Method and apparatus for electromagnetic transient simulation of power electronic systems

Abstract

The application discloses an electromagnetic transient simulation method and device of a power electronic system, relates to the technical field of numerical simulation, and aims to improve the efficiency and stability of electromagnetic transient simulation. The technical scheme is as follows: a discrete working condition set corresponding to each sub-module of the power electronic system is constructed; response data of each sub-module under different discrete working condition points is respectively generated by using numerical simulation; a simulation behavior database of each sub-module is constructed based on transient sequence data and steady-state incremental data; in the process of electromagnetic transient simulation, target transient sequence data or target steady-state incremental data corresponding to each sub-module is inquired from the simulation behavior database according to real-time working condition points and real-time switching behaviors of each sub-module; the internal state variables of each sub-module are updated by using the target transient sequence data or the target steady-state incremental data, and the port electrical quantities of each updated sub-module are merged according to the topological connection relationship of each sub-module, so as to construct a system-level electromagnetic transient waveform of the power electronic system.

Description

Technical Field

[0001] This application relates to the field of numerical simulation technology, and in particular to an electromagnetic transient simulation method and apparatus for power electronic systems. Background Technology

[0002] Electromagnetic transient simulation (EMT) of power electronic systems is an important technical means for the operation analysis, equipment design, and control strategy verification of new power systems. With the increasing complexity of converter topologies and the continuous improvement of power device switching speeds, the impact of switching behavior (such as the turn-on and turn-off processes of IGBTs / MOSFETs) on system voltage and current waveforms has significantly increased, thus placing higher demands on the electromagnetic transient simulation of power electronic systems.

[0003] Currently, existing electromagnetic transient simulation techniques typically employ circuit equation solving methods based on discrete step sizes, primarily including the nodal voltage method (NVM) and the state-space method (SSM). While these two methods can reconstruct the physical logic of the circuit, they suffer from several drawbacks when dealing with large-scale power electronic topologies. Firstly, the large number of switching devices in the system directly increases simulation complexity. Secondly, the need for full iterative calculations and cumbersome switching state determination at each simulation step leads to an exponential increase in computational load compared to the system size. This directly limits the efficiency and stability of electromagnetic transient simulations, making it difficult to meet the dual requirements of high fidelity and high timeliness for large-scale system-level simulations. Summary of the Invention

[0004] In view of the above problems, this application provides an electromagnetic transient simulation method and apparatus for power electronic systems. The main purpose is to improve the efficiency and stability of electromagnetic transient simulation in order to meet the dual requirements of high fidelity and high timeliness for large-scale system-level simulation.

[0005] To solve the above-mentioned technical problems, this application proposes the following solution: In a first aspect, this application provides an electromagnetic transient simulation method for a power electronic system, the method comprising: Construct a set of discrete operating conditions corresponding to each sub-module of the power electronic system. The set of discrete operating conditions includes external input parameters and internal state variables corresponding to each discrete operating point. The discrete operating point is obtained by discretizing the external input parameters and internal state variables of the standard operating point. Numerical simulation is used to generate response data for each of the sub-modules under different discrete operating conditions. The response data includes transient sequence data in response to switch switching behavior and steady-state incremental data in response to switch holding behavior. A simulation behavior database for each submodule is constructed based on the transient sequence data and the steady-state incremental data; During the electromagnetic transient simulation of the power electronic system, the target transient sequence data or target steady-state incremental data corresponding to each sub-module are queried in the simulation behavior database according to the real-time operating point and real-time switching behavior of each sub-module. The internal state variables of each submodule are updated using the target transient sequence data or the target steady-state incremental data, and the port electrical quantities of each updated submodule are merged according to the topological connection relationship of each submodule to construct the system-level electromagnetic transient waveform of the power electronic system.

[0006] Secondly, this application provides an electromagnetic transient simulation device for a power electronic system, the device comprising: The processing unit is used to construct a set of discrete operating conditions corresponding to each sub-module of each power electronic system. The set of discrete operating conditions includes external input parameters and internal state variables corresponding to each discrete operating point. The discrete operating point is obtained by discretizing the external input parameters and internal state variables of the standard operating point. The simulation unit is used to generate response data of each of the sub-modules under different discrete operating conditions obtained by the processing unit using numerical simulation. The response data includes transient sequence data in response to switch switching behavior and steady-state incremental data in response to switch holding behavior. The first construction unit is used to construct a simulation behavior database for each of the sub-modules based on the transient sequence data and the steady-state incremental data obtained by the simulation unit. The query unit is used to query the target transient sequence data or target steady-state incremental data of each sub-module in the simulation behavior database obtained by the first construction unit, based on the real-time operating point and real-time switching behavior of each sub-module during the electromagnetic transient simulation of the power electronic system. The second construction unit is used to update the internal state variables of each sub-module using the target transient sequence data or the target steady-state incremental data obtained by the query unit, and to merge the port electrical quantities of each updated sub-module according to the topological connection relationship of each sub-module, so as to construct the system-level electromagnetic transient waveform of the power electronic system.

[0007] To achieve the above objectives, according to a third aspect of this application, a storage medium is provided, the storage medium including a stored program, wherein, when the program is executed, the device where the storage medium is located is controlled to perform the electromagnetic transient simulation method of the power electronic system described in the first aspect.

[0008] To achieve the above objectives, according to a fourth aspect of this application, a processor is provided for running a program, wherein the program executes the electromagnetic transient simulation method for the power electronic system described in the first aspect.

[0009] By employing the above technical solutions, this application provides an electromagnetic transient simulation method and apparatus for power electronic systems. By constructing a set of discrete operating conditions for each sub-module of the power electronic system and simulating each sub-module individually based on the discrete operating conditions, it achieves modular decoupling of complex topologies from a physical structure perspective. It transforms the unified solution of globally complex circuit equations into independent calculations of local sub-modules, eliminating the problem of exponential growth in computational load due to global iteration. This not only reduces the solution complexity and enhances simulation stability but also enables the simulation process to comprehensively cover various operating conditions such as rated, overload, and boundary limits, ensuring the physical integrity and comprehensiveness of subsequent simulation data. Building upon this foundation, transient sequence data and steady-state incremental data were generated using offline numerical simulation, and a simulation behavior database was constructed. This enabled a shift in simulation logic from "online nonlinear iteration" to "offline high-precision pre-storage." By finely decoupling the transient response and steady-state increment, the complex physical logic of switching was pre-extracted into deterministic simulation behaviors and stored in the database. An efficient database query mechanism replaced the cumbersome switching state determination and circuit equation iteration process, simplifying the execution logic of each simulation step and significantly improving simulation efficiency. Finally, by topologically merging the updated internal state variables to construct system-level electromagnetic transient waveforms, a logical closed loop from high-fidelity microscopic reconstruction to high-efficiency macroscopic operation was achieved, meeting the dual requirements of high fidelity and high timeliness in large-scale system simulation.

[0010] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0011] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart of an electromagnetic transient simulation method for a power electronic system provided in an embodiment of this application is shown. Figure 2 A flowchart of another electromagnetic transient simulation method for a power electronic system provided in an embodiment of this application is shown; Figure 3 This paper shows a block diagram of an electromagnetic transient simulation device for a power electronic system provided in an embodiment of this application. Figure 4 This paper presents a block diagram of another electromagnetic transient simulation device for a power electronic system provided in an embodiment of this application. Detailed Implementation

[0012] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the present application to those skilled in the art.

[0013] Currently, existing electromagnetic transient simulation techniques typically employ circuit equation solving methods based on discrete step sizes, primarily including the nodal voltage method (NVM) and the state-space method (SSM). While these two methods can reconstruct the physical logic of the circuit, they suffer from several drawbacks when dealing with large-scale power electronic topologies. Firstly, the large number of switching devices in the system directly increases simulation complexity. Secondly, the need for full iterative calculations and cumbersome switching state determination at each simulation step leads to an exponential increase in computational load compared to the system size. This directly limits the efficiency and stability of electromagnetic transient simulations, making it difficult to meet the dual requirements of high fidelity and high timeliness for large-scale system-level simulations.

