An uncertain economic dispatch method considering new energy access to power system

Abstract

The application discloses a kind of considering the uncertainty economic dispatch method of new energy access power system, first, the initial polyhedral mathematical model of new energy output uncertainty considering active / reactive coupling relationship is established;Then, through sampling identification safety violation sample point, drive adaptive polyhedral partitioning to initial polyhedral mathematical model, and explicitly create special polyhedral partitioning, form the partitioning set covering complete uncertainty space;Linear mapping model is constructed for each partition, and the uncertainty is mapped to power system state variable;Based on this, generate robust safe operation constraint considering worst case, with adaptive safety margin;Finally, integrate all partition constraints, construct a two-stage minimax robust optimization framework, to realize the unified optimization of safety and economy.The application can realize the fine description and efficient processing of new energy uncertainty space.

Description

Technical Field

[0001] This invention relates to the technical field of power system operation and control, and in particular to an economic dispatch method that takes into account the uncertainty of new energy access to the power system. Background Technology

[0002] As the penetration rate of new energy sources in the power system continues to increase, the uncertainty and volatility of their output pose a huge challenge to the safe and stable operation of the system. As the core operational decision-making link of the power system, economic dispatch must pursue economic operation while ensuring system safety.

[0003] Existing technologies have the following shortcomings when dealing with the above problems: Traditional robust optimization methods use simple box-type uncertainty sets, which are difficult to accurately characterize the spatiotemporal correlation of new energy output, resulting in excessively robust results that seriously affect economic costs; Monte Carlo sampling-based methods can approximate the real safety domain, but the computational burden is heavy and it is difficult to meet the requirements of real-time scheduling; Traditional convex relaxation methods have gaps when dealing with non-convex power flow constraints, and the constructed safety constraints are not accurate enough, making the scheduling results infeasible in actual power systems; Existing set theory methods usually adopt uniform or static strategies when partitioning uncertainty sets, which cannot be dynamically adjusted according to the sensitive direction of power system security, resulting in excessive linearization errors in the constructed safety constraints.

[0004] Therefore, there is an urgent need in this field for an innovative technical solution that can accurately characterize high-dimensional uncertainty in a compact and efficient manner and quickly construct operational constraints for safe operation, so as to maximize the economic operation level of the system while ensuring absolute safety. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings and deficiencies of the prior art and to provide an economic dispatch method that considers the uncertainty of new energy access to the power system, so as to achieve a refined characterization and efficient processing of the uncertainty space of new energy.

[0006] To achieve the above objectives, the technical solution provided by this invention is: an economic dispatch method considering the uncertainty of new energy access to the power system, comprising the following steps: S1. Refined modeling of the uncertainty space of new energy, namely, establishing an initial polyhedral mathematical model of the uncertainty of new energy output. The initial polyhedral mathematical model adopts a mathematical structure combining the center point and the generator matrix. The center point represents the expected or typical value of new energy output, and the generator matrix describes the correlation of the range and direction of the change of new energy output in each dimension of new energy units. S2. In order to refine the initial polyhedral mathematical model constructed in step S1 into a set of multiple sub-polyhedra that can balance computational complexity and security, the initial polyhedral mathematical model is sampled and verified. A large number of new energy power output scenario sample points are generated by Monte Carlo method. Power system security constraints are checked for each sample point to identify safety violation sample points that violate line capacity constraints, node voltage constraints, and generator output constraints. Based on the spatial distribution characteristics of the identified safety violation sample points, the initial polyhedral mathematical model is adaptively partitioned into polyhedra. S3. Considering that the adaptive polyhedral partitioning in step S2 is based on random sampling and violation sample driving, it cannot fully cover all low-probability, high-impact extreme operating scenarios. Therefore, a special polyhedral partitioning for extreme operating scenarios is explicitly created to establish a complete extreme scenario database and ensure that low-probability, high-impact events are fully covered. S4. In order to efficiently and accurately map the adaptive polyhedral partitions and special polyhedral partitions obtained in steps S2 and S3, as well as their uncertainties, to the state variables of the power system, namely the voltage magnitude and phase angle of each node, a linearized mapping model for each partition of the adaptive polyhedral partitions and special polyhedral partitions is formed. S5. Based on the linearized mapping model established for each partition in step S4, the uncertainty in each partition is transformed into the fluctuation range of the power system state variables through linear mapping, and operation constraints are generated to ensure that the power system can operate safely in the obtained adaptive polyhedral partition and special polyhedral partition. S6. In order to achieve economic optimization under the premise of satisfying the operational constraints of safe operation of the power system, the operational constraints of safe operation obtained by adaptive polyhedral partitioning and special polyhedral partitioning are integrated into the robust optimization model, and a two-stage minimax robust optimization framework is constructed to achieve unified optimization of safety and economy.

