Power system power flow calculation method and device based on trust domain model switching

Abstract

The application relates to a power system flow calculation method and device based on trust domain model switching, wherein the method comprises the following steps: a unified mathematical model of power system flow is established, a complex alternating current flow problem is converted into a standard nonlinear optimization problem; a numerical solver based on a trust region framework is constructed to ensure that an iteration process is carried out in a reliable region; dynamic model selection and parameter adaptive adjustment are realized, a system real-time state optimization calculation strategy is adopted, and therefore the convergence and robustness of the flow calculation are significantly improved. Thus, the problems in the prior art, such as single model, insufficient adaptability, limited convergence reliability, lack of numerical stability, lack of perfect adaptive mechanism and the like when a sick state power system is processed, are solved, and the reliable application of the flow calculation method in the sick state power system is seriously restricted.

Description

Technical Field

[0001] This application relates to the field of power system technology, and in particular to a power system power flow calculation method and apparatus based on trust domain model switching. Background Technology

[0002] Power flow calculation is a fundamental algorithm in power system analysis, used to determine the voltage magnitude, phase angle, and power distribution of the power grid during steady-state operation. The reliability and efficiency of power flow calculation directly affect the accuracy of power system planning, operation, and control. Currently, existing technologies can address the convergence difficulties in power flow calculations in microgrid systems caused by singular Jacobian matrices or the absence of balancing nodes, based on algorithms that combine the Levenberg-Marquardt method with trust region techniques. In addition, existing technologies can also maintain frequency and voltage stability through power flow calculation methods for islanded AC / DC hybrid microgrids based on improved adaptive droop control, which also employ trust region-based LMTR solution algorithms.

[0003] In recent years, with the development of artificial intelligence technology, existing techniques can map power flow information into high-dimensional, topologically decoupled node representations in a high-dimensional space by constructing a large pre-trained power system model that includes feature upscaling, encoders, decoders, and feature downscaling. This model can then be applied to downstream tasks such as future power flow generation and analysis. While this method provides a new approach to power flow analysis, it relies on extensive historical data training, and its convergence reliability for ill-conditioned systems remains uncertain. Trust domain methods, as an important branch of numerical optimization, have shown promising application prospects in power system power flow calculation. However, existing trust domain methods have limitations in model selection and adaptive mechanisms. Specifically, existing power flow calculation methods suffer from the following main drawbacks when dealing with ill-conditioned power systems: 1. Insufficient model adaptability: Existing trust domain methods usually use a single quadratic model or Gauss-Newton model, which cannot dynamically adjust the model structure according to the system state and lacks flexibility when dealing with systems with different degrees of ill-conditioning.

[0004] 2. Limited convergence reliability: The traditional Newton-Raphson method has poor convergence performance when dealing with ill-conditioned systems with high R / X ratios and heavy loads, especially when the initial values ​​are not ideal, the convergence rate drops significantly.

[0005] 3. Lack of numerical stability: Under strongly nonlinear system conditions, the Jacobian matrix of existing methods is prone to ill-conditioning, leading to unstable calculation processes or even complete failure. 4. Imperfect adaptive mechanism: Existing trust domain methods lack an intelligent model switching mechanism and cannot automatically select the optimal calculation model based on the real-time status of the system, which limits the application effect of the method in complex power systems.

[0006] In summary, existing technologies for handling ill-conditioned power systems suffer from limitations in model simplification, insufficient adaptability, limited convergence reliability, lack of numerical stability, and absence of a robust adaptive mechanism. These shortcomings severely restrict the reliable application of power flow calculation methods in ill-conditioned power systems and urgently require solutions. Summary of the Invention

[0007] This application provides a power flow calculation method and apparatus based on trust domain model switching to solve the problems of existing technologies in dealing with ill-conditioned power systems, such as single model, insufficient adaptability, limited convergence reliability, lack of numerical stability, and lack of a sound adaptive mechanism, which seriously restrict the reliable application of power flow calculation methods in ill-conditioned power systems.

[0008] The first aspect of this application provides a power system power flow calculation method based on trust domain model switching, comprising the following steps: constructing nodal power balance equations of the target power system based on Kirchhoff's laws, and decomposing the nodal power balance equations into real-valued power equations; constructing a unified mathematical model of power system power flow based on the real-valued power equations and preset nodal constraints; establishing a corresponding numerical solver based on a pre-constructed trust domain optimization framework, and solving the constraint optimization problem of the unified mathematical model of power system power flow using the numerical solver to obtain corresponding solution results; generating predicted power flow decreases corresponding to the quadratic model and the Gauss-Newton model based on preset quadratic models, Gauss-Newton models, and the solution results; dynamically selecting the quadratic model or the Gauss-Newton model as the corresponding target prediction model based on preset numerical stability conditions, obtaining the predicted decrease generated by the target prediction model and the actual decrease of the power system power flow, calculating the decrease ratio between the actual decrease and the predicted decrease, and dynamically adjusting the trust domain radius using the decrease ratio to calculate the power system power flow using the adjusted trust domain radius.

[0009] Optionally, in one embodiment of this application, the step of constructing the node power balance equations of the target power system based on Kirchhoff's laws, decomposing the node power balance equations into real-valued power equations, and constructing a unified mathematical model of power flow based on the real-valued power equations and preset node constraints includes: determining the node admittance matrix, node voltage vector, and node power injection vector corresponding to the target power system based on Kirchhoff's laws, and constructing the node power balance equations based on the node admittance matrix, the node voltage vector, and the node power injection vector; performing real-domain decomposition on the node power balance equations to obtain the real-valued power equations, and determining multiple node types corresponding to the target power system to construct node constraints for each of the multiple node types, wherein the multiple node types include PQ node type, PV node type, and slack node type; and constructing a unified mathematical model of power flow based on the real-valued power equations and the node constraints.