[0014] The inventors discovered that large-scale power electronic systems are physically modular and decomposable, and their microscopic nonlinear responses can be accurately characterized by pre-enumerating different operating conditions. Therefore, discrete operating points can be set for each submodule of the power electronic system. Transient and steady-state data corresponding to different switching behaviors can be determined through numerical simulation and stored in a simulation database. During simulation, the pre-stored transient sequences or steady-state increments are quickly retrieved from the database to update the submodule states. Finally, these are merged at the system topology level to reconstruct the system-level transient simulation waveforms. This transforms the unified solution of globally complex circuit equations into independent calculations of local submodules, reducing solution complexity, enhancing simulation stability, and ensuring the physical integrity and comprehensiveness of subsequent simulation data. Simultaneously, it realizes the transformation of simulation logic from "online nonlinear iteration" to "offline high-precision pre-storage," replacing the cumbersome switching state discrimination and circuit equation iteration process with an efficient database query mechanism. This simplifies the execution logic of each simulation step and greatly improves simulation efficiency. This satisfies the dual requirements of high fidelity and high timeliness in large-scale system simulation.

[0015] Based on the above considerations, this application provides an electromagnetic transient simulation method for power electronic systems. This method can improve the efficiency and stability of electromagnetic transient simulation, thereby meeting the dual requirements of high fidelity and high timeliness for large-scale system-level simulation. The specific execution steps are as follows: Figure 1 As shown, it includes: 101. Construct a set of discrete operating conditions corresponding to each sub-module of the power electronic system.

[0016] The discrete operating condition set contains the external input parameters and internal state variables corresponding to each discrete operating condition point. The discrete operating condition point is obtained by discretizing the external input parameters and internal state variables of the standard operating condition point.

[0017] In this step, the original topology of the power electronic system can be modularly decomposed to obtain several independent sub-modules. Specifically, it can be divided according to functional branches, directly defining physical branches with independent PWM switching control logic and containing power switching devices and LC energy storage elements as independent sub-modules; it can also be divided according to topological units, directly encapsulating standardized topological units (half-bridge, full-bridge, DAB, LLC, MMC bridge arm units) in the power electronic system as independent sub-modules; or it can be divided according to electrical decoupling, obtaining independent sub-modules based on electrical isolation and power flow boundaries. A sub-module specifically refers to the smallest independent computational unit decomposed in the above ways. For example, in a multi-terminal DC transmission system, each converter or controlled voltage source converter can be defined as an independent sub-module.

[0018] After the submodules are decomposed, the standard operating condition range of external input parameters and internal state variables for each type of submodule throughout its entire lifecycle can be statistically analyzed. This standard operating condition range refers to the continuous multidimensional boundary space that the external input parameters and internal state variables of each type of submodule may cover under real physical operating scenarios. It is usually determined by comprehensively analyzing the factory design rated parameters, allowable overload limits, and expected boundary operating conditions of the power electronic system. The standard operating condition point refers to the instantaneous operating state corresponding to any sampling step within the above standard operating condition range. This instantaneous operating state refers to the external input parameters and internal state variables at a certain moment. Based on this, the external input parameters and internal state variables corresponding to each standard operating condition point are discretized and sampled to obtain the discrete operating condition points for each standard operating condition point. Based on this, a set of discrete operating conditions specific to each submodule is constructed. The set of discrete operating conditions contains the external input parameters and internal state variables corresponding to each discrete operating condition point, and all discrete operating condition points follow Kirchhoff's current / voltage law, device volt-ampere characteristics, and energy conservation physical constraints to ensure the physical consistency of subsequent simulation data.

[0019] During the discretization sampling process, a uniform sampling method with equal intervals can be used to uniformly distribute sampling points across the entire operating condition range; alternatively, an adaptive non-uniform encrypted sampling method can be used to encrypt sampling points in the transient sensitive range of switching and the range of sudden changes in energy storage voltage and current, and to sparsely sample in the smooth steady-state range, thereby controlling the database storage scale while ensuring the integrity of operating condition coverage; alternatively, a cluster sampling method can be used to generate representative discrete operating condition points based on historical operating condition data to reduce the number of redundant operating conditions.

[0020] It should be noted that a discrete operating point refers to a discrete slice of a power electronic system in a multi-dimensional operating state space. Specifically, external input parameters (such as source voltage, bus current, and load impedance) represent the external excitation and boundary conditions of the system, i.e., the macroscopic operating condition settings. Internal state variables (the initial values ​​of energy storage elements within a submodule, such as capacitor voltage and inductor current) represent the internal energy storage state and dynamic response results exhibited by the system under the external excitation. Therefore, a complete discrete operating point is defined by a specific set of external input parameters and their corresponding internal state variables. Combining these two allows for a unique and complete characterization of the system's physical operating state at that discrete moment, thus providing an unambiguous index dimension for subsequently constructing a high-precision simulation behavior database.

[0021] 102. Use numerical simulation to generate response data for each submodule under different discrete operating conditions.

[0022] The response data includes transient sequence data in response to switch switching behavior and steady-state incremental data in response to switch holding behavior.

[0023] In this step, for each discrete operating point in the discrete operating condition set, the switching behavior and the switching holding behavior are applied as excitations to the submodule simulation model to conduct high-precision numerical simulation. The numerical simulation solution algorithm can employ Newton's iteration method, implicit trapezoidal integration method, Runge-Kutta method, etc.

[0024] When a switching excitation is applied, the entire switching event process of the sub-power device is simulated. Starting from the initial switching state, numerical iterative simulation is continuously performed until the internal state variables of the sub-module converge to a steady state. The response sequence of the internal state variables from the initial switching state to the re-steady-state convergence is completely recorded. This response sequence is defined as the transient sequence. The transient sequences at all discrete operating points are summarized to form the transient sequence data of the sub-module. This transient sequence is used to fully characterize the strongly nonlinear transient response features of the switch.

[0025] When applying the switching holding behavior excitation, the control signals of all switching devices in the submodule are kept constant to eliminate strong nonlinear disturbances caused by switching behavior jumps. Using the preset standard simulation step size of the electromagnetic transient simulation as a benchmark, the changes in the internal state variables of the submodule within a single standard simulation step size are recorded. These changes within the step size are defined as the steady-state increment. After traversing all discrete operating points, the steady-state increment data of the submodule is summarized. The steady-state increment is used to characterize the steady-state evolution of the slow energy throughput of energy storage components such as capacitors and inductors, and the gradual evolution of node voltage and current when the switch is not activated.

[0026] 103. Construct a simulation behavior database for each submodule based on transient sequence data and steady-state incremental data.

[0027] In this step, a dedicated simulation behavior database can be built for each independent submodule, or similar submodules can be built in batches using a template-based unified database creation and reused calls. The constructed simulation behavior database can be divided into independent transient behavior databases and steady-state behavior databases, used to store transient sequence data corresponding to switch switching behavior and steady-state incremental data corresponding to switch holding behavior, respectively. The database can be deployed using various methods such as local storage, distributed storage, and lightweight cloud storage.

[0028] This simulation behavior database transforms complex physical evolution processes into a structured key-value storage model for rapid retrieval during subsequent simulations. Specifically, it uses "discrete operating point + switching behavior (switching or holding behavior)" as the query key and the corresponding "transient sequence data" or "steady-state incremental data" as the value. The database pre-stores transient sequence data and steady-state incremental data that respond to different switching behaviors at different discrete operating points. Both transient sequence data and steady-state incremental data are uniformly converted into structured data format and archived offline. In other words, the database pre-stores transient response characteristics to switching behaviors at different discrete operating points, as well as steady-state change patterns in response to holding behaviors, providing rapid retrieval and direct data access for subsequent online simulations, eliminating the need for repeated circuit equation derivation and numerical iteration calculations during simulation.

[0029] 104. During the electromagnetic transient simulation of the power electronic system, based on the real-time operating point and real-time switching behavior of each sub-module, query the target transient sequence data or target steady-state incremental data corresponding to each sub-module in the simulation behavior database.

