097] As an optional implementation, the recursive least squares method includes:

[0098] Construct an observation matrix in least square format based on historical current data and historical voltage data;

[0099] The weights of the historical current data and the historical voltage data in parameter updating are dynamically adjusted through the forgetting factor, and the system gain coefficient vector and covariance matrix are iteratively calculated.

[0100] This implementation method dynamically adjusts the weight of historical data in the recursive least squares method through the forgetting factor, thereby resolving the contradiction between speed and stability in the parameter identification process. This solution introduces the forgetting factor to dynamically adjust the contribution of historical data based on the latest data. For example, when system parameters change rapidly, the weight of historical data is reduced to accelerate parameter updates, while under steady-state conditions, the weight of historical data is increased to suppress noise interference. Through the iterative update of the covariance matrix and the gain coefficient vector, the algorithm can balance identification speed and accuracy under different working conditions, solving the problems of parameter update lag and noise sensitivity. This dynamic adjustment mechanism not only improves the convergence speed of parameter identification, but also enhances the robustness of the algorithm to sudden changes in working conditions, providing a reliable guarantee for the continuous optimization of the prediction model.

[0101] As an optional implementation, the method further includes:

[0102] determining the convergence of the recursive least squares method in real time;

[0103] If the recursive least squares method does not converge to a preset operating condition within a preset time, controlling the on and off of each bridge arm submodule of the MMC by using a standby fixed parameter model;

[0104] Fault data is recorded, and the recursive least squares method is reinitialized after the fault is resolved.

[0105] This embodiment provides a fault switching and recording mechanism to solve the risk of system loss of control when parameter identification has not converged. Specifically, if there is no emergency switching mechanism when parameter identification fails, it may lead to control instability. This embodiment monitors the convergence of the recursive least squares method in real time. If it does not converge to the preset working conditions within the preset time, it automatically switches to the backup fixed parameter model to ensure continuous control operation. At the same time, the recording function of fault data provides a basis for subsequent diagnosis and algorithm optimization. For example, in the event of extreme grid disturbances or sensor failures, the system can quickly switch to the backup model to avoid current tracking failures caused by parameter mismatch. After the fault is resolved, the recursive least squares method is reinitialized to restore the online identification function. This fault-tolerant mechanism not only improves the reliability of the system, but also optimizes the long-term operating performance through data accumulation, providing double protection for MMC control under complex working conditions.

[0106] See also Figure 3 , Figure 3 This is an overall control block diagram corresponding to a modular multi-level converter current control method based on parameter identification provided in one embodiment of the present application, which is used to illustrate an example of combining various implementation methods according to actual application scenarios.

[0107] In this scenario, the main parameters can be configured as follows: the voltage level of the medium voltage AC bus 1 and 2 are both 10kV, the output power required by the MMC converter is obtained through the dispatch instruction, the DC bus voltage is 20kV, and the MMC converter bridge arm inductance L mj is 2mH, the filter inductor L sm is 2mH, the number of sub-modules in each half-bridge arm is X is 10, and the sub-module capacitance voltage rating is U c The voltage is 2000V and the control frequency is 10kHz.

[0108] As mentioned above, the control method mainly includes two parts: a parameter identification module based on the recursive least squares method and a model predictive current control module. The following is an example of the model predictive current control strategy adopted on the AC bus 1 side.

[0109] The parameter identification module based on recursive least squares method consists of the following steps:

[0110] (1) According to the MMC structure, the current equation can be derived:

[0111]

[0112] Among them, U om is the three-phase output voltage of MMC, i gm MMC output three-phase current, U sm is the three-phase voltage on the distribution network side, m is a, b, c, L m is the single-phase total inductance value;

[0113] Among them, L m =L sm +(L m 1+L m 2) / 2, ΔL m is the inductance change value.

[0114] (2) Discretizing the current equation, we can obtain the current prediction model:

[0115]

[0116] Among them, i gm (k+1) is the predicted output current value at the next moment, i gm (k), U om (k), U sm (k) is the current measured value of output current, output voltage and distribution network voltage, T s is the sampling period;

[0117] (3) The current prediction equation is a single-input single-output system and can be rewritten in the least squares format:

[0118]

[0119] Where θ(k) is the system gain coefficient vector, To collect historical data of previously sampled inputs and outputs, F is the system output error;

[0120] (4) Real-time detection of MMC output voltage U om (k), output current i gm (k), and use the recursive least squares method to estimate the system gain coefficient vector online to obtain the system gain coefficients a1 and b0. The recursive least squares method is as follows:

[0121]

[0122] in, is the gain coefficient vector that needs to be estimated; Error between system output and estimated output; To collect historical data of previously sampled inputs and outputs; is the estimated parameter vector covariance matrix; parameter update matrix; For the forgetting factor, is the identity matrix.