[0010] Optionally, in one embodiment of this application, the step of solving the constrained optimization problem of the unified mathematical model of power system flow using the numerical solver to obtain the corresponding solution result, and generating the predicted decrease in power system flow corresponding to the quadratic model and the Gauss-Newton model based on the preset quadratic model, the Gauss-Newton model, and the solution result, respectively, includes: iteratively performing a constrained optimization problem solving operation on the unified mathematical model of power system flow using the numerical solver, wherein, in each iteration, the solution result of the current iteration is calculated using the Jacobian matrix of the current iteration point and the radius of the trust region of the current iteration; the solution result is input into the quadratic model and the Gauss-Newton model respectively, so as to perform a preset second-order approximation estimation operation on the solution result using the quadratic model to generate the predicted decrease in power system flow corresponding to the quadratic model, and performing a preset residual norm comparison operation on the solution result using the Gauss-Newton model to obtain the predicted decrease in power system flow corresponding to the Gauss-Newton model.

[0011] Optionally, in one embodiment of this application, the step of dynamically selecting the quadratic model or the Gauss-Newton model as the corresponding target prediction model based on a preset numerical stability condition, obtaining the predicted decrease amount generated by the target prediction model and the actual decrease amount of the power system flow, calculating the decrease ratio between the actual decrease amount and the predicted decrease amount, and dynamically adjusting the trust region radius using the decrease ratio to calculate the power system flow using the adjusted trust region radius includes: calculating the reciprocal of the condition number of the trust region radius of the current iteration, and determining whether the predicted decrease amount is greater than zero and whether the reciprocal of the condition number is greater than a preset numerical stability threshold; if the predicted decrease amount is greater than zero... If the reciprocal of the condition number is greater than the numerical stability threshold, then the quadratic model is selected as the target prediction model for the current iteration; otherwise, the Gauss-Newton model is selected as the target prediction model for the current iteration. The ratio between the actual decrease and the predicted decrease corresponding to the target prediction model in the current iteration is calculated, and this ratio is compared with multiple preset ratio thresholds to obtain corresponding comparison results. The trust region radius of the current iteration is adjusted based on the comparison results to obtain the trust region radius for the next iteration. It is then determined whether the next iteration meets the preset iteration end requirements. If the next iteration meets the preset iteration end requirements, the iteration operation is stopped to obtain the power system power flow.

[0012] A second aspect of this application provides a power system power flow calculation device based on trust domain model switching, comprising: a modeling module, used to construct the nodal power balance equations of the target power system based on Kirchhoff's laws, and decompose the nodal power balance equations into real-valued power equations, and construct a unified mathematical model of power system power flow based on the real-valued power equations and preset nodal constraints; and an optimization module, used to establish a corresponding numerical solver based on a pre-constructed trust domain optimization framework, and solve the constraint optimization problem of the unified mathematical model of power system power flow through the numerical solver to obtain the corresponding solution results. Based on the preset quadratic model, Gauss-Newton model, and the solution results, the predicted decrease in power system flow corresponding to the quadratic model and the Gauss-Newton model are generated respectively. The calculation module is used to dynamically select the quadratic model or the Gauss-Newton model as the corresponding target prediction model based on the preset numerical stability conditions, and obtain the predicted decrease generated by the target prediction model and the actual decrease in power system flow. The module also calculates the decrease ratio between the actual decrease and the predicted decrease, and dynamically adjusts the trust domain radius through the decrease ratio to calculate the power system flow using the adjusted trust domain radius.

[0013] Optionally, in one embodiment of this application, the modeling module includes: a first construction unit, configured to determine the node admittance matrix, node voltage vector, and node power injection vector corresponding to the target power system based on the Kirchhoff laws, and construct the node power balance equation according to the node admittance matrix, the node voltage vector, and the node power injection vector; a decomposition unit, configured to perform real-domain decomposition on the node power balance equation to obtain the power equation in real-number form, and determine multiple node types corresponding to the target power system to construct node constraints corresponding to each of the multiple node types, wherein the multiple node types include PQ node type, PV node type, and slack node type; and a second construction unit, configured to construct a unified mathematical model of power system flow based on the power equation in real-number form and the node constraints.

[0014] Optionally, in one embodiment of this application, the optimization module includes: a solution unit, configured to perform constrained optimization problem solving operations on the unified mathematical model of the power system power flow iteratively through the numerical solver, wherein, in each iteration, the solution result of the current iteration is calculated using the Jacobian matrix of the current iteration point and the confidence region radius of the current iteration; and a comparison unit, configured to input the solution result into the quadratic model and the Gauss-Newton model respectively, to perform a preset second-order approximation estimation operation on the solution result through the quadratic model to generate the predicted decrease in power system power flow corresponding to the quadratic model, and to perform a preset residual norm comparison operation on the solution result using the Gauss-Newton model to obtain the predicted decrease in power system power flow corresponding to the Gauss-Newton model.

[0015] Optionally, in one embodiment of this application, the calculation module includes: a first judgment unit, configured to calculate the reciprocal of the condition number of the trust domain radius of the current iteration, and determine whether the predicted decrease is greater than zero and whether the reciprocal of the condition number is greater than a preset numerical stability threshold; a selection unit, configured to select the quadratic model as the target prediction model for the current iteration if the predicted decrease is greater than zero and the reciprocal of the condition number is greater than the numerical stability threshold, otherwise select the Gauss-Newton model as the target prediction model for the current iteration; an adjustment unit, configured to calculate the ratio between the actual decrease and the predicted decrease of the target prediction model corresponding to the current iteration, and compare the ratio with a plurality of preset ratio thresholds to obtain a corresponding comparison result, and adjust the trust domain radius of the current iteration according to the comparison result to obtain the trust domain radius of the next iteration; and a second judgment unit, configured to determine whether the next iteration meets a preset iteration end requirement, wherein, if the next iteration meets the preset iteration end requirement, the iteration operation is stopped to obtain the power system power flow.

[0016] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the power flow calculation method for trust domain model switching as described in the above embodiments.

[0017] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power flow calculation method for a power system based on trust domain model switching as described above.

[0018] A fifth aspect of this application provides a computer program product, including a computer program that is executed to implement the above-described power flow calculation method based on trust domain model switching.