[0007] Furthermore, step S1 includes: An initial polyhedral mathematical model incorporating uncertainties in the active and reactive power of new energy sources is established, employing a mathematical structure that combines the center point with the generator matrix. : ; In the formula, The vector at the center point represents the expected value of the new energy output; The generator matrix describes the correlation between the range and direction of the change in new energy output across various new energy units. For the generator coefficient vector, in ; The number of generators; The Monte Carlo method is used to generate test scenario sample points in a polyhedral space, with each test scenario sample point corresponding to a specific combination of new energy power output. : ; In the formula, For the first A coefficient vector randomly generated from sample points in a test scenario. This represents the number of samples.

[0008] Furthermore, step S2 includes: For each test scenario sample point, a power system safety constraint check is performed to identify safety violation sample points that violate line capacity constraints, node voltage constraints, and generator output constraints, forming a set of violation samples. : ; In the formula, The function representing the degree of violation of the j-th safety constraint; Establish a zoned accuracy control mechanism, which involves determining the maximum permissible linearization error through error tolerance parameters and power system sensitivity analysis. : ; In the formula, The preset error tolerance parameter has a range of values. , is the condition number of the Jacobian matrix; The spectral norm of the generator matrix represents the precision of the current partition. A multi-strategy fusion approach is adopted to select the optimal partitioning direction by weighting and combining different scoring strategies. , means as follows: ; In the formula, The three scoring functions correspond to: the direction of the right singular vector of the minimum singular value of the Jacobian matrix calculated based on the power flow equation; the direction of analysis based on equipment capacity margin analysis and unit output margin; and the direction of principal component analysis results based on violation sample points. These are the weight coefficients corresponding to each scoring function, used to adjust the relative importance of different strategies in the optimal partitioning direction; they satisfy the following relationship: and ; By adaptive step size Generate candidate partitioning schemes: ; In the formula, For adaptive step size, it is dynamically adjusted based on the current partition depth, including: ; In the formula, The initial step size, As the attenuation factor, The current partitioning depth is set; each candidate partitioning scheme undergoes rigorous feasibility verification, including physical constraint satisfaction checks, linearization error boundary verification, and power system operability assessment.

[0009] Furthermore, in step S3, a special polyhedral partition is explicitly created for extreme operating scenarios, including: Based on the characteristics of the power output combination of new energy units, extreme scenarios are divided into: boundary extreme scenarios where all units are simultaneously at the upper or lower limit of output; unit component failure scenarios where a single line, a single unit, or a single transformer is damaged; correlation extreme scenarios where the output of all units is highly correlated or dispersed; and high-risk scenario combinations identified based on historical operating data. Different polyhedron range ratio parameters are adopted for different types of extreme scenarios to ensure that the extreme scenario partitioning fully covers the risk situation while avoiding excessive conservatism. At the same time, the physical feasibility of each extreme scenario polyhedron is verified by power flow calculation to ensure that the partitioning is within the feasible operating domain of the power system.

[0010] Furthermore, in step S4, the process of constructing the linearized mapping model includes: Optimal reference point weighting based on voltage stability The linearization reference point is selected by evaluating the voltage stability margin and constraint satisfaction of the candidate points: ; In the formula, This refers to the voltage stability margin, specifically the width between the initial reference point and the voltage amplitude boundary. The weighting factor is the minimum voltage stability margin. The weighting coefficients for the average voltage margin are used to select the operating point with a high stability score as the linearization benchmark point through the above weighting scoring method. Tikhonov regularization is used to treat ill-conditioned Jacobian matrices to ensure computational and numerical stability. ; In the formula, The Jacobian matrix is ​​calculated based on the power flow at the reference point. It is the minimum singular value of the Jacobian matrix; It is a small constant with a range of values. ; It is an unit diagonal matrix with the same row and column dimensions as the Jacobian matrix; For Tikhonov regularization parameters, Let be the inverse of the Jacobian matrix after Tikhonov regularization; Evaluate the accuracy of the linearized model at each partition vertex, boundary point, and interior point: ; In the formula, The maximum error calculated for linearization, This is the set of test points selected from vertices, boundary points, and interior points on each partition. To linearize the calculated voltage amplitude, To obtain the voltage amplitude at the current test point from the AC power flow, the linearization model error is checked by calculating the maximum voltage amplitude; through partitioned linearization quality assessment, partitions with errors exceeding the threshold are marked or further partitioned.