[0030] In this step, during the electromagnetic transient simulation, the simulation process is advanced frame by frame according to the standard simulation step size. Within each simulation step size, real-time operating data of each submodule is synchronously collected, including real-time external input parameters and real-time internal state variables, which are combined to form the real-time operating point of each submodule. Simultaneously, the real-time switching control signals of each submodule are collected and analyzed to obtain the corresponding real-time switching behavior. By using "real-time operating point + real-time switching behavior" as the target query key, and querying the target value corresponding to this target query key in the simulation behavior database, the target transient sequence data or target steady-state incremental data of each submodule's response to real-time switching behavior at the real-time operating point can be obtained.

[0031] It should be noted that if the real-time operating points do not perfectly match the discrete operating points in the simulation behavior database, a mapping strategy can be set up. This can be achieved by using the "nearest neighbor search" algorithm to quickly locate the closest discrete operating point; or by using "multidimensional linear interpolation" or "radial basis function (RBF) interpolation" algorithms to calculate a more accurate target transient sequence data or target steady-state incremental data by weighting multiple neighboring discrete operating points.

[0032] 105. Update the internal state variables of each submodule using the target transient sequence data or the target steady-state incremental data, and combine the port electrical quantities of each updated submodule according to the topological connection relationship of each module to construct the system-level electromagnetic transient waveform of the power electronic system.

[0033] In this step, when the target transient sequence data is obtained, the internal state variables of each sub-module can be directly updated based on the target transient sequence data to quickly reproduce the switching transient nonlinear evolution characteristics; when the target steady-state incremental data is obtained, the internal state variables of the sub-module can be incrementally accumulated based on the target steady-state incremental data to deduce the gradual energy evolution process under steady-state conditions step by step.

[0034] After updating the internal state variables of all submodules, the port electrical quantities of each updated submodule can be combined using the topological connection relationship of each submodule to obtain the system-level electromagnetic parameters of the power electronic system, and thus obtain the system-level electromagnetic transient waveform.

[0035] It should be noted that during the convergence process, the system's constraint equations can be constructed based on port equivalent aggregation using either the nodal voltage method or the state-space method. Specifically, the port electrical quantities of each submodule are first aggregated according to topological relationships such as series, parallel, and bridge arm connections. Based on this, the nodal voltage method can be used to construct the system constraint equations, substituting the updated port electrical quantities of each submodule and the current external excitation parameters of the power electronic system into the constraint equations to solve for the system nodal voltages. Alternatively, the state-space method can be used, assembling the system state equations by associating the updated state variables of each submodule with topological relationships to solve for the system-level electromagnetic parameters. After obtaining the system-level electromagnetic parameters such as voltage, current, and power through the above methods, frame-by-frame fitting according to the simulation time series can construct and output a complete system-level electromagnetic transient waveform, covering switching transient waveforms, steady-state evolution waveforms, and transient-steady-state boundary waveforms. Simultaneously, the updated internal state variables of each submodule are used as the initial conditions for the next simulation step, and the simulation is iteratively advanced until the preset simulation termination time is reached, completing the entire process of device-level electromagnetic transient simulation of the power electronic system.

[0036] Based on the above Figure 1 As can be seen from the implementation method, the electromagnetic transient simulation method for power electronic systems provided in this application achieves modular decoupling of complex topologies from a physical structure perspective by constructing discrete operating condition sets for each sub-module of the power electronic system. This transforms the unified solution of globally complex circuit equations into independent calculations of local sub-modules, eliminating the problem of exponential growth in computational load due to global iteration. This not only reduces solution complexity and enhances simulation stability but also enables the simulation process to comprehensively cover various operating conditions such as rated, overload, and boundary limits, ensuring the physical integrity and comprehensiveness of subsequent simulation data. Based on this, transient sequence data and steady-state incremental data are generated using offline numerical simulation, and a simulation behavior database is constructed. This realizes the transformation of simulation logic from "online nonlinear iteration" to "offline high-precision pre-storage." By finely decoupling and describing transient responses and steady-state increments, the complex switching physical logic influences are pre-extracted into deterministic simulation behaviors and stored in the database. An efficient database query mechanism replaces the cumbersome switching state discrimination and circuit equation iteration process, simplifying the execution logic of each simulation step and greatly improving simulation efficiency. Finally, by performing topological merging on the updated internal state variables to construct system-level electromagnetic transient waveforms, a logical closed loop from high-fidelity microscopic reconstruction to high-efficiency macroscopic operation was achieved, satisfying the dual requirements of high fidelity and high timeliness for large-scale system simulation.

[0037] Furthermore, the preferred embodiments of this application are based on the above... Figure 1 Based on this, a detailed explanation of the electromagnetic transient simulation process for power electronic systems is provided, including the specific steps as follows: Figure 2As shown, it includes: 201. Construct the set of discrete operating conditions corresponding to each sub-module of the power electronic system.

[0038] This step combines the description of step 101 in the above method, and the same content will not be repeated here.

[0039] Furthermore, the specific implementation method for constructing the discrete operating condition set corresponding to each sub-module of the power electronic system is as follows: Based on the topology of the power electronic system, the branch with at least one switching element or at least one energy storage element is divided into several independent sub-modules; the standard operating condition range of external input parameters and internal state variables of each sub-module during operation is statistically analyzed, and the standard operating condition range includes the external input parameters and internal state variables corresponding to each standard operating condition point; within the standard operating condition range, the standard operating condition points are discretized into discrete operating condition points to construct the discrete operating condition set of each sub-module.

[0040] In this embodiment, based on the overall topology of the power electronic system, functional branches with at least one switching element or at least one energy storage element are preferentially divided into independent simulation sub-modules. This division method ensures electrical decoupling and state autonomy of each sub-module, adapting to the requirements of subsequent standardized modeling and independent library construction.

[0041] Based on this, a comprehensive statistical analysis of the complete variation range of external input parameters and internal state variables during the actual operation of each type of submodule was conducted to determine the standard operating condition range corresponding to each parameter, covering rated operation, light load, overload, and boundary limit conditions. External input parameters include, but are not limited to, source voltage, bus current, and load impedance; internal state variables mainly refer to the initial state values ​​of energy storage elements within the submodule, such as capacitor voltage and inductor current. Within the determined standard operating condition range, any one of the following methods—equal-interval sampling, adaptive non-uniform density sampling, or operating condition clustering sampling—can be used to discretize the standard operating condition points included in the submodule, generating several discretely distributed discrete operating condition points, thereby constructing a discrete operating condition set specific to each submodule. All discrete operating condition points strictly adhere to Kirchhoff's laws and the physical constraints of energy conservation, laying the physical foundation for the subsequent generation of high-precision simulation data.

[0042] By decoupling complex power electronic systems into multiple independent sub-modules based on their electrical characteristics, the limitations of traditional global unified modeling are overcome, effectively reducing the difficulty of modeling and solving large-scale topologies. Simultaneously, based on discrete sampling within standard operating conditions, it can comprehensively cover various operating points such as rated, overload, and boundary limits, ensuring the physical integrity and comprehensiveness of subsequent simulation data.

[0043] To ensure comprehensive coverage of the entire operating range of the submodule while maintaining the regularity of discrete operating point distribution and the accuracy of representing strongly nonlinear regions, the specific implementation of the discretization sampling algorithm is as follows: Points are traversed and sampled within the standard operating range at a preset fixed step interval to obtain uniformly distributed discrete operating points; for sensitive regions where the output response gradient caused by changes in external input parameters exceeds a preset threshold, the distribution density of discrete operating points is increased; for non-sensitive regions where the output response gradient caused by changes in external input parameters does not exceed the preset threshold, the distribution density of discrete operating points is decreased; based on the final distribution density, a sample operating point set is constructed.