[0019] Therefore, the embodiments of this application have the following beneficial effects: The embodiments of this application can construct the nodal power balance equations of the target power system based on Kirchhoff's laws, decompose the nodal power balance equations into real-valued power equations, and construct a unified mathematical model of power flow based on the real-valued power equations and preset nodal constraints. Based on a pre-constructed trust domain optimization framework, a corresponding numerical solver is established, and the constrained optimization problem of the unified mathematical model of power flow is solved by the numerical solver to obtain the corresponding solution results. Based on preset quadratic models, Gauss-Newton models, and solution results, the predicted decrease in power flow corresponding to the quadratic model and the Gauss-Newton model are generated respectively. Based on preset numerical stability conditions, the quadratic model or the Gauss-Newton model is dynamically selected as the corresponding target prediction model, and the predicted decrease and the actual decrease in power flow generated by the target prediction model are obtained. The decrease ratio between the actual decrease and the predicted decrease is calculated, and the trust domain radius is dynamically adjusted by the decrease ratio to calculate the power flow using the adjusted trust domain radius. This application significantly improves the convergence and robustness of power flow calculation by employing dynamic model switching and adaptive trust domain adjustment operations, thereby enhancing the accuracy and reliability of power system analysis. This addresses the problems of existing technologies in handling ill-conditioned power systems, such as the reliance on a single model, insufficient adaptability, limited convergence reliability, lack of numerical stability, and the absence of a robust adaptive mechanism, which severely restrict the reliable application of power flow calculation methods in ill-conditioned power systems.

[0020] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0021] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart illustrating a power flow calculation method for a power system based on trust domain model switching, according to an embodiment of this application. Figure 2 A schematic diagram illustrating the comparison of computation time and convergence analysis of a loading mode provided for an embodiment of this application; Figure 3 A schematic diagram illustrating the comparison of iteration counts and convergence analysis of a loading mode according to an embodiment of this application; Figure 4 A schematic diagram illustrating the comparison of computation time and convergence analysis of a melting mode provided for one embodiment of this application; Figure 5 A schematic diagram illustrating the comparison of iteration counts and convergence analysis of a melting mode, provided for one embodiment of this application; Figure 6 A schematic diagram illustrating the computation time comparison and convergence analysis of a refining mode provided for an embodiment of this application; Figure 7 A schematic diagram illustrating the comparison of iteration counts and convergence analysis of a refining mode provided for an embodiment of this application; Figure 8 A schematic diagram illustrating the comparison and analysis of iteration counts provided in one embodiment of this application; Figure 9 A schematic diagram of convergence rate improvement analysis provided for one embodiment of this application; Figure 10 A schematic diagram of voltage distribution results for power flow calculation under three electric arc furnace operating modes is provided as an embodiment of this application; Figure 11 This is an example diagram of a power system power flow calculation device based on trust domain model switching according to an embodiment of this application; Figure 12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0022] Among them, 10-Power system power flow calculation device based on trust domain model switching; 100-Modeling module, 200-Optimization module, 300-Calculation module; 1201-Memory, 1202-Processor, 1203-Communication interface. Detailed Implementation

[0023] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0024] The following describes, with reference to the accompanying drawings, a power system power flow calculation method and apparatus based on trust domain model switching according to embodiments of this application. To address the problems mentioned in the background, this application provides a power system flow calculation method based on trust domain model switching. In this method, based on Kirchhoff's laws, the nodal power balance equations of the target power system are constructed and decomposed into real-valued power equations. A unified mathematical model of power system flow is constructed based on the real-valued power equations and preset nodal constraints. Based on the pre-constructed trust domain optimization framework, a corresponding numerical solver is established, and the constraint optimization problem of the unified mathematical model of power system flow is solved to obtain the corresponding solution results. Based on preset quadratic models, Gauss-Newton models, and the solution results, the predicted decrease in power system flow corresponding to the quadratic model and the Gauss-Newton model are generated, respectively. Based on preset numerical stability conditions, either the quadratic model or the Gauss-Newton model is dynamically selected as the corresponding target prediction model. The predicted decrease generated by the target prediction model and the actual decrease in power system flow are obtained, and the ratio of the actual decrease to the predicted decrease is calculated. The trust domain radius is dynamically adjusted based on the decrease ratio to calculate the power system flow using the adjusted trust domain radius. This application significantly improves the convergence and robustness of power flow calculation by employing dynamic model switching and adaptive trust domain adjustment operations, thereby enhancing the accuracy and reliability of power system analysis. This addresses the problems of existing technologies in handling ill-conditioned power systems, such as the reliance on a single model, insufficient adaptability, limited convergence reliability, lack of numerical stability, and the absence of a robust adaptive mechanism, which severely restrict the reliable application of power flow calculation methods in ill-conditioned power systems.

[0025] Specifically, Figure 1 This is a flowchart illustrating a power flow calculation method for a power system based on trust domain model switching, provided as an embodiment of this application.

[0026] like Figure 1 As shown, the power flow calculation method for power systems based on trust domain model switching includes the following steps: In step S101, based on Kirchhoff's laws, the nodal power balance equations of the target power system are constructed, and the nodal power balance equations are decomposed into power equations in real form. Based on the power equations in real form and the preset nodal constraints, a unified mathematical model of power flow of the power system is constructed.

[0027] The embodiments of this application first utilize Kirchhoff's laws to construct the nodal power balance equations of the power system and decompose them in the real domain to obtain the corresponding real-number power equations; secondly, the embodiments of this application can establish corresponding constraints according to the node type to construct a unified mathematical model of power flow based on the real-number power equations and nodal constraints.

[0028] Optionally, in one embodiment of this application, based on Kirchhoff's laws, the nodal power balance equations of the target power system are constructed, and the nodal power balance equations are decomposed into power equations in real form. A unified mathematical model of power flow is constructed based on the real-form power equations and preset nodal constraints. This includes: determining the nodal admittance matrix, nodal voltage vector, and nodal power injection vector corresponding to the target power system based on Kirchhoff's laws, and constructing nodal power balance equations based on these equations; performing real-domain decomposition on the nodal power balance equations to obtain power equations in real form, and determining multiple nodal types corresponding to the target power system to construct nodal constraints for each nodal type, where the multiple nodal types include PQ nodal type, PV nodal type, and slack nodal type; and constructing a unified mathematical model of power flow based on the real-form power equations and nodal constraints.

[0029] Specifically, the process of establishing a unified mathematical model for power system power flow in the embodiments of this application is as follows: 1. Establishing network equations: The embodiments of this application can construct nodal power balance equations based on Kirchhoff's laws:

[0030] Where Y is the node admittance matrix; V is the node voltage vector; and S is the node power injection vector.