[0011] Furthermore, step S5 includes: Adaptive safety margin adjustment strategy based on partition quality Taking into account the influence of the generator matrix norm and the maximum error of linearization calculation, it is expressed as follows: ; In the formula, The norm of the generator matrix The maximum error of linearization calculation The corresponding weighting coefficients; Calculate robust voltage boundaries by combining standard interval arithmetic and vertex sampling methods: ; In the formula, The minimum and maximum voltage boundaries are calculated using the standard interval arithmetic and vertex sampling methods. For the robust voltage minimum and maximum boundaries of the standard interval arithmetic, The minimum and maximum robust voltage boundaries are calculated for vertex sampling; Transform the above linearization constraints into robust linear constraints. : ; In the formula, This is the voltage sensitivity matrix.

[0012] Furthermore, in step S6, the specific details of the two-stage min-max robust optimization framework are as follows: The first stage consists of decision variables, including unit start-up and shutdown status and basic output plan; the second stage consists of adjustable variables, including real-time rescheduling and renewable energy curtailment; the objective function is to minimize the total cost, which includes the cost related to the decision variables in the first stage, and the maximum cost of the adjustable variables in the second stage under the worst-case uncertainty scenario.

[0013] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1. This invention systematically characterizes the uncertainty space of new energy output by constructing an initial polyhedral mathematical model and subsequent adaptive polyhedral partitioning and special polyhedral partitioning; based on the partitioned linearized mapping model, robust safe operation constraints are generated, which can more accurately approximate the complex safety boundary composed of power flow equations of the power system, effectively avoiding the problem of poor economic cost caused by excessive robustness of traditional robust optimization methods.

[0014] 2. The adaptive polyhedral partitioning of this invention is not a simple exhaustive approach, but rather an intelligent selection of the adaptive partitioning direction based on key information such as power system sensitivity analysis and violation distribution of sample points in the test scenario. It constructs high-precision operational constraints for safe operation with near-linear complexity. Its computational efficiency is far higher than that of the Monte Carlo method, which relies on a large number of scenario simulations, and also superior to the traditional convex relaxation method. It can meet the strict requirements of real-time power system dispatch for computational speed.

[0015] 3. The present invention adopts a partitioning strategy that can dynamically adjust based on the real-time power flow Jacobian matrix and the power system operating point, track changes in the power grid state, and handle high-dimensional uncertainties. At the same time, the entire method system is based on rigorous polyhedral set operations, linearization theory, and Tikhonov regularization, which ensures the numerical stability of the model and the clear physical meaning of the solution results, and achieves a reliable unified optimization of safety and economy. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention.

[0017] Figure 2 This is a flowchart of the adaptive polyhedral partitioning method in the present invention. Detailed Implementation

[0018] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0019] This embodiment discloses an economic dispatch method that considers the uncertainty of new energy access to the power system, such as... Figure 1 As shown, the specific details are as follows: S1. Refined modeling of the uncertainty space of new energy sources, namely, establishing an initial polyhedral mathematical model of the uncertainty of new energy output. The initial polyhedral mathematical model adopts a mathematical structure combining a center point and a generator matrix, where the center point represents the expected or typical value of new energy output, and the generator matrix describes the correlation of the range and direction of change of new energy output in various dimensions of new energy units, as detailed below: An initial polyhedral mathematical model incorporating uncertainties in the active and reactive power of new energy sources is established, employing a mathematical structure that combines the center point with the generator matrix. : ; In the formula, The vector at the center point represents the expected value of the new energy output; The generator matrix describes the correlation between the range and direction of the change in new energy output across various new energy units. For the generator coefficient vector, in ; The number of generators; The Monte Carlo method is used to generate test scenario sample points in a polyhedral space, with each test scenario sample point corresponding to a specific combination of new energy power output. : ; In the formula, For the first A coefficient vector randomly generated from sample points in a test scenario. This represents the number of samples.