[0123] The control objectives of the model-predictive current control strategy based on parameter identification are composed of three parts: current tracking, internal circulating current suppression, and submodule capacitor balancing. Taking phase A as an example (phases B and C are the same as phase A), the control steps are as follows:

[0124] (1) Measure the output current i ga , circulation i za , submodule capacitor voltage U ca_xi , upper and lower bridge arm current and voltage e ax 、i ax , the prediction model is established according to the control target as follows:

[0125]

[0126] Among them, i ga (k+1) is the predicted output current value of phase A at the next moment, i ga (k), U sa (k) is the current measured value of the output current and distribution network voltage, U oa (k) is the output voltage, i za (k+1) is the predicted circulation value of phase A at the next moment, i za (k), U dc (k) are the measured values ​​of the circulating current and DC voltage of phase A, respectively, e pa (k), e na (k) is the voltage measurement value of the upper and lower bridge arms, U za (k) is the output loop voltage, U ca_xi (k+1) is the predicted value of the capacitor voltage of the submodule at the next moment of phase A, x is p, n represents the upper and lower bridge arms, i=[1,2,…,10] is the number of submodules, U ca_xi (k) is the measured value of the submodule capacitance voltage, S is the switching constant, S=1 submodule is on, S=0 submodule is off, i ax (k) is the measured value of the upper and lower bridge arm currents;

[0127] (3) Taking phase A as an example (phases B and C are the same as phase A), establish the objective function according to the control target:

[0128]

[0129] Among them, g1 is the output current objective function, i* ga The output current reference value is given as the dispatch instruction, P AC is the AC output power; g2 is the circulating current suppression objective function; g 3i is the submodule capacitor voltage balancing objective function;

[0130] (4) Calculate the minimum value g of the objective function g1 1_min, get the output voltage U required by MMC oa The voltage level is determined by U oa The voltage level determines the number of sub-modules Xi and i that need to be turned on in the upper and lower bridge arms;

[0131] (5) Calculate the minimum value g of the objective function g2 2_min , since the output voltage U oa = is the voltage difference between the upper and lower bridge arms, so the upper and lower bridge arms increase or decrease one submodule at the same time, and its output voltage U oa Remain unchanged, according to g 2_min The number of submodules that are increased or decreased simultaneously in the upper and lower arms when the circulating current is minimum can be determined as y=-1, 0, 1. It should be noted that when i=0 or i=10, y≡0.

[0132] (6) Number the submodules and calculate the objective function g 3xi Its size can reflect the closeness between the submodule capacitor voltage and the rated voltage. 3xi Sort the upper and lower bridge arm submodules, select the X-i+y and i+y submodules with the closest capacitor voltages according to the number of submodules that need to be turned on, and record the selected submodule numbers.

[0133] (7) According to the submodule number selected in step (6), the corresponding switch signal S is output a 、S b 、S c , and ultimately acts on MMC.

[0134] Figures 4 to 8 1 is a schematic diagram showing the effect of the modular multi-level converter current control method based on parameter identification corresponding to the present application, which is used to illustrate the execution effect of the aforementioned embodiment.

[0135] in, Figure 4 and Figure 5 The local waveforms of the output A-phase current under the two control strategies, namely the traditional model predictive current control and the model predictive current control based on parameter identification, are shown. By comparing the two, it can be clearly seen that there is a more significant difference between the output current and the reference value under the traditional model predictive current control, while the output current corresponding to the model predictive current control method based on parameter identification proposed in this application can better track the reference value.

[0136] Figure 6The absolute error between the A-phase output current and the reference value under the two corresponding control strategies is 0s to 0.1s for traditional model-predicted current control, and 0.1s to 0.2s for model-predicted current control based on parameter identification. The absolute error range of the output current using traditional model-predicted current control is ±5.3A. Using the control strategy provided in this application, the absolute error range of the output current is ±3.2A, a 40% reduction in output current error.