[0031] 2. Real number field decomposition: The nodal power balance equations are decomposed in the real domain to obtain the power equations in real form, as shown below:

[0032] 3. Constraint Integration: This application embodiment can determine multiple node types corresponding to a power system, so as to construct node constraint conditions corresponding to different node types, as described below: (1) PQ node type: fixed active power and reactive power ; (2) PV node type: fixed active power and voltage amplitude ; (3) Balance node type; fixed voltage amplitude and phase angle.

[0033] Therefore, the embodiments of this application can construct a unified mathematical model corresponding to power flow in a power system based on real-number power equations and node constraints, thereby transforming the complex AC power flow problem into a standard nonlinear optimization problem.

[0034] In step S102, based on the pre-constructed trust domain optimization framework, a corresponding numerical solver is established, and the constrained optimization problem of the unified mathematical model of power flow is solved by the numerical solver to obtain the corresponding solution results. Based on the preset quadratic model, Gauss-Newton model and solution results, the predicted decrease in power flow corresponding to the quadratic model and Gauss-Newton model are generated respectively.

[0035] Furthermore, embodiments of this application also require the construction of a numerical solver based on a trust domain optimization framework to solve the constrained optimization problem of the unified mathematical model of power flow in the power system, so as to obtain the corresponding solution results, thereby ensuring that the iterative process is carried out in a reliable region; in addition, embodiments of this application can generate the predicted decrease in power flow corresponding to the quadratic model and the Gauss-Newton model, respectively, based on the preset quadratic model, Gauss-Newton model and solution results.

[0036] Therefore, the embodiments of this application can design a trust domain algorithm based on a dual model switching mechanism, thereby ensuring reliable convergence under various initial value conditions, significantly improving the computation success rate, and enhancing the convergence reliability of power flow calculation in ill-conditioned systems.

[0037] Optionally, in one embodiment of this application, a constrained optimization problem of the unified mathematical model of power system flow is solved using a numerical solver to obtain the corresponding solution results. Based on a preset quadratic model, a Gauss-Newton model, and the solution results, the predicted decrease in power system flow corresponding to the quadratic model and the Gauss-Newton model are generated respectively. This includes: iteratively solving the constrained optimization problem of the unified mathematical model of power system flow using a numerical solver, wherein in each iteration, the solution result of the current iteration is calculated using the Jacobian matrix of the current iteration point and the radius of the trust region of the current iteration; the solution results are input into the quadratic model and the Gauss-Newton model respectively, so that a preset second-order approximation estimation operation is performed on the solution results using the quadratic model to generate the predicted decrease in power system flow corresponding to the quadratic model, and a preset residual norm comparison operation is performed on the solution results using the Gauss-Newton model to obtain the predicted decrease in power system flow corresponding to the Gauss-Newton model.

[0038] In practical implementation, embodiments of this application can construct a trust domain optimization framework to optimize trust domains in each iteration. k Solving the constrained optimization problem in the following equation:

[0039] in, This represents the Jacobian matrix at the current iteration point; This indicates the dynamically adjusted radius of the trust domain.

[0040] Furthermore, embodiments of this application can design two complementary prediction models, namely a quadratic model and a Gauss-Newton model, to construct a dual-model prediction mechanism. The quadratic model can approximate the decrease in the objective function of the power system flow during each iteration (i.e., the solution result corresponding to the quadratic model), as shown in the following equation:

[0041] Secondly, the Gauss-Newton model can conservatively estimate the decrease in the objective function of the power flow in each iteration based on the residual norm (i.e., the solution result corresponding to the Gauss-Newton model), as shown in the following equation:

[0042] Therefore, the embodiments of this application construct a dual-model prediction mechanism, thereby providing reliable data guidance and theoretical support for the subsequent calculation of power flow in the power system.

[0043] In step S103, based on the preset numerical stability conditions, a quadratic model or a Gauss-Newton model is dynamically selected as the corresponding target prediction model, and the predicted decrease and the actual decrease of the power flow generated by the target prediction model are obtained. The ratio of the decrease between the actual decrease and the predicted decrease is calculated, and the radius of the trust domain is dynamically adjusted by the ratio of the decrease, so as to calculate the power flow using the adjusted radius of the trust domain.

[0044] Subsequently, embodiments of this application can dynamically select a quadratic model or a Gauss-Newton model as the corresponding target prediction model based on the value stability condition, obtain the corresponding predicted decrease and the actual decrease of the power system flow, and calculate the decrease ratio between the predicted decrease and the actual decrease, thereby dynamically adjusting the trust domain radius and using the adjusted trust domain radius to calculate the power system flow.

[0045] Therefore, the embodiments of this application provide a trust domain framework that can dynamically switch the computational model according to the system state, thereby maintaining optimal computational performance in systems with different degrees of ill-conditioning, thus solving the problems of single model and insufficient adaptability of existing trust domain methods. Secondly, the embodiments of this application can also effectively avoid the difficulty of power flow convergence caused by ill-conditioning of the Jacobian matrix through intelligent model selection and adaptive adjustment of the trust domain radius, ensuring the stability of the computation process under strongly nonlinear conditions. In addition, the embodiments of this application can also establish model evaluation and switching criteria based on the real-time state of the system, achieving optimal matching between the computational model and system characteristics, improving applicability in complex power systems, providing a general, reliable and efficient solution for power system power flow calculation, and improving the accuracy and reliability of power system analysis.

[0046] Optionally, in one embodiment of this application, based on a preset numerical stability condition, a quadratic model or a Gauss-Newton model is dynamically selected as the corresponding target prediction model, and the predicted decrease and the actual decrease of the power system flow generated by the target prediction model are obtained. The ratio of the actual decrease to the predicted decrease is calculated, and the trust region radius is dynamically adjusted using this ratio to calculate the power system flow. This includes: calculating the reciprocal of the condition number of the trust region radius in the current iteration, and determining whether the predicted decrease is greater than zero and whether the reciprocal of the condition number is greater than a preset numerical stability threshold; if the predicted decrease is greater than zero and the condition number is greater than a preset numerical stability threshold... If the reciprocal of the number of items is greater than the numerical stability threshold, a quadratic model is selected as the target prediction model for the current iteration; otherwise, a Gauss-Newton model is selected. The ratio between the actual decrease and the predicted decrease corresponding to the target prediction model for the current iteration is calculated, and the ratio is compared with several preset ratio thresholds to obtain the corresponding comparison results. The confidence region radius of the current iteration is adjusted according to the comparison results to obtain the confidence region radius for the next iteration. It is then determined whether the next iteration meets the preset iteration end requirements. If the next iteration meets the preset iteration end requirements, the iteration operation is stopped to obtain the power system power flow.