[0020] S2. To refine the initial polyhedral mathematical model constructed in step S1 into a set of multiple sub-polyhedra that balances computational complexity and security, the initial polyhedral mathematical model is systematically sampled and verified. A large number of sample points from new energy power output scenarios are generated using the Monte Carlo method. Each sample point is checked for power system safety constraints to identify safety violations that violate line capacity constraints, node voltage constraints, and generator output constraints. Based on the spatial distribution characteristics of the identified safety violation sample points, the initial polyhedral mathematical model is adaptively partitioned into polyhedral sections, such as... Figure 2 As shown, the details are as follows: For each test scenario sample point, a power system safety constraint check is performed to identify safety violation sample points that violate line capacity constraints, node voltage constraints, and generator output constraints, forming a set of violation samples. : ; In the formula, The function representing the degree of violation of the j-th safety constraint; Establish a zoned accuracy control mechanism, which involves determining the maximum permissible linearization error through error tolerance parameters and power system sensitivity analysis. : ; In the formula, The preset error tolerance parameter has a range of values. , is the condition number of the Jacobian matrix; The spectral norm of the generator matrix represents the precision of the current partition. A multi-strategy fusion approach is adopted to select the optimal partitioning direction by weighting and combining different scoring strategies. , means as follows: ; In the formula, The three scoring functions correspond to: the direction of the right singular vector of the minimum singular value of the Jacobian matrix calculated based on the power flow equation; the direction of analysis based on equipment capacity margin analysis and unit output margin; and the direction of principal component analysis results based on violation sample points. These are the weight coefficients corresponding to each scoring function, used to adjust the relative importance of different strategies in the optimal partitioning direction; they satisfy the following relationship: and ; By adaptive step size Generate candidate partitioning schemes: ; In the formula, For adaptive step size, it is dynamically adjusted based on the current partition depth, including: ; In the formula, The initial step size, As the attenuation factor, The current partitioning depth is set; each candidate partitioning scheme undergoes rigorous feasibility verification, including physical constraint satisfaction checks, linearization error boundary verification, and power system operability assessment.

[0021] S3. Considering that the adaptive polyhedral partitioning in step S2 is based on random sampling and violation sample-driven methods, it cannot fully cover all low-probability, high-impact extreme operating scenarios. Therefore, a special polyhedral partitioning for extreme operating scenarios is explicitly created to establish a complete extreme scenario database, ensuring that low-probability, high-impact events are fully covered, as follows: Based on the characteristics of the power output combination of new energy units, extreme scenarios are divided into: boundary extreme scenarios where all units are simultaneously at the upper or lower limit of output; unit component failure scenarios where a single line, a single unit, or a single transformer is damaged; correlation extreme scenarios where the output of all units is highly correlated or dispersed; and high-risk scenario combinations identified based on historical operating data. Different polyhedron range ratio parameters are adopted for different types of extreme scenarios to ensure that the extreme scenario partitioning fully covers the risk situation while avoiding excessive conservatism. At the same time, the physical feasibility of each extreme scenario polyhedron is verified by power flow calculation to ensure that the partitioning is within the feasible operating domain of the power system.

[0022] S4. To efficiently and accurately map the adaptive polyhedral partitioning and specific polyhedral partitioning obtained in steps S2 and S3, along with their uncertainties, to the state variables of the power system, namely the voltage magnitude and phase angle of each node, a linearized mapping model for each partition of the adaptive polyhedral partitioning and specific polyhedral partitioning is formed, as follows: Optimal reference point weighting based on voltage stability The linearization reference point is selected by evaluating the voltage stability margin and constraint satisfaction of the candidate points: ; In the formula, This refers to the voltage stability margin, specifically the width between the initial reference point and the voltage amplitude boundary. The weighting factor is the minimum voltage stability margin. The weighting coefficients for the average voltage margin are used to select the operating point with a high stability score as the linearization benchmark point through the above weighting scoring method. Tikhonov regularization is used to treat ill-conditioned Jacobian matrices to ensure computational and numerical stability. ; In the formula, The Jacobian matrix is ​​calculated based on the power flow at the reference point. It is the minimum singular value of the Jacobian matrix; It is a small constant with a range of values. ; It is an unit diagonal matrix with the same row and column dimensions as the Jacobian matrix; For Tikhonov regularization parameters, Let be the inverse of the Jacobian matrix after Tikhonov regularization; Evaluate the accuracy of the linearized model at each partition vertex, boundary point, and interior point: ; In the formula, The maximum error calculated for linearization, This is the set of test points selected from vertices, boundary points, and interior points on each partition. To linearize the calculated voltage amplitude, To obtain the voltage amplitude at the current test point from the AC power flow, the linearization model error is checked by calculating the maximum voltage amplitude; through partitioned linearization quality assessment, partitions with errors exceeding the threshold are marked or further partitioned.