[0137] Figure 7 and Figure 8 Used to indicate the parameter identification result based on the recursive least squares method after running the control strategy provided by this application for 1 second, where: Figure 7 and Figure 8 Parameters a1 and b0, respectively, represent the output current and input voltage control parameters of the MMC converter's predicted current model. As can be seen from the figure, when the control system first starts running, the control parameters continuously change, gradually converging after 0.1s. If convergence is poor, the failover mechanism provided in this application can be employed to ensure control effectiveness.

[0138] The present application also provides a parameter identification-based modular multilevel converter current control device, which is used in a target flexible interconnected system. The target flexible interconnected system includes two groups of modular multilevel converters (MMCs) connected back-to-back via DC busbars and two groups of three-phase AC systems corresponding to the MMCs. Each of the MMC converters adopts a three-phase six-bridge arm structure, where each phase includes an upper and lower bridge arm, and each bridge arm includes multiple cascaded half-bridge sub-modules and a bridge arm inductor. The upper and lower bridge arms of each phase are connected and then connected to the corresponding phase busbar in the corresponding three-phase AC system via a filter inductor.

[0139] The device comprises:

[0140] a determination module, configured to collect operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determine current target parameters based on the operating parameters;

[0141] a processing module, configured to perform parameter identification on the current model by a recursive least square method according to the target parameters, and determine a system gain coefficient;

[0142] The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the device;

[0143] The processing module is further configured to establish a target prediction model according to the system gain coefficient and the target parameter, and obtain a target parameter prediction value for the next cycle based on the target prediction model;

[0144] The processing module is further configured to establish a multi-objective function based on the error corresponding to the target parameter prediction value, and generate corresponding switching signals to control the on and off of each bridge arm sub-module of the MMC based on the multi-objective function.

[0145] This embodiment provides a target-flexible system topology for the application scenarios of this application. Using recursive least squares, the system parameters are dynamically identified online, fundamentally resolving the prediction error problem caused by parameter mismatch in traditional model predictive control. Traditional methods rely on fixed-parameter models and are unable to adapt to parameter drift caused by device aging or environmental changes, resulting in reduced current tracking accuracy and insufficient circulating current suppression. This solution constructs a least-squares current model by collecting real-time operating parameters such as output current, output voltage, and grid voltage. The system gain coefficients are then iteratively updated using recursive least squares to dynamically correct the current prediction equation. This process compensates for the impact of inductance drift on the model in real time, ensuring a close match between the prediction model and actual operating conditions. Compared to traditional fixed-parameter models, this method significantly improves current tracking accuracy and enhances the system's adaptability to dynamic operating conditions. Based on multiple control objectives, a corresponding multi-objective function is established. Based on the parameter identification results, corresponding sub-module operation strategies are generated in real time to generate corresponding switching signals to control the corresponding sub-modules. The seamless integration of the parameter identification process with predictive control avoids additional modulation steps, shortens the control cycle, and thus improves dynamic response speed. By updating parameters online, the system's adaptability to complex working conditions is improved, and it can remain stable in the face of grid voltage fluctuations or sudden load changes, providing key technical guarantees for the reliable operation of the flexible interconnected system.

[0146] As an optional implementation manner, the target parameters include output current, circulating current and submodule capacitor voltage;

[0147] The multi-objective function includes a first objective function, a second objective function, and a third objective function. The specific manner in which the processing module establishes the multi-objective function according to the error corresponding to the target parameter prediction value includes:

[0148] Calculating an absolute error between the output current and a reference value as a first objective function;

[0149] calculating a deviation between the circulation and a reference circulation as a second objective function;

[0150] The deviation between the submodule capacitor voltage and the rated voltage is calculated as a third objective function.

[0151] This implementation achieves comprehensive optimization of MMC control by constructing multiple objective functions (output current error, circulating current deviation, and capacitor voltage deviation). Traditional approaches often focus solely on current tracking or circulating current suppression, failing to address multi-objective collaborative optimization, leading to capacitor-voltage imbalances in submodules or excessive circulating current amplitudes. This solution defines independent objective functions for output current, circulating current, and capacitor voltage, quantifying the deviations from each control objective. The output current error (the first objective function) directly reflects current tracking accuracy, the circulating current deviation (the second objective function) suppresses unbalanced current between bridge arms and reduces device losses, and the capacitor voltage deviation (the third objective function) ensures voltage balance in submodules and avoids overvoltage risks. By jointly optimizing multiple objective functions, the system can dynamically adjust control priorities under different operating conditions. For example, prioritizing current error reduction during sudden load changes, while prioritizing circulating current suppression and capacitor voltage balance during steady-state operation. This multi-objective design not only enhances control flexibility but also reduces interference between subsystems through the synergistic effect of the objective functions, thereby improving overall system stability and efficiency.