[0047] As one possible approach, embodiments of this application can be based on an intelligent model switching strategy to dynamically select the prediction model according to numerical stability conditions, as shown in the following equation:

[0048] in, κ (·) denotes the reciprocal of the condition number of a matrix; This is the preset numerical stability threshold.

[0049] Furthermore, embodiments of this application can perform adaptive trust domain management, that is, based on the ratio of the actual decrease to the predicted decrease. The trust domain radius is dynamically adjusted as shown in the following formula:

[0050] The execution logic of the power system power flow calculation method based on trust domain model switching according to the embodiments of this application is described below. In the embodiments of this application, a power system power flow calculation operation based on trust domain model switching is mainly performed in three stages: initialization, iterative solution, and convergence judgment. The details are as follows: 1. Initialization phase: (1) Set the initial voltage vector V 0, typically employs a flat start or hot start strategy; (2) Calculate the initial Jacobian matrix J 0 and residual vector F 0; (3) Initialize the trust domain radius .

[0051] 2. Iterative solution stage: (1) Calculate the Jacobian matrix and residual vector at the current iteration point; (2) Solve the trust domain subproblem to obtain the trial step. ; (3) Calculate the predicted decrease for both models respectively; (4) Select the optimal prediction model based on the numerical stability criterion; (5) Calculate the ratio of the actual decrease to the predicted decrease. ; (6) According to The value determines whether to accept the trial step; (7) Dynamically adjust the radius of the trust domain.

[0052] 3. Convergence Judgment Phase: (1) Main convergence criteria: ; (2) Auxiliary convergence criterion: The number of iterations reaches the preset upper limit. N max .

[0053] Therefore, compared with the prior art, the embodiments of this application have the following advantages: 1. Improved convergence reliability: By using a hybrid prediction strategy and intelligent model switching, the convergence rate of PV node phase angle initial value scanning is increased from 80%-88% in traditional methods to 100% when dealing with ill-conditioned systems; 2. Enhanced numerical stability: The dual-model design ensures stable prediction results under any system state, effectively avoiding numerical instability caused by ill-conditioned Jacobian matrix; 3. Improved robustness to initial values: The adaptive trust region mechanism significantly reduces the dependence on initial values, ensuring that effective solutions can be obtained across a wide range of initial values; 4. Optimized computational efficiency: The intelligent model selection mechanism ensures convergence reliability while preserving quadratic convergence characteristics to the maximum extent, achieving the best balance between convergence speed and stability. In summary, the technical solution of this application can adopt a layered design architecture including a basic problem modeling layer, a trust domain optimization layer, and an intelligent decision-making layer. Specifically, in the basic problem modeling layer, a unified mathematical model of power system flow can be established, transforming the complex AC power flow problem into a standard nonlinear optimization problem. In the trust domain optimization layer, a numerical solver based on the trust domain framework can be constructed to ensure that the iteration process is carried out within a reliable region. In the intelligent decision-making layer, dynamic model selection and adaptive parameter adjustment can be implemented to optimize the calculation strategy according to the real-time state of the system. Thus, through dynamic model switching and adaptive trust domain adjustment, the convergence and robustness of power flow calculation are significantly improved, solving the technical problems of insufficient convergence reliability, sensitivity to initial values, poor numerical stability, and limited model adaptability in existing technologies when dealing with ill-conditioned power systems.

[0054] Furthermore, the technical solutions described in this application have been verified by the IEEE standard testing system, demonstrating excellent performance in challenging scenarios such as handling fluctuating industrial load grid connection, and providing reliable technical support for power system planning, operation and control.

[0055] Specifically, the test results of the power system power flow calculation method based on trust domain model switching of this application are described below through a specific embodiment and in conjunction with the accompanying drawings.

[0056] This application verifies performance through a case study of a constructed IEEE 14-bus H-DRI-EAF system. The application provides a test case in which the IEEE 14-bus H-DRI-EAF system considers multiple energy demands, including hydrogen production load, hydrogen heating load, and electric arc furnace load, with the IEEE 14 configuration as follows: 1. Integration of Multiple Industrial Load Types: The system fully integrates three key load types in the hydrogen-based direct reduced iron (H-DRI) production process: (1) Hydrogen production load: Electrolysis hydrogen production units with a total capacity of 36MW are configured at nodes 3, 4 and 5, with a power factor of 0.95; (2) Hydrogen heating load: A hydrogen heating system with a total capacity of 7MW is configured at nodes 9 and 10, with a power factor of 0.90. (3) Electric arc furnace load: Variable capacity EAF equipment is configured at nodes 7 and 8 to accurately simulate the steel smelting process.

[0057] 2. Refined Modeling of EAF Operation Status: Fully considering the actual working characteristics of the electric arc furnace, the system completely simulates three typical operation stages: (1) Loading stage (0-10 minutes): load 3.0 MW, power factor 0.75, corresponding to the raw material loading process; (2) Melting stage (10-40 minutes): load 12.0 MW, power factor 0.75, corresponding to high-intensity melting operation; (3) Refining stage (40-60 minutes): Load 8.0 MW, power factor 0.75, corresponding steel quality adjustment.

[0058] 3. Coordinated Configuration of Power Generation System: A total of 315 MW of generator units will be configured at nodes 1, 2, 3, 6, and 8 to ensure system power balance, taking into account the gradeability and regulation capacity of each generator unit. Therefore, in the embodiments of this application, the total system load changes dynamically within the range of 315-324 MW, which fully reflects the time-varying characteristics of the actual power demand of steel enterprises and provides a real and reliable test platform for algorithm verification.