[0044] In this embodiment, by traversing the sampling points with a fixed step size, the standard operating condition range of the submodule can be fully covered without omissions. This ensures that the discrete operating condition points are distributed in a regular and logically consistent manner, facilitating subsequent database indexing and operating condition retrieval. Using the gradient of the internal electrical output response corresponding to the external input parameter disturbance as the basis for region division, it is possible to accurately identify highly nonlinear sensitive regions in the power electronics submodule, such as power device switching transients and energy storage state abrupt changes. By increasing the density of sampling points, the transient evolution characteristics of such regions can be finely characterized, ensuring the fidelity of device-level simulation. For non-sensitive regions with smooth electrical response changes and weak nonlinear effects, appropriately reducing the sampling distribution density can effectively reduce the number of redundant discrete operating condition points, compress the storage overhead of the simulation behavior database, and reduce the computational load of offline numerical simulation database construction. This sampling method combines the regularity of uniform grid sampling with the accuracy of adaptive non-uniform sampling. While achieving complete coverage of the entire operating condition range, it also takes into account the characterization accuracy of highly nonlinear regions and the lightweight design requirements of the database, providing physically complete and quantitatively simplified sample operating conditions for subsequent data-driven electromagnetic transient simulations.

[0045] 202. At each discrete operating point, numerical simulations are performed by applying switching and holding behaviors as stimuli to each submodule.

[0046] In this step, each discrete operating point in the discrete operating case set is traversed. The external input parameters and internal state variables corresponding to the discrete operating point are used as the initial electrical conditions. Two types of switching behaviors (i.e., switch switching behavior and switch holding behavior) are applied to excite offline high-precision numerical simulation. Among them, switch switching behavior refers to a state reversal signal given by the user (e.g., instructing a switch to change from "off" to "on"), while switch holding behavior refers to maintaining the existing switch state unchanged within the current step (the on state continues to be on, or the off state continues to be off).

[0047] Numerical simulation can employ any solution algorithm, such as Newton's iteration method, implicit trapezoidal integration method, or Runge-Kutta method, to achieve high-precision numerical solutions for the switching transient process. Simulation excitation can be applied using a time-sharing independent excitation method, where switching and holding behaviors are applied sequentially and independently at the same discrete operating point without interference; alternatively, a multi-threaded parallel simulation method can be used to simultaneously simulate both types of switching behaviors at the same operating point, improving the efficiency of offline data generation.

[0048] By applying two types of switching behavior excitations to each discrete operating point, the dynamic simulation behavior data of the submodule under different discrete operating points and different switching behaviors can be fully collected, achieving full-dimensional coverage of operating conditions and switching behaviors.

[0049] 203. When applying the switching behavior, record the response sequence of the internal state variables of each submodule after the numerical iteration converges, and use the response sequence as transient sequence data.

[0050] In this step, the complete transient process of the power device's switching control signal from on to off and back to on is simulated. Starting from the initial moment of the switching state transition, numerical iterative simulation calculations are continuously executed, recording the changes in internal state variables such as node voltage, branch current, and energy storage state within each submodule at each simulation step size, until all internal state variables meet the preset convergence threshold and the system re-enters a stable operating state. The evolution trajectory of internal state variables from the initial moment of switching to the convergent steady state is completely acquired, and this continuous time trajectory is defined as the transient sequence under the corresponding operating condition and switching behavior. After traversing all discrete operating points and combinations of switching behaviors, all transient sequences are summarized to form the complete transient sequence data of the submodule.

[0051] It should be noted that various methods can be used to determine iterative convergence, such as fixed tolerance threshold determination or state variable change rate threshold determination, all of which can guarantee the physical authenticity and numerical convergence of transient sequence data.

[0052] 204. When applying the switch holding behavior, record the state increment of the internal state variable of each submodule within a single standard simulation step, and use the state increment as steady-state increment data.

[0053] In this step, the switching control signals of all power devices in the lock module are kept constant, without any state transitions, avoiding strong nonlinear disturbances caused by switching transients, and only retaining the slow energy throughput process of energy storage components such as capacitors and inductors. Using a fixed standard simulation step size preset by the electromagnetic transient simulation as the time reference, the changes in internal state variables at the start and end times of a single standard simulation step are precisely recorded, and these changes are defined as the state increments under the corresponding operating conditions and switching behaviors. After traversing all combinations of switching hold behaviors, all state increments are summarized to form the complete steady-state incremental data of the submodule.

[0054] According to the detailed implementation steps 202-204, by using switching behavior and switching holding behavior as excitations respectively, the highly nonlinear transient process of power devices and the smooth evolution process of energy storage elements are accurately distinguished, fully conforming to the real physical operating laws of power electronic systems. By completing full-condition numerical simulation offline in one go and fixing the data, the mode of step-by-step iterative calculation in online simulation is replaced, which significantly reduces the amount of calculation in online derivation while ensuring the fidelity of electromagnetic transient simulation.

[0055] 205. At each discrete operating point, based on the state signal corresponding to the applied switching behavior, as well as the external input parameters and internal state variables before the application of the switching behavior, an initial state vector is constructed. Based on the response sequence after the application of the switching behavior, a convergent state vector is constructed. The initial state vector and the convergent state vector are stored in the transient behavior database discrete operating points in a key-value pair structure.

[0056] It should be noted that the external input parameters and internal state variables before applying the switching behavior at the discrete operating point refer specifically to the external input parameters and internal state variables at that discrete operating point. The initial state vector is obtained by encoding the external input parameters and internal state variables of the discrete operating point as the initial electrical state, and then encoding the initial electrical state together with the state signal corresponding to the switching behavior applied at that discrete operating point. Furthermore, in the key-value pairs of the transient behavior database, the initial state vector is the transient query key, and the convergent state vector is the value corresponding to the transient query key.

[0057] It should be noted that in this embodiment, the simulation behavior database is subdivided into a transient behavior database and a steady-state behavior database, which are independent of each other.

[0058] In this step, the initial state vector fully integrates the initial electrical state and switch state signals of the submodule, uniquely identifying a specific transient triggering scenario. Simultaneously, the complete response sequence obtained through numerical iteration and convergence under the corresponding switch switching behavior is defined as the converged state vector. Following the key-value mapping logic, the initial state vector is used as the transient query key, and the converged state vector is used as the value corresponding to that transient query key, and these are batch-stored into the transient behavior database using a standard key-value pair structure.

[0059] In terms of encoding implementation, various methods such as floating-point vector unique hash encoding and multi-dimensional feature serialization encoding can be used to generate the initial state vector.

[0060] 206. At each discrete operating point, based on the state signal corresponding to the applied switch holding behavior, as well as the external input parameters and internal state variables before applying the switch holding behavior, an operating condition vector is constructed. The operating condition vector and the state increment after applying the switch holding behavior are stored in the steady-state behavior database in a key-value pair structure.

[0061] It should be noted that the external input parameters and internal state variables before applying the switch-hold behavior at the discrete operating point refer to the external input parameters and internal state variables of that discrete operating point. The operating condition vector is obtained by encoding the external input parameters, internal state variables, and the state signal corresponding to the applied switch-hold behavior at the discrete operating point. In the key-value pairs stored in the steady-state behavior database, the operating condition vector serves as the steady-state lookup key, and the state increment serves as the value corresponding to the steady-state lookup key.

[0062] In this step, for the construction of the steady-state behavior database, switching holding behavior can be incorporated into the encoding of the operating condition vector. Although the switching state remains fixed within a specific steady-state stage, different combinations of switching holding (e.g., the on and off states of different bridge arms) correspond to completely different circuit topologies and system state equations, leading to significant differences in steady-state evolution patterns. Therefore, fusing and encoding the applied switching holding behavior together with external input parameters and internal state variables can accurately distinguish the steady-state characteristics under different topologies, ensuring the global uniqueness of the generated operating condition vector and avoiding state confusion during queries. Furthermore, using switching behavior as the unified encoding dimension of the operating condition vector helps maintain the consistency of the transient behavior database and the steady-state behavior database in terms of data structure, thereby simplifying the unified management and query logic of the database. The state increments collected at the corresponding discrete operating condition points are extracted, and the operating condition vector is used as the steady-state query key, with the state increment as the corresponding value, and stored in the steady-state behavior database in a key-value pair structure.