[0152] As an optional implementation manner, the specific manner in which the processing module generates corresponding switching signals according to the multi-objective function to control the on and off of each bridge arm sub-module of the MMC includes:

[0153] Determine the output voltage level and submodule conduction strategy according to the multi-objective function, and generate corresponding switching signals to control the conduction and shutdown of the MMC bridge arm submodule;

[0154] The output voltage level and the submodule conduction strategy are used to indicate the expected output effect of the MMC on the target flexible interconnection system and the expected operating state of the MMC.

[0155] This implementation directly generates switching signals through a multi-objective function, eliminating the complex modulation steps in traditional methods and significantly improving control efficiency. Traditional MMC control requires indirect generation of switching signals through methods such as pulse width modulation or space vector modulation, resulting in control delays and computational overhead. This implementation directly determines the output voltage level and submodule conduction strategy based on the minimum value of the multi-objective function, and generates corresponding switching signals to directly control the corresponding submodules. In specific application scenarios, the output voltage level can be dynamically adjusted by the minimum value of the first objective function to determine the total number of submodules that need to be turned on in the upper and lower bridge arms. Circulating current suppression further optimizes the increase or decrease in the number of submodules through the second objective function. The capacitor voltage sorting mechanism, the third objective function, selects the submodule combination closest to the rated voltage. This integrated control strategy not only reduces delays in intermediate links but also shortens the control cycle to a high-frequency range, significantly improving dynamic response speed. Furthermore, directly optimizing the switching signal avoids the accumulation of modulation errors and improves control accuracy. This is particularly suitable for the rapid regulation requirements of grids with a high proportion of renewable energy connected.

[0156] As an optional implementation manner, the processing module determines the specific manner of output voltage level and submodule conduction strategy according to the multi-objective function, including:

[0157] Determining an output voltage level according to the minimum value of the first objective function, and determining the total number of submodules that need to be turned on in the upper and lower bridge arms of the MMC according to the output voltage level;

[0158] According to the minimum value of the second objective function, adjusting the increase or decrease in the number of upper and lower bridge arm submodules of the MMC based on the circulation suppression target;

[0159] sorting the proximity between the capacitor voltage of each submodule and the rated voltage according to the value of the third objective function corresponding to each submodule to obtain a sorting result;

[0160] A target submodule combination is determined according to the total number of submodules, the increase and decrease, and the sorting result.

[0161] This embodiment specifically proposes a target submodule selection mechanism based on a multi-objective function, which determines the output voltage level through the minimum value of the first objective function, determines the total number of submodules that need to be turned on in the upper and lower bridge arms, and further optimizes the increase or decrease in the number of submodules through the minimum value of the second objective function to achieve circulation suppression. The capacitor voltage sorting mechanism, i.e., the third objective function, sorts the degree of proximity between the capacitor voltage of each submodule and the rated voltage, and dynamically selects the optimal submodule combination based on the total number of submodules that need to be turned on in the upper and lower bridge arms and the circulation suppression requirements. This mechanism gives priority to submodules with voltages close to the rated value, and adjusts the increase or decrease in the number of upper and lower bridge arm submodules according to the circulation suppression target. This dynamic selection strategy not only optimizes the capacitor charge and discharge balance, but also reduces the impact of voltage fluctuations on device life. In addition, the sorting mechanism reduces the submodule switching frequency, reduces switching losses, and thus improves the overall efficiency and reliability of the system.

[0162] As an optional implementation manner, the processing module is further configured to:

[0163] Determining influence weights of the first objective function, the second objective function, and the third objective function;

[0164] Determining a comprehensive objective function according to the impact weights;

[0165] The output voltage level and the submodule conduction strategy are determined according to the comprehensive objective function.