[0059] This application analyzes the test results of the power flow calculation method based on trust domain model switching applied to the IEEE 14-busH-DRI-EAF system. Specifically, it compares the calculation time and convergence analysis for the charging mode (i.e., the stage where the furnace lid is opened and raw materials such as scrap steel are added to the furnace). Figure 2 As shown, the comparison of iteration times and convergence analysis of the charging mode are as follows: Figure 3 As shown, the computation time comparison and convergence analysis of the melting mode (i.e., the stage where the electrode descends to ignite the arc and electrical energy is converted into thermal energy to melt the furnace charge) are as follows: Figure 4 As shown, the comparison of iteration times and convergence analysis of the smelting mode are as follows: Figure 5 As shown, the computation time comparison and convergence analysis of the refining mode (i.e., the stage after the furnace charge has completely melted, in which the composition and temperature of the molten steel are adjusted) are as follows: Figure 6 As shown, the comparison of iteration counts and convergence analysis of the refining mode are as follows: Figure 7 As shown, the comparison and analysis of iteration counts are as follows: Figure 8 As shown, the convergence rate improvement analysis is as follows: Figure 9 The voltage distribution results of power flow calculations under the three electric arc furnace operating modes are shown in the figure. Figure 10 As shown.

[0060] The power flow calculation method for power systems based on dual-model trust domain switching provided in this application has the following significant advantages over existing technologies: 1. Innovative Dual-Model Dynamic Switching Mechanism: This application pioneers a dual-model trust domain power flow calculation method based on system state. By intelligently switching between the quadratic model and the Gauss-Newton model, the optimal balance between computational accuracy and numerical stability is achieved. The quadratic model is used in regions with good system state to obtain more accurate curvature information, while the Gauss-Newton model is automatically switched in regions with ill-conditioned Jacobian matrix or strong nonlinearity to ensure the numerical stability of the algorithm throughout the entire process. 2. Significantly improved convergence reliability: Table 1 shows the comparison of convergence performance under different electric arc furnace operating modes. This application achieved a 100% convergence rate in the initial value scan of the PV node phase angle, which is a significant improvement of 12%-20% compared with the 80%-88% convergence rate of the traditional method. In particular, under the most challenging smelting operating mode, the convergence rate improvement reached 20.0%, thus effectively solving the convergence difficulty problem of the traditional method when dealing with fluctuating industrial loads.

[0061] Table 1

[0062] 3. Multi-mode adaptive capability: This application demonstrates superior performance in different electric arc furnace operating modes, especially in the smelting mode. This characteristic is highly consistent with the physical characteristics of the electric arc furnace load. The severe harmonic distortion and power fluctuations generated during smelting operation pose a great challenge to traditional methods, while this application effectively addresses these challenges through a dual-model trust domain mechanism. 4. Significantly improved robustness to initial values: This application can maintain reliable convergence over a wide range of initial values, effectively overcoming the shortcomings of traditional methods that are highly sensitive to the initial voltage estimate; under high load operating conditions, the trust domain mechanism can effectively handle the sensitivity to the increase of the initial voltage value, ensuring that effective power flow solutions can be obtained under various harsh initial conditions.

[0063] 5. Solving industry technical challenges: This application provides a reliable solution to the special technical challenges brought about by grid connection of hydrogen-based direct reduction iron-electric arc furnace system in green steel production, and provides key technical support for the decarbonization process of high-energy-consuming industries.

[0064] 6. Broad prospects for promotion and application: This application is not only applicable to electric arc furnace load scenarios, but can also be widely used in grid connection analysis of other fluctuating industrial loads. It has important engineering application value and broad market prospects. The promotion and application of this technology will significantly improve the power system's ability to accept fluctuating industrial loads and promote energy transformation and industrial upgrading. In summary, this application, through an innovative dual-model trust domain switching mechanism, significantly improves the convergence reliability and numerical stability of power flow calculation while maintaining the computational efficiency of traditional methods, providing a reliable technical guarantee for the safe and stable operation of power systems.

[0065] According to the power system power flow calculation method based on trust domain model switching proposed in this application, the nodal power balance equations of the target power system are constructed based on Kirchhoff's laws, and the nodal power balance equations are decomposed into real-valued power equations. A unified mathematical model of power system power flow is constructed based on the real-valued power equations and preset nodal constraints. Based on the pre-constructed trust domain optimization framework, a corresponding numerical solver is established, and the constraint optimization problem of the unified mathematical model of power system power flow is solved by the numerical solver to obtain the corresponding solution results. Based on preset quadratic models, Gauss-Newton models, and solution results, the predicted decrease in power system power flow corresponding to the quadratic model and the Gauss-Newton model are generated respectively. Based on preset numerical stability conditions, the quadratic model or the Gauss-Newton model is dynamically selected as the corresponding target prediction model, and the predicted decrease and the actual decrease in power system power flow generated by the target prediction model are obtained. The decrease ratio between the actual decrease and the predicted decrease is calculated, and the trust domain radius is dynamically adjusted by the decrease ratio to calculate the power system power flow using the adjusted trust domain radius. This application significantly improves the convergence and robustness of power flow calculation through dynamic model switching and adaptive trust domain adjustment operations, thereby enhancing the accuracy and reliability of power system analysis.

[0066] Secondly, with reference to the accompanying drawings, a power flow calculation device based on trust domain model switching for a power system according to an embodiment of this application is described.

[0067] Figure 11 This is a block diagram of a power system power flow calculation device based on trust domain model switching according to an embodiment of this application.

[0068] like Figure 11 As shown, the power flow calculation device 10 based on trust domain model switching includes: a modeling module 100, an optimization module 200, and a calculation module 300.

[0069] Among them, the modeling module 100 is used to construct the nodal power balance equations of the target power system based on Kirchhoff's laws, decompose the nodal power balance equations into real power equations, and construct a unified mathematical model of power flow of the power system based on the real power equations and preset nodal constraints.

[0070] The optimization module 200 is used to establish a corresponding numerical solver based on a pre-built trust domain optimization framework, and solve the constrained optimization problem of the unified mathematical model of power flow in the power system through the numerical solver to obtain the corresponding solution results. Based on the preset quadratic model, Gauss-Newton model and solution results, it generates the predicted decrease in power flow in the power system corresponding to the quadratic model and Gauss-Newton model, respectively.

[0071] The calculation module 300 is used to dynamically select a quadratic model or a Gauss-Newton model as the corresponding target prediction model based on preset numerical stability conditions, and to obtain the predicted decrease and the actual decrease of the power flow generated by the target prediction model. It also calculates the ratio of the decrease between the actual decrease and the predicted decrease, and dynamically adjusts the trust domain radius based on the decrease ratio, so as to calculate the power flow using the adjusted trust domain radius.