[0063] In terms of encoding implementation, various methods such as floating-point vector unique hash encoding and multi-dimensional feature serialization encoding can be used to generate working condition vectors.

[0064] It should be noted that, in conjunction with steps 205-206, the steady-state behavior database and the transient behavior database can be stored locally using key-value stores, distributed indexes, or other methods to ensure efficient retrieval and matching in subsequent online simulations. The steady-state behavior database and the transient behavior database can adopt an architecture of independent partitioned storage and unified index management, together forming a complete sub-module simulation behavior database. This provides refined and structured data support for pattern identification, data lookup, and state updates in the subsequent online simulation phase.

[0065] According to the detailed implementation methods in steps 205-206, by dividing the transient behavior database and the steady-state behavior database into independent databases, transient sequence data and steady-state incremental data are classified, stored, and accessed independently. This facilitates separate database maintenance, expansion, and reuse. Furthermore, the use of a standardized key-value pair structure to organize the data further improves the efficiency of data retrieval and recall during the online simulation phase. Differentiated vector encoding rules are designed for transient and steady-state scenarios: In transient scenarios, the focus is on the high-frequency dynamics at the moment of switch switching, incorporating internal state variables, external input parameters, and switch switching behavior into the transient query dimension; in steady-state scenarios, the focus is on the macroscopic evolution during the switch holding period, incorporating external input parameters, internal state variables, and switch holding behavior into the steady-state query dimension. This targeted feature extraction method accurately represents the physical mechanisms at different stages while effectively eliminating non-sensitive redundant variables in each scenario (such as external input fluctuations during transient processes), compressing the vector dimension and the overall database storage size, and avoiding matching deviations caused by redundancy in operating condition combinations. By clearly defining the correspondence between keys and values ​​in the database, the query logic can be standardized and normalized, effectively preventing data mismatch problems during the query process and further ensuring the stability and reliability of the simulation.

[0066] To reduce database storage redundancy and improve the efficiency of modeling and reusing similar sub-modules, it is also possible to determine and jointly build a set of discrete working conditions and a database for similar sub-modules: determine whether there are multiple target sub-modules with the same topology and physical parameters in each sub-module; if so, build a single common set of discrete working conditions for the target sub-modules and establish a unified simulation behavior database.

[0067] In this implementation, submodules with identical topologies, device selections, and electrical physical parameters exhibit high homogeneity in their operating condition evolution patterns, switching transient responses, and steady-state characteristics. Therefore, it is unnecessary to perform discrete sampling, offline numerical simulation, and database construction for each submodule of the same type. By identifying and classifying similar target submodules and configuring a unique set of common discrete operating conditions and a unified simulation behavior database, the data redundancy storage problem caused by repeated sampling and database construction can be avoided at the source, significantly reducing database storage space and offline database construction computational overhead in multi-submodule scenarios. Furthermore, during subsequent electromagnetic transient simulations, all similar target submodules can share and call the same set of common discrete operating conditions and simulation behavior database to complete state updates. This enables standardized submodule models to be modeled once and reused across the entire domain, effectively simplifying the modeling process for large-scale power electronic systems, improving the overall efficiency of simulation modeling and online derivation, and ensuring the consistency and reliability of multi-submodule collaborative simulation without altering the original physical constraints and simulation accuracy.

[0068] 207. During the electromagnetic transient simulation of the power electronic system, based on the real-time operating point and real-time switching behavior of each sub-module, query the target transient sequence data or target steady-state incremental data corresponding to each sub-module in the simulation behavior database.

[0069] This step combines the description of step 104 in the above method, and the same content will not be repeated here.

[0070] Furthermore, the specific implementation method for querying the target transient sequence data or target steady-state increment data corresponding to each submodule in the simulation behavior database is as follows: The target operating mode of each submodule is determined based on the real-time switching behavior. The target operating mode includes transient mode and steady-state mode. The real-time switching behavior includes real-time switch switching behavior or real-time switch holding behavior. In transient mode, the transient behavior database is invoked. Based on the real-time internal state variables and real-time external input parameters of the real-time operating point, and the state signal corresponding to the real-time switch switching behavior, a target initial state vector is constructed. The corresponding target converged state vector is matched in the transient behavior database, and the target converged state vector is used as the target transient sequence data. In steady-state mode, the steady-state behavior database is invoked. Based on the real-time external input parameters and real-time internal state variables of the real-time operating point, and the state signal corresponding to the real-time switch holding behavior, a target operating condition vector is constructed. The corresponding target state increment is matched in the steady-state behavior database, and the target state increment is used as the target steady-state increment data.

[0071] In this implementation, based on the state changes of real-time switching behavior, two operating conditions are defined: transient mode and steady-state mode, and the corresponding database is matched to complete the query. The real-time switching behavior includes real-time switching switching behavior or real-time switching holding behavior. If the real-time switching behavior is detected as a switching behavior, the determination submodule enters transient mode and then calls the transient behavior database for subsequent query operations. If the real-time switching behavior is a holding behavior, the determination submodule enters steady-state mode and then calls the steady-state behavior database for subsequent query operations.

[0072] Following the vector encoding rules used in constructing the simulation behavior database, in transient mode, the real-time internal state variables, real-time external input parameters, and state signals corresponding to applied real-time switching behaviors of the real-time operating points are encoded to obtain the target initial state vector, i.e., the target transient query key. In steady-state mode, the real-time external input parameters, real-time internal state variables, and state signals corresponding to applied real-time switching hold behaviors of the real-time operating points are encoded to obtain the target operating condition vector, i.e., the target steady-state query key. By matching the value corresponding to the query key in the corresponding transient behavior database or steady-state behavior database, the target converged state vector corresponding to the target initial state vector or the target state increment corresponding to the target operating condition vector can be obtained. During the matching process, various matching methods can be used, such as precise key-value matching, nearest neighbor operating point retrieval, and multi-dimensional interpolation fitting, to adapt to various simulation scenarios where real-time operating points and discrete operating points completely overlap or have deviations, ensuring the accuracy and applicability of the query results. By distinguishing between transient and steady-state modes based on real-time switching behavior and matching them with a dedicated database and coding rules to complete the query, the simulation operation mode and data call logic are adaptively linked, breaking away from the rigid mode of traditional unified simulation and allowing data calls to be highly matched with the current state of the submodule.

[0073] Furthermore, the specific implementation method for determining the target operating mode of each submodule based on real-time switching behavior is as follows: obtain the real-time switching control signal of each submodule; if the switching control signal changes abruptly, the real-time switching behavior is determined to be a switching behavior, and the submodule is determined to enter the transient mode; if the switching control signal remains unchanged within a preset time and the target convergence state vector is called up, the real-time switching behavior is determined to be a holding behavior, and the submodule is determined to enter the steady-state mode.

[0074] In this embodiment, the switch control signal can be obtained through any of the following methods: PWM modulation signal edge detection, switch instruction register status reading, or real-time communication bus message parsing. If a state transition is detected in the switch control signal, the determination submodule enters transient mode and then calls the transient behavior database for subsequent query operations. If the real-time switch behavior remains unchanged for a long time and is in a held state, the determination submodule enters steady-state mode and then calls the steady-state behavior database for subsequent query operations. Simultaneously, the physical time-based judgment logic is abandoned, and the inherent simulation step size of electromagnetic transient simulation is used as the discrimination criterion. The steady-state switching condition is defined as the absence of transitions in the switch control signal for multiple consecutive steps and the completion of the current transient sequence. This better aligns with the operating mechanism of discrete step-length simulation, avoids discrimination errors caused by time threshold settings, effectively prevents erroneous switching to steady state before the transient process is completed, ensures continuous and distortion-free simulation waveforms during the transient and steady-state transition phases, and improves the overall physical consistency of the simulation.