[0166] This embodiment determines the overall comprehensive objective function based on each objective function through weight distribution, thereby realizing dynamic priority adjustment and comprehensive regulation of multi-objective control. According to the importance of current tracking, circulating current suppression and capacitor voltage balancing, the influence weights of the first, second and third objective functions are dynamically allocated to construct a comprehensive objective function. For example, when the load suddenly changes, the weight of current tracking is increased to quickly respond to instructions, while the weight of circulating current suppression is increased during steady-state operation to reduce losses. This dynamic weight mechanism not only balances the optimization requirements of different objectives, but also avoids conflicts between objectives through weight adjustment. In addition, the design of the comprehensive objective function simplifies the optimization process, reduces the computational complexity, and enables the system to quickly complete multi-objective decision-making within a high-frequency control cycle, thereby enhancing the real-time and adaptability of the control.

[0167] As an optional implementation, the recursive least squares method includes:

[0168] Construct an observation matrix in least square format based on historical current data and historical voltage data;

[0169] The weights of the historical current data and the historical voltage data in parameter updating are dynamically adjusted through the forgetting factor, and the system gain coefficient vector and covariance matrix are iteratively calculated.

[0170] This implementation method dynamically adjusts the weight of historical data in the recursive least squares method through the forgetting factor, thereby resolving the contradiction between speed and stability in the parameter identification process. This solution introduces the forgetting factor to dynamically adjust the contribution of historical data based on the latest data. For example, when system parameters change rapidly, the weight of historical data is reduced to accelerate parameter updates, while under steady-state conditions, the weight of historical data is increased to suppress noise interference. Through the iterative update of the covariance matrix and the gain coefficient vector, the algorithm can balance identification speed and accuracy under different working conditions, solving the problems of parameter update lag and noise sensitivity. This dynamic adjustment mechanism not only improves the convergence speed of parameter identification, but also enhances the robustness of the algorithm to sudden changes in working conditions, providing a reliable guarantee for the continuous optimization of the prediction model.

[0171] As an optional implementation manner, the processing module is further configured to:

[0172] determining the convergence of the recursive least squares method in real time;

[0173] If the recursive least squares method does not converge to a preset operating condition within a preset time, controlling the on and off of each bridge arm submodule of the MMC by using a standby fixed parameter model;

[0174] Fault data is recorded, and the recursive least squares method is reinitialized after the fault is resolved.

[0175] This embodiment provides a fault switching and recording mechanism to solve the risk of system loss of control when parameter identification has not converged. Specifically, if there is no emergency switching mechanism when parameter identification fails, it may lead to control instability. This embodiment monitors the convergence of the recursive least squares method in real time. If it does not converge to the preset working conditions within the preset time, it automatically switches to the backup fixed parameter model to ensure continuous control operation. At the same time, the recording function of fault data provides a basis for subsequent diagnosis and algorithm optimization. For example, in the event of extreme grid disturbances or sensor failures, the system can quickly switch to the backup model to avoid current tracking failures caused by parameter mismatch. After the fault is resolved, the recursive least squares method is reinitialized to restore the online identification function. This fault-tolerant mechanism not only improves the reliability of the system, but also optimizes the long-term operating performance through data accumulation, providing double protection for MMC control under complex working conditions.

[0176] It should be noted that it should be understood that the division of the various modules of the above device is merely a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. Moreover, these modules can all be implemented in the form of software called by a processing element; or they can all be implemented in the form of hardware; or some modules can be implemented in the form of software called by a processing element, and some modules can be implemented in the form of hardware. For example, the processing module can be a separately established processing element, or it can be integrated into a chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called by a processing element of the above device to perform the functions of the above-mentioned module. The implementation of other modules is similar. In addition, these modules can all or partly be integrated together, or they can be implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the hardware integrated logic circuit in the processor element or by instructions in the form of software.

[0177] Schematically, as Figure 9 As shown, Figure 9 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of the present application. The computer device 300 can be provided as a server. Figure 9 Computer device 300 includes a processing component 302, which further includes one or more processors, and a memory resource represented by memory 301 for storing instructions executable by processing component 302, such as an application. The application stored in memory 301 may include one or more modules, each corresponding to a set of instructions. In addition, processing component 302 is configured to execute the instructions to perform the method of any of the above embodiments.

[0178] The computer device 300 may further include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate based on an operating system stored in the memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or the like.

[0179] Those skilled in the art will understand that Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0180] An embodiment of the present application provides a storage medium storing computer-readable instructions. When the computer-readable instructions are executed by one or more processors, the one or more processors execute a method as provided in any embodiment.

[0181] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0182] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referenced to each other.