[0072] Optionally, in one embodiment of this application, the modeling module 100 includes: a first building unit, a decomposition unit, and a second building unit.

[0073] The first building unit is used to determine the node admittance matrix, node voltage vector and node power injection vector of the target power system based on Kirchhoff's laws, and to construct the node power balance equation based on the node admittance matrix, node voltage vector and node power injection vector.

[0074] The decomposition unit is used to perform real-domain decomposition of the nodal power balance equations to obtain power equations in real form, and to determine the various nodal types corresponding to the target power system in order to construct the nodal constraints for each of the various nodal types, including PQ nodal type, PV nodal type and slack nodal type.

[0075] The second building unit is used to construct a unified mathematical model of power system flow based on real-form power equations and nodal constraints.

[0076] Optionally, in one embodiment of this application, the optimization module 200 includes a solution unit and a comparison unit.

[0077] The solution unit is used to perform constrained optimization problem solving operations on the unified mathematical model of power flow in the power system through numerical solver. In each iteration, the solution result of the current iteration is calculated using the Jacobian matrix of the current iteration point and the radius of the trust domain of the current iteration.

[0078] The comparison unit is used to input the solution results into the quadratic model and the Gauss-Newton model respectively. The quadratic model performs a preset second-order approximation estimation operation on the solution results to generate the predicted decrease in power flow corresponding to the quadratic model. The Gauss-Newton model performs a preset residual norm comparison operation on the solution results to obtain the predicted decrease in power flow corresponding to the Gauss-Newton model.

[0079] Optionally, in one embodiment of this application, the calculation module 300 includes: a first judgment unit, a selection unit, an adjustment unit, and a second judgment unit.

[0080] The first judgment unit is used to calculate the reciprocal of the condition number of the trust domain radius of the current iteration, and to determine whether the predicted decrease is greater than zero and whether the reciprocal of the condition number is greater than a preset numerical stability threshold.

[0081] The selection unit is used to select a quadratic model as the target prediction model for the current iteration if the predicted decrease is greater than zero and the reciprocal of the condition number is greater than the numerical stability threshold; otherwise, the Gauss-Newton model is selected as the target prediction model for the current iteration.

[0082] The adjustment unit is used to calculate the ratio between the actual decrease and the predicted decrease of the target prediction model in the current iteration, and compare the ratio with multiple preset ratio thresholds to obtain the corresponding comparison results. Based on the comparison results, the trust domain radius of the current iteration is adjusted to obtain the trust domain radius of the next iteration.

[0083] The second judgment unit is used to determine whether the next iteration meets the preset iteration end requirements. If the next iteration meets the preset iteration end requirements, the iteration operation is stopped to obtain the power system power flow.

[0084] It should be noted that the foregoing explanation of the power system power flow calculation method based on trust domain model switching also applies to the power system power flow calculation device based on trust domain model switching in this embodiment, and will not be repeated here.

[0085] The power flow calculation device based on trust domain model switching proposed in this application includes a modeling module 100, used to construct the nodal power balance equations of the target power system based on Kirchhoff's laws, decompose the nodal power balance equations into real-valued power equations, and construct a unified mathematical model of power flow based on the real-valued power equations and preset nodal constraints; and an optimization module 200, used to establish a corresponding numerical solver based on a pre-constructed trust domain optimization framework, and solve the constraint optimization problem of the unified mathematical model of power flow using the numerical solver to obtain the corresponding... The solution results, based on preset quadratic and Gauss-Newton models and the solution results, generate predicted power flow decline amounts corresponding to the quadratic and Gauss-Newton models, respectively. The calculation module 300 dynamically selects either the quadratic model or the Gauss-Newton model as the corresponding target prediction model based on preset numerical stability conditions. It obtains the predicted decline amount generated by the target prediction model and the actual decline amount of the power flow, calculates the decline ratio between the actual and predicted decline amounts, and dynamically adjusts the trust domain radius using the adjusted trust domain radius to calculate the power flow. This application significantly improves the convergence and robustness of power flow calculation through dynamic model switching and adaptive trust domain adjustment operations, thereby enhancing the accuracy and reliability of power system analysis.

[0086] Figure 12 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 1201, the processor 1202, and the computer program stored on the memory 1201 and executable on the processor 1202.

[0087] When the processor 1202 executes the program, it implements the power flow calculation method based on trust domain model switching provided in the above embodiments.

[0088] Furthermore, electronic devices also include: Communication interface 1203 is used for communication between memory 1201 and processor 1202.

[0089] The memory 1201 is used to store computer programs that can run on the processor 1202.

[0090] The memory 1201 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage.

[0091] If the memory 1201, processor 1202, and communication interface 1203 are implemented independently, then the communication interface 1203, memory 1201, and processor 1202 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be divided into address buses, data buses, control buses, etc. For ease of representation, Figure 12 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0092] Optionally, in a specific implementation, if the memory 1201, processor 1202, and communication interface 1203 are integrated on a single chip, then the memory 1201, processor 1202, and communication interface 1203 can communicate with each other through an internal interface.

[0093] The processor 1202 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0094] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the above-described power flow calculation method based on trust domain model switching.

[0095] This application also provides a computer program product, including a computer program, which, when executed, is used to implement the above-described power flow calculation method based on trust domain model switching.

[0096] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0097] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0098] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0099] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0100] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0101] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0102] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0103] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A power flow calculation method for power systems based on trust domain model switching, characterized in that, Includes the following steps: Based on Kirchhoff's laws, the nodal power balance equations of the target power system are constructed, and the nodal power balance equations are decomposed into power equations in real form. Based on the power equations in real form and the preset nodal constraints, a unified mathematical model of power flow of the power system is constructed. Based on the pre-constructed trust domain optimization framework, a corresponding numerical solver is established, and the constrained optimization problem of the unified mathematical model of the power flow of the power system is solved by the numerical solver to obtain the corresponding solution results. Based on the pre-constructed quadratic model, Gauss-Newton model and the solution results, the predicted decrease in power flow of the power system corresponding to the quadratic model and the Gauss-Newton model is generated respectively. Based on preset numerical stability conditions, the quadratic model or the Gauss-Newton model is dynamically selected as the corresponding target prediction model. The predicted decrease generated by the target prediction model and the actual decrease of the power system flow are obtained. The ratio of the decrease between the actual decrease and the predicted decrease is calculated. The radius of the trust domain is dynamically adjusted using the ratio of the decrease, so as to calculate the power system flow using the adjusted radius of the trust domain.