[0075] To balance the retrieval response speed of real-time operating point matching, the fitting efficiency of smooth operating regions, and the accuracy of simulation waveforms under strongly nonlinear operating conditions, while avoiding the technical defects of insufficient matching accuracy or high computational overhead of single matching algorithms, a hierarchical adaptive operating condition fusion matching algorithm can be used for fitting and solving: When the real-time operating point cannot be accurately matched with the discrete operating point in the simulation behavior database, the set of neighboring discrete operating points around the real-time operating point is selected through nearest neighbor search; the degree of nonlinearity of the electrical response corresponding to the set of neighboring discrete operating points is quantitatively evaluated and compared with a preset threshold; if the degree of nonlinearity is less than the preset threshold, multidimensional linear interpolation is used to fit and obtain the target transient sequence data or the target steady-state increment data; if the degree of nonlinearity is greater than or equal to the preset threshold, radial basis function interpolation is switched to fit and solve the target transient sequence data or the target steady-state increment data.

[0076] In this embodiment, the rapid coarse localization of neighboring samples in the multi-dimensional operating condition space is achieved by relying on nearest neighbor search, which can significantly reduce the range of samples to be fitted and reduce the time overhead caused by global traversal retrieval. By quantifying the nonlinearity of the electrical output response within the neighborhood of the operating condition, the steady-state smooth linear region and the strongly nonlinear region of the switching transient of the power electronic submodule can be accurately distinguished. For the smooth operating condition with weak nonlinearity, multi-dimensional linear weighted interpolation with lower computational complexity is adopted to improve the online simulation inference speed while ensuring that the fitting error is controllable. For the transient operating condition with significant nonlinear characteristics, radial basis function interpolation is used to construct a nonlinear mapping model to accurately characterize the evolution law of the strongly nonlinear transient of the switch and ensure the physical fidelity of the simulation waveform. By adaptively fusing neighbor selection, linear interpolation and nonlinear interpolation, and intelligently switching the matching strategy according to the operating condition characteristics, a dynamic balance between simulation matching efficiency and fitting accuracy is achieved, effectively eliminating the matching gap between discrete operating point and real-time operating point, smoothing the electromagnetic transient simulation output waveform, and thus adapting to the device-level simulation requirements of large-scale power electronic systems under long-term and variable operating conditions.

[0077] 208. Update the internal state variables of each submodule using the target transient sequence data or the target steady-state incremental data, and combine the port electrical quantities of each updated submodule according to the topological connection relationship of each submodule to construct the system-level electromagnetic transient waveform of the power electronic system.

[0078] This step combines the description of step 105 in the above method, and the same content will not be repeated here.

[0079] Furthermore, the specific real-time method for updating the internal state variables of each submodule using the target transient sequence data or the target steady-state incremental data is as follows: when the target transient sequence data is acquired, the internal state variables of each submodule are assigned state values ​​based on the target steady-state incremental data; when the target steady-state incremental data is acquired, the internal state variables of each submodule are incremented based on the target steady-state incremental data.

[0080] In this embodiment, when the target transient sequence data is acquired, the current operating mode of the submodule can be determined to be transient mode. During the transient process, a direct assignment of the overall state is used, which can reproduce the strong nonlinear transient characteristics of the system or module at the moment of switch transition in one go, without the need for step-by-step iteration. When the target steady-state incremental data is acquired, the current operating mode of the submodule can be determined to be steady-state mode. During the steady-state process, a step-size incremental accumulation method is used to adapt to the gradual evolution of the slow energy throughput of the energy storage element, achieving high-precision extrapolation of long-term steady-state conditions with low computational load.

[0081] By adopting differentiated state update logic for different operating modes, namely relying on direct assignment to fully reproduce the strong nonlinear characteristics of the switching transient, and relying on incremental accumulation to adapt to the steady-state smooth evolution law, the high-precision requirements of transient simulation and the low computational overhead of long-term steady-state simulation are taken into account, thus achieving bidirectional optimization of simulation accuracy and operating efficiency.

[0082] Furthermore, the specific implementation method for constructing the system-level electromagnetic transient waveform of the power electronic system by synchronously merging the updated sub-modules according to the topological connection relationship of the power electronic system is as follows: Based on the updated internal state variables of each sub-module, determine the port electrical quantities of each sub-module; construct system constraint equations based on the topological connection relationship of each sub-module; substitute the port electrical quantities of each sub-module and the current external excitation parameters of the power electronic system into the system constraint equations for solving to obtain the system-level electromagnetic parameters of the power electronic system, so as to construct the system-level electromagnetic transient waveform.

[0083] In this embodiment, port electrical quantities include external electrical characteristics such as port voltage and port current exhibited by the submodule, which have a mapping relationship with internal state variables based on the submodule's physical topology. That is, the internal state variables reflect the real-time state of the energy storage elements inside the submodule, while the port electrical quantities are the external electrical characteristics exhibited by this real-time state through the submodule's topology under specific switching behavior. By combining the current switching behavior of the submodule, the preset topological logic equations are solved using the updated internal state variables, or the corresponding port feature values ​​are directly extracted from the matched target transient sequence data, thus determining the submodule's port electrical quantities. The system constraint equations can be constructed based on various methods such as the node voltage method, state-space method, and port equivalent circuit method to adapt to the coupling solution requirements of different power electronic topologies, thereby obtaining system-level electromagnetic parameters. These system-level electromagnetic parameters characterize the macroscopic physical quantities of the overall topology network operation characteristics of the power electronic system, specifically including DC bus voltage, AC three-phase current, bridge arm circulating current, and electromagnetic characteristic values ​​of key nodes. These parameters are obtained by globally solving the port electrical quantities of each submodule after merging using the system constraint equations, which can completely characterize the electromagnetic energy evolution process of the system during transient or steady-state processes. The calculated system-level electromagnetic parameters such as voltage, current, and power are arranged continuously according to the simulation time sequence, allowing for the plotting of transient change waveforms, steady-state evolution waveforms, and transient-steady-state boundary waveforms. Simultaneously, the internal state variables of the updated submodules are cached as the initial conditions for the next simulation step, and the simulation is iteratively advanced until the preset simulation termination time is reached, completing the entire electromagnetic transient simulation process of the power electronic system and fully realizing a closed-loop electromagnetic transient simulation from switching devices to the system level.

[0084] By constructing constraint equations based on the system topology connections to achieve the merging of multiple submodules, and solving the states of independent submodules simultaneously according to circuit physical constraints, the electrical coupling relationships between submodules are ensured to conform to basic circuit laws such as Kirchhoff's laws. This method can be adapted to various complex power electronic topologies, achieving seamless connection from the fine state at the submodule level to the electromagnetic transient waveform at the system level, further improving the rigor and completeness of the system-level simulation results.

[0085] To improve the reliability and safety of power electronic systems under complex operating conditions such as extreme loads and external fault impacts, after obtaining the system-level electromagnetic transient waveform of the power electronic system, effective guidance can be provided based on this waveform: Based on the system-level electromagnetic transient waveform, target electromagnetic characteristic parameters are extracted; the target electromagnetic characteristic parameters are compared with preset system safety operation thresholds to determine the current operating state of the power electronic system; if the current operating state is abnormal, the abnormal submodule causing the abnormal state is located based on the system constraint equations and topological connections; corresponding control adjustment instructions are generated, which contain handling measures corresponding to different abnormal states. These handling measures are used to guide the adjustment of the switching control logic of the abnormal submodule or trigger circuit breaker protection actions.

[0086] The target electromagnetic characteristic parameters include at least one of transient voltage peak value, transient current change rate, and harmonic distortion rate. The control adjustment instructions include handling measures corresponding to different abnormal states, which guide the adjustment of the switching control logic of the abnormal submodule or trigger circuit breaker actions.

[0087] In this embodiment, after constructing the system-level electromagnetic transient waveform of the power electronic system, time-domain and frequency-domain feature analysis can be performed on the waveform to extract target electromagnetic characteristic parameters, namely at least one of transient voltage peak value, transient current rate of change, and harmonic distortion rate. The transient voltage peak value characterizes the withstand voltage limit of the switching transistor, the transient current rate of change characterizes the short-circuit or transient impact intensity, and the harmonic distortion rate characterizes the power quality.