[0183] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A modular multi-level converter current control method based on parameter identification, characterized in that: The method is used for a target flexible interconnection system; The method comprises: collecting operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determining current target parameters according to the operating parameters; The target parameters include output current, circulating current and submodule capacitor voltage; According to the target parameters, the current model is identified by a recursive least square method to determine the system gain coefficient; The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the method; Establishing a target prediction model according to the system gain coefficient and the target parameter, and obtaining a target parameter prediction value for the next cycle based on the target prediction model; Calculating the absolute error between the predicted value of the output current and the reference value as the first objective function, calculating the deviation between the predicted value of the circulating current and the reference circulating current as the second objective function, and calculating the deviation between the predicted value of the submodule capacitor voltage and the rated voltage as the third objective function; Determining an output voltage level according to the minimum value of the first objective function, and determining the total number of submodules that need to be turned on in the upper and lower bridge arms of the MMC according to the output voltage level; According to the minimum value of the second objective function, adjusting the increase or decrease in the number of upper and lower bridge arm submodules of the MMC based on the circulation suppression target; sorting the proximity between the capacitor voltage of each submodule and the rated voltage according to the value of the third objective function corresponding to each submodule to obtain a sorting result; According to the total number of submodules, the increase and decrease situation and the sorting result, a target submodule combination is determined, and a corresponding switch signal is generated to control the on and off of the MMC bridge arm submodule.

2. The method according to claim 1, characterized in that The method further comprises: Determining influence weights of the first objective function, the second objective function, and the third objective function; Determining a comprehensive objective function according to the impact weights; The output voltage level and the submodule conduction strategy are determined according to the comprehensive objective function.

3. The method according to claim 1, characterized in that The recursive least squares method includes: Construct an observation matrix in least square format based on historical current data and historical voltage data; The weights of the historical current data and the historical voltage data in parameter updating are dynamically adjusted through the forgetting factor, and the system gain coefficient vector and covariance matrix are iteratively calculated.

4. The method according to any one of claims 1 to 3, characterized in that The method further comprises: determining the convergence of the recursive least squares method in real time; If the recursive least squares method does not converge to a preset operating condition within a preset time, controlling the on and off of each bridge arm submodule of the MMC by using a standby fixed parameter model; Fault data is recorded, and the recursive least squares method is reinitialized after the fault is resolved.

5. A modular multi-level converter current control device based on parameter identification, characterized in that: The device is used in a target flexible interconnection system; The device comprises: a determination module, configured to collect operating parameters corresponding to the MMC in the target flexible interconnection system in real time, and determine current target parameters based on the operating parameters; The target parameters include output current, circulating current and submodule capacitor voltage; a processing module, configured to perform parameter identification on the current model by a recursive least square method according to the target parameters, and determine a system gain coefficient; The current model is used to indicate a least squares model determined according to the structure of the MMC, the operating parameter is used to indicate a current working state of the target flexible interconnection system, and the target parameter is used to indicate a control target corresponding to the device; The processing module is further configured to establish a target prediction model according to the system gain coefficient and the target parameter, and obtain a target parameter prediction value for the next cycle based on the target prediction model; The processing module is further configured to calculate an absolute error between the predicted value of the output current and a reference value as a first objective function, calculate a deviation between the predicted value of the circulating current and a reference circulating current as a second objective function, and calculate a deviation between the predicted value of the submodule capacitor voltage and a rated voltage as a third objective function; The processing module is further configured to determine an output voltage level according to a minimum value of the first objective function, and determine a total number of submodules of the upper and lower bridge arms of the MMC that need to be turned on according to the output voltage level; The processing module is further configured to adjust the number of upper and lower bridge arm submodules of the MMC based on a circulation suppression target according to the minimum value of the second objective function; The processing module is further configured to sort the proximity between the capacitor voltage of each submodule and the rated voltage according to the value of the third objective function corresponding to each submodule, to obtain a sorting result; The processing module is further configured to determine a target submodule combination according to the total number of submodules, the increase / decrease status, and the sorting result, and generate a corresponding switch signal to control the on / off of the MMC bridge arm submodule.

6. A computer device, characterized in that: The method comprises one or more processors and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the one or more processors, the steps of the method according to any one of claims 1 to 4 are performed.

7. A storage medium, characterized in that: The storage medium stores computer-readable instructions, which, when executed by one or more processors, enable the one or more processors to perform the steps of the method according to any one of claims 1 to 4.

Images