2. The method according to claim 1, characterized in that, Based on Kirchhoff's laws, the nodal power balance equations of the target power system are constructed, and these equations are decomposed into real-valued power equations. A unified mathematical model of the power system flow is then constructed based on these real-valued power equations and preset nodal constraints, including: Based on Kirchhoff's laws, the node admittance matrix, node voltage vector, and node power injection vector corresponding to the target power system are determined, and the node power balance equation is constructed based on the node admittance matrix, the node voltage vector, and the node power injection vector. The nodal power balance equation is decomposed in the real domain to obtain the power equation in real form, and multiple nodal types corresponding to the target power system are determined to construct nodal constraints for each of the multiple nodal types, wherein the multiple nodal types include PQ nodal type, PV nodal type and slack nodal type. Based on the real-number form of the power equation and the node constraints, a unified mathematical model of power flow in the power system is constructed.

3. The method according to claim 2, characterized in that, The process involves solving the constrained optimization problem of the unified mathematical model of the power system flow using the numerical solver to obtain the corresponding solution results. Based on a preset quadratic model, a Gauss-Newton model, and the solution results, the predicted decrease in power system flow corresponding to the quadratic model and the Gauss-Newton model is generated, respectively, including: The numerical solver iteratively solves the constrained optimization problem of the unified mathematical model of the power flow of the power system. In each iteration, the solution result of the current iteration is calculated using the Jacobian matrix of the current iteration point and the radius of the trust region of the current iteration. The solution results are input into the quadratic model and the Gauss-Newton model respectively. The solution results are then subjected to a preset second-order approximation estimation operation by the quadratic model to generate the predicted decrease in power flow corresponding to the quadratic model. The solution results are then subjected to a preset residual norm comparison operation by the Gauss-Newton model to obtain the predicted decrease in power flow corresponding to the Gauss-Newton model.

4. The method according to claim 3, characterized in that, Based on preset numerical stability conditions, the quadratic model or the Gauss-Newton model is dynamically selected as the corresponding target prediction model. The predicted decrease generated by the target prediction model and the actual decrease in power system flow are obtained. The ratio of the actual decrease to the predicted decrease is calculated, and the confidence region radius is dynamically adjusted using this ratio to calculate the power system flow. This includes: Calculate the reciprocal of the condition number of the trust region radius of the current iteration, and determine whether the predicted decrease is greater than zero and whether the reciprocal of the condition number is greater than a preset numerical stability threshold. If the predicted decrease is greater than zero and the reciprocal of the condition number is greater than the numerical stability threshold, then the quadratic model is selected as the target prediction model for the current iteration; otherwise, the Gauss-Newton model is selected as the target prediction model for the current iteration. Calculate the ratio between the actual decrease and the predicted decrease of the target prediction model in the current iteration, and compare the ratio with a number of preset ratio thresholds to obtain the corresponding comparison results. Adjust the trust domain radius of the current iteration according to the comparison results to obtain the trust domain radius of the next iteration. Determine whether the next iteration meets the preset iteration end requirement. If the next iteration meets the preset iteration end requirement, stop the iteration operation to obtain the power system power flow.

5. A power flow calculation device for a power system based on trust domain model switching, characterized in that, include: The modeling module is used to construct the nodal power balance equations of the target power system based on Kirchhoff's laws, decompose the nodal power balance equations into real-valued power equations, and construct a unified mathematical model of power flow of the power system based on the real-valued power equations and preset nodal constraints. The optimization module is used to establish a corresponding numerical solver based on a pre-built trust domain optimization framework, and solve the constrained optimization problem of the unified mathematical model of the power flow of the power system through the numerical solver to obtain the corresponding solution results. Based on the preset quadratic model, Gauss-Newton model and the solution results, the module generates the predicted decrease in power flow of the power system corresponding to the quadratic model and the Gauss-Newton model, respectively. The calculation module is used to dynamically select the quadratic model or the Gauss-Newton model as the corresponding target prediction model based on preset numerical stability conditions, and to obtain the predicted decrease amount generated by the target prediction model and the actual decrease amount of the power system flow. It also calculates the decrease ratio between the actual decrease amount and the predicted decrease amount, and dynamically adjusts the trust domain radius through the decrease ratio to calculate the power system flow using the adjusted trust domain radius.

6. The apparatus according to claim 5, characterized in that, The modeling module includes: The first construction unit is used to determine the node admittance matrix, node voltage vector and node power injection vector corresponding to the target power system based on the Kirchhoff laws, and to construct the node power balance equation based on the node admittance matrix, the node voltage vector and the node power injection vector. The decomposition unit is used to perform real-domain decomposition on the nodal power balance equation to obtain the power equation in real form, and to determine the multiple nodal types corresponding to the target power system, so as to construct the nodal constraints corresponding to each of the multiple nodal types, wherein the multiple nodal types include PQ nodal type, PV nodal type and slack nodal type; The second building unit is used to construct a unified mathematical model of power system flow based on the real-number form of the power equation and the node constraints.

7. The apparatus according to claim 6, characterized in that, The optimization module includes: The solution unit is used to perform constrained optimization problem solving operations on the unified mathematical model of the power flow of the power system through the numerical solver. In each iteration, the solution result of the current iteration is calculated by using the Jacobian matrix of the current iteration point and the radius of the trust domain of the current iteration. The comparison unit is used to input the solution results into the quadratic model and the Gauss-Newton model respectively, so as to perform a preset second-order approximation estimation operation on the solution results through the quadratic model to generate the predicted decrease in power flow corresponding to the quadratic model, and to perform a preset residual norm comparison operation on the solution results using the Gauss-Newton model to obtain the predicted decrease in power flow corresponding to the Gauss-Newton model.

8. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, the processor executing the program to implement the power flow calculation method for power system based on trust domain model switching as described in any one of claims 1-4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the power flow calculation method based on trust domain model switching as described in any one of claims 1-4.

10. A computer program product, comprising a computer program, characterized in that, The computer program is executed to implement the power flow calculation method based on trust domain model switching as described in any one of claims 1-4.

Images