[0088] The extracted target electromagnetic characteristic parameters are compared with a preset system safety operation threshold. This system safety operation threshold can be set according to the target electromagnetic characteristic parameters, such as the maximum withstand voltage of the device, the maximum allowable harmonic content, and other standards. When the target electromagnetic characteristic parameters exceed the system safety operation threshold, i.e., the current operating state is determined to be abnormal, the system will trigger an anomaly tracing mechanism. Since the system-level electromagnetic transient waveform is obtained by combining and solving the topological connection relationship of each submodule and the system constraint equation, the abnormal electrical node that caused the abnormal state is found, and the specific submodule that caused the anomaly is accurately located by tracing back according to the topological connection relationship, and it is identified as the abnormal submodule.

[0089] After locating the abnormal submodule, the system can directly generate corresponding control adjustment instructions based on a pre-defined fault intervention strategy library. It should be noted that these control adjustment instructions encapsulate handling measures for different abnormal states. For example, when the abnormal state is an excessive transient voltage peak, the handling measures included in the control adjustment instructions could be adjusting the dead time of the corresponding submodule or reducing the PWM switching frequency; when the abnormal state is a sharp increase in the transient current change rate, the handling measures included in the instructions would be blocking the drive signal of the abnormal submodule or triggering a circuit breaker protection action.

[0090] By automatically outputting complete control commands from precise anomaly location to matching specific handling measures, the efficiency of guiding the safety protection of physical power electronic systems is greatly improved, thereby enhancing the operational reliability and safety of power electronic systems under complex operating conditions such as extreme loads and external fault impacts.

[0091] Furthermore, as a response to the above Figure 1-2 The implementation of the method embodiment shown in this application provides an electromagnetic transient simulation device for power electronic systems. This device is used to improve the efficiency and stability of electromagnetic transient simulation to meet the dual requirements of high fidelity and high timeliness for large-scale system-level simulation. The embodiment of this device corresponds to the aforementioned method embodiment. For ease of reading, this embodiment will not repeat the details of the aforementioned method embodiment, but it should be clear that the device in this embodiment can correspondingly implement all the contents of the aforementioned method embodiment. Specifically, as shown... Figure 3 As shown, the device includes: Processing unit 31 is used to construct a set of discrete operating conditions corresponding to each sub-module of the electronic power system. The set of discrete operating conditions includes external input parameters and internal state variables corresponding to each discrete operating point. The discrete operating point is obtained by discretizing the external input parameters and internal state variables of the standard operating point. Simulation unit 32 is used to generate response data under different discrete operating points obtained by processing unit 31 using numerical simulation at discrete operating points. The response data includes transient sequence data of each submodule in response to switch switching behavior and steady-state incremental data in response to switch holding behavior. The first construction unit 33 is used to construct a simulation behavior database for each of the sub-modules based on the transient sequence data and the steady-state incremental data obtained by the simulation unit 32. The query unit 34 is used to query the target transient sequence data or target steady-state incremental data corresponding to each sub-module in the simulation behavior database obtained by the first construction unit 33, based on the real-time operating point and real-time switching behavior corresponding to each sub-module during the electromagnetic transient simulation of the power electronic system. The second construction unit 35 is used to update the internal state variables of each sub-module using the target transient sequence data or the target steady-state incremental data obtained by the query unit 34, and to merge the port electrical quantities of each updated sub-module according to the topological connection relationship of each sub-module, so as to construct the system-level electromagnetic transient waveform of the power electronic system.

[0092] Furthermore, such as Figure 4 As shown, the processing unit 31 includes: The splitting module 311 is used to divide a branch having at least one switching element or at least one energy storage element into several independent sub-modules based on the topology of the power electronic system. The statistics module 312 is used to statistically analyze the standard operating condition range of the external input parameters and internal state variables of each sub-module obtained by the splitting module 311 during operation. The standard operating condition range includes the external input parameters and internal state variables corresponding to each standard operating condition point. Discrete module 313 is used to discretize the standard operating condition points obtained by statistical module 312 into discrete operating condition points within the standard operating condition range, and construct the discrete operating condition set of each sub-module.

[0093] Furthermore, such as Figure 4 As shown, the simulation unit 32 includes: Simulation module 321 is used to apply switching behavior and switching holding behavior as stimuli to each of the sub-modules at each of the discrete operating points to perform numerical simulation. The first recording module 322 is used to record the response sequence of the internal state variables of each sub-module after numerical iteration convergence when the switching behavior is applied during the simulation process of the simulation module 321, and to use the response sequence as the transient sequence data. The second recording module 323 is used to record the state increment of the internal state variable of each sub-module within a single standard simulation step when the switch holding behavior is applied during the simulation process of the simulation module 321, and to use the state increment as the steady-state increment data.

[0094] Furthermore, such as Figure 4 As shown, the simulation behavior database includes a transient behavior database and a steady-state behavior database; the first construction unit 33 includes: The first processing module 331 is used to construct an initial state vector based on the state signal corresponding to the applied switch switching behavior, as well as the external input parameters and internal state variables before applying the switch switching behavior, at each discrete operating point; construct a converged state vector based on the response sequence after applying the switch switching behavior; and store the initial state vector and the converged state vector in the transient behavior database discrete operating points in a key-value pair structure. The second processing module 332 is used to construct a working condition vector based on the state signal corresponding to the applied switch holding behavior, as well as the external input parameters and internal state variables before applying the switch holding behavior, at each of the discrete working condition points, and to store the working condition vector and the state increment after applying the switch holding behavior in the steady-state behavior database discrete working condition points in a key-value pair structure.

[0095] Furthermore, such as Figure 4 As shown, the query unit 34 includes: The first determining module 341 is used to determine the target operating mode of each of the sub-modules based on the real-time switching behavior. The target operating mode includes a transient mode and a steady-state mode. The real-time switching behavior includes real-time switching behavior or real-time switching holding behavior. The first matching module 342 is used to call the transient behavior database when the first determining module 341 determines that the transient mode is being used. Based on the real-time internal state variables and real-time external input parameters of the real-time operating point, as well as the state signal corresponding to the real-time switch switching behavior, it constructs a target initial state vector, matches the corresponding target converged state vector in the transient behavior database, and uses the target converged state vector as the target transient sequence data. The second matching module 343 is used to, when the first determining module 341 determines that the steady-state mode is in the steady-state mode, call the steady-state behavior database, construct a target operating condition vector based on the real-time external input parameters and real-time internal state variables of the real-time operating point, and the state signal corresponding to the real-time switch switching behavior, match the corresponding target state increment in the steady-state behavior database, and use the target state increment as the target steady-state increment data.

[0096] Furthermore, such as Figure 4 As shown, the first determining module 341 is specifically used for, Obtain the real-time switching control signals of each of the sub-modules; If the switch control signal changes abruptly, the real-time switch behavior is determined to be a switching behavior, and the submodule is determined to enter the transient mode. If the switch control signal remains unchanged within a preset time and the target convergence state vector is invoked, then the real-time switch behavior is determined to be a hold behavior, and the submodule is determined to enter the steady-state mode.

[0097] Furthermore, such as Figure 4 As shown, the second building unit 35 includes: Update module 351 is used to assign state values ​​to the internal state variables of each submodule based on the target transient sequence data when the target transient sequence data is acquired. The update module 351 is further configured to perform incremental accumulation calculation on the internal state variables of each sub-module based on the target steady-state incremental data when the target steady-state incremental data is acquired.

[0098] Furthermore, such as Figure 4 As shown, the second building unit 35 includes: The second determining module 352 is used to determine the port electrical quantities of each sub-module based on the updated internal state variables of each sub-module obtained by the updating module 351. Module 353 is used to construct system constraint equations based on the topological connection relationships of each of the sub-modules; The solver module 354 is used to solve the system constraint equations obtained by the construction module 353 by substituting the port electrical quantities of each submodule obtained by the second determining module 352 and the current external excitation parameters of the power electronic system into the system-level electromagnetic parameters of the power electronic system, so as to construct the system-level electromagnetic transient waveform.

[0099] Furthermore, embodiments of this application also provide a storage medium for storing a computer program, wherein the computer program, when running, controls the device where the storage medium is located to execute the above-described... Figure 1-2 The electromagnetic transient simulation method for power electronic systems described in the paper.

[0100] Furthermore, embodiments of this application also provide a processor for running a program, wherein the program executes the above-described... Figure 1-2 The electromagnetic transient simulation method for power electronic systems described in the paper.

[0101] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0102] It is understood that the relevant features in the above methods and apparatus can be referenced interchangeably. Furthermore, the terms "first," "second," etc., in the above embodiments are used to distinguish between embodiments and do not represent the superiority or inferiority of any particular embodiment.

[0103] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0104] The algorithms and displays provided herein are not inherently related to any particular computer, virtual system, or other device. Various general-purpose systems can also be used in conjunction with the teachings herein. The required structure for constructing such systems is apparent from the above description. Furthermore, this application is not directed to any particular programming language. It should be understood that the content of this application described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of this application.

[0105] In addition, the memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0106] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0107] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0108] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0109] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0110] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0111] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0112] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0113] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0114] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0115] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for electromagnetic transient simulation of a power electronic system, characterized in that, The method includes: Construct a set of discrete operating conditions corresponding to each sub-module of the power electronic system. The set of discrete operating conditions includes external input parameters and internal state variables corresponding to each discrete operating point. The discrete operating point is obtained by discretizing the external input parameters and internal state variables of the standard operating point. Numerical simulation is used to generate response data for each of the sub-modules under different discrete operating conditions. The response data includes transient sequence data in response to switch switching behavior and steady-state incremental data in response to switch holding behavior. A simulation behavior database for each submodule is constructed based on the transient sequence data and the steady-state incremental data; During the electromagnetic transient simulation of the power electronic system, the target transient sequence data or target steady-state incremental data corresponding to each sub-module are queried in the simulation behavior database according to the real-time operating point and real-time switching behavior of each sub-module. The internal state variables of each submodule are updated using the target transient sequence data or the target steady-state incremental data, and the port electrical quantities of each updated submodule are merged according to the topological connection relationship of each submodule to construct the system-level electromagnetic transient waveform of the power electronic system.

2. The method according to claim 1, characterized in that, The set of discrete operating conditions corresponding to each sub-module of the power electronic system includes: Based on the topology of the power electronic system, a branch having at least one switching element or at least one energy storage element is divided into several independent sub-modules. The standard operating condition range of the external input parameters and internal state variables of each submodule during operation is statistically analyzed. The standard operating condition range includes the external input parameters and internal state variables corresponding to each standard operating condition point. Within the standard operating condition range, the standard operating condition points are discretized into discrete operating condition points to construct the discrete operating condition set for each sub-module.

3. The method according to claim 1, characterized in that, The method of generating response data for each sub-module under different discrete operating conditions using numerical simulation includes: At each of the discrete operating points, numerical simulations are performed by applying switching behavior and switching holding behavior as stimuli to each of the sub-modules. When the switch-to-switch behavior is applied, the response sequence corresponding to the internal state variable of each sub-module after numerical iteration convergence is recorded, and the response sequence is used as the transient sequence data. When the switch holding behavior is applied, the state increment corresponding to the internal state variable of each submodule within a single standard simulation step is recorded, and the state increment is used as the steady-state increment data.

4. The method according to claim 3, characterized in that, The simulation behavior database includes a transient behavior database and a steady-state behavior database; the construction of the simulation behavior database for each sub-module based on the transient sequence data and the steady-state incremental data includes: At each discrete operating point, an initial state vector is constructed based on the state signal corresponding to the applied switch switching behavior, as well as the external input parameters and internal state variables before the switch switching behavior is applied. A converged state vector is constructed based on the response sequence after the switch switching behavior is applied. The initial state vector and the converged state vector are stored in the transient behavior database in a key-value pair structure. At each discrete operating point, an operating condition vector is constructed based on the state signal corresponding to the applied switch holding behavior, as well as the external input parameters and internal state variables before applying the switch holding behavior. The operating condition vector and the state increment after applying the switch holding behavior are stored in the steady-state behavior database in a key-value pair structure.

5. The method according to claim 4, characterized in that, The step of querying the simulation behavior database for the target transient sequence data or target steady-state increment data corresponding to each sub-module includes: The target operating mode of each submodule is determined based on the real-time switching behavior. The target operating mode includes transient mode and steady-state mode. The real-time switching behavior includes real-time switch switching behavior or real-time switch holding behavior. In the transient mode, the transient behavior database is invoked, and a target initial state vector is constructed based on the real-time internal state variables and real-time external input parameters of the real-time operating point, as well as the state signal corresponding to the real-time switch switching behavior. The corresponding target converged state vector is matched in the transient behavior database, and the target converged state vector is used as the target transient sequence data. In the steady-state mode, the steady-state behavior database is invoked, and a target operating condition vector is constructed based on the real-time external input parameters and real-time internal state variables of the real-time operating point, as well as the state signal corresponding to the real-time switch holding behavior. The corresponding target state increment is matched in the steady-state behavior database, and the target state increment is used as the target steady-state increment data.

6. The method according to claim 5, characterized in that, The step of determining the target operating mode of each sub-module based on the real-time switching behavior includes: Obtain the real-time switching control signals of each of the sub-modules; If the switch control signal changes abruptly, the real-time switch behavior is determined to be a switching behavior, and the submodule is determined to enter the transient mode. If the switch control signal remains unchanged within a preset time and the target convergence state vector is invoked, then the real-time switch behavior is determined to be a hold behavior, and the submodule is determined to enter the steady-state mode.

7. The method according to claim 1, characterized in that, The step of updating the internal state variables of each submodule using the target transient sequence data or the target steady-state increment data includes: When the target transient sequence data is acquired, the internal state variables of each submodule are assigned state values ​​based on the target transient sequence data; When the target steady-state incremental data is acquired, the internal state variables of each sub-module are incrementally accumulated based on the target steady-state incremental data.

8. The method according to any one of claims 1-7, characterized in that, The step of aggregating the port electrical quantities of each updated submodule according to the topological connection relationship of each submodule to construct the system-level electromagnetic transient waveform of the power electronic system includes: Based on the updated internal state variables of each submodule, determine the port electrical quantities of each submodule; The system constraint equations are constructed based on the topological connection relationships of each sub-module; The port electrical quantities of each submodule and the current external excitation parameters of the power electronic system are substituted into the system constraint equations for solution to obtain the system-level electromagnetic parameters of the power electronic system, thereby constructing the system-level electromagnetic transient waveform.

9. An electromagnetic transient simulation device for a power electronic system, characterized in that, The device includes: The processing unit is used to construct a set of discrete operating conditions corresponding to each sub-module of the power electronic system. The set of discrete operating conditions includes external input parameters and internal state variables corresponding to each discrete operating point. The discrete operating point is obtained by discretizing the external input parameters and internal state variables of the standard operating point. The simulation unit is used to generate response data of each of the sub-modules under different discrete operating conditions obtained by the processing unit using numerical simulation. The response data includes transient sequence data in response to switch switching behavior and steady-state incremental data in response to switch holding behavior. The first construction unit is used to construct a simulation behavior database for each of the sub-modules based on the transient sequence data and the steady-state incremental data obtained by the simulation unit. The query unit is used to query the target transient sequence data or target steady-state incremental data corresponding to each sub-module in the simulation behavior database obtained by the first construction unit, based on the real-time operating point and real-time switching behavior of each sub-module during the electromagnetic transient simulation of the power electronic system. The second construction unit is used to update the internal state variables of each sub-module using the target transient sequence data or the target steady-state incremental data obtained by the query unit, and to merge the port electrical quantities of each updated sub-module according to the topological connection relationship of each sub-module, so as to construct the system-level electromagnetic transient waveform of the power electronic system.

10. A storage medium, characterized in that, The storage medium includes a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the electromagnetic transient simulation method for a power electronic system as described in any one of claims 1 to 8.

11. A processor, characterized in that, The processor is used to run a program, wherein the program executes the electromagnetic transient simulation method for a power electronic system as described in any one of claims 1 to 8.

Images