[0023] S5. Based on the linearized mapping model established for each partition in step S4, the uncertainties within each partition are transformed into the fluctuation range of the power system state variables through linear mapping, and based on this, operating constraints are generated to ensure the safe operation of the power system within the obtained adaptive polyhedral partitions and special polyhedral partitions, as follows: Adaptive safety margin adjustment strategy based on partition quality Taking into account the influence of the generator matrix norm and the maximum error of linearization calculation, it is expressed as follows: ; In the formula, The norm of the generator matrix The maximum error of linearization calculation The corresponding weighting coefficients; Calculate robust voltage boundaries by combining standard interval arithmetic and vertex sampling methods: ; In the formula, The minimum and maximum voltage boundaries are calculated using the standard interval arithmetic and vertex sampling methods. For the robust voltage minimum and maximum boundaries of the standard interval arithmetic, The minimum and maximum robust voltage boundaries are calculated for vertex sampling; Transform the above linearization constraints into robust linear constraints. : ; In the formula, This is the voltage sensitivity matrix.

[0024] S6. To achieve economic optimization while satisfying the operational constraints for safe operation of the power system, the operational constraints for safe operation obtained from adaptive polyhedral partitioning and specialized polyhedral partitioning are integrated into the robust optimization model. A two-stage minimax robust optimization framework is constructed to achieve unified optimization of safety and economy. The specific details of the two-stage minimax robust optimization framework are as follows: The first stage consists of decision variables, including unit start-up and shutdown status and basic output plan; the second stage consists of adjustable variables, including real-time rescheduling and renewable energy curtailment; the objective function is to minimize the total cost, which includes the cost related to the decision variables in the first stage, and the maximum cost of the adjustable variables in the second stage under the worst-case uncertainty scenario.

[0025] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A method for economic dispatch considering the uncertainty of new energy sources accessing the power system, characterized in that, Includes the following steps: S1. Refined modeling of the uncertainty space of new energy, namely, establishing an initial polyhedral mathematical model of the uncertainty of new energy output. The initial polyhedral mathematical model adopts a mathematical structure combining the center point and the generator matrix. The center point represents the expected or typical value of new energy output, and the generator matrix describes the correlation of the range and direction of the change of new energy output in each dimension of new energy units. S2. In order to refine the initial polyhedral mathematical model constructed in step S1 into a set of multiple sub-polyhedra that can balance computational complexity and security, the initial polyhedral mathematical model is sampled and verified. A large number of new energy power output scenario sample points are generated by Monte Carlo method. Power system security constraints are checked for each sample point to identify safety violation sample points that violate line capacity constraints, node voltage constraints, and generator output constraints. Based on the spatial distribution characteristics of the identified safety violation sample points, the initial polyhedral mathematical model is adaptively partitioned into polyhedra. S3. Considering that the adaptive polyhedral partitioning in step S2 is based on random sampling and violation sample driving, it cannot fully cover all low-probability, high-impact extreme operating scenarios. Therefore, a special polyhedral partitioning for extreme operating scenarios is explicitly created to establish a complete extreme scenario database and ensure that low-probability, high-impact events are fully covered. S4. In order to efficiently and accurately map the adaptive polyhedral partitions and special polyhedral partitions obtained in steps S2 and S3, as well as their uncertainties, to the state variables of the power system, namely the voltage magnitude and phase angle of each node, a linearized mapping model for each partition of the adaptive polyhedral partitions and special polyhedral partitions is formed. S5. Based on the linearized mapping model established for each partition in step S4, the uncertainty in each partition is transformed into the fluctuation range of the power system state variables through linear mapping, and operation constraints are generated to ensure that the power system can operate safely in the obtained adaptive polyhedral partition and special polyhedral partition. S6. In order to achieve economic optimization under the premise of satisfying the operational constraints of safe operation of the power system, the operational constraints of safe operation obtained by adaptive polyhedral partitioning and special polyhedral partitioning are integrated into the robust optimization model, and a two-stage minimax robust optimization framework is constructed to achieve unified optimization of safety and economy.

2. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 1, characterized in that, Step S1 includes: An initial polyhedral mathematical model incorporating uncertainties in the active and reactive power of new energy sources is established, employing a mathematical structure that combines the center point with the generator matrix. : ； In the formula, The vector at the center point represents the expected value of the new energy output; The generator matrix describes the correlation between the range and direction of the change in new energy output across various new energy units. For the generator coefficient vector, in ; The number of generators; The Monte Carlo method is used to generate test scenario sample points in a polyhedral space, with each test scenario sample point corresponding to a specific combination of new energy power output. : ； In the formula, For the first A coefficient vector randomly generated from sample points in a test scenario. This represents the number of samples.

3. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 2, characterized in that, Step S2 includes: For each test scenario sample point, a power system safety constraint check is performed to identify safety violation sample points that violate line capacity constraints, node voltage constraints, and generator output constraints, forming a set of violation samples. : ； In the formula, The function representing the degree of violation of the j-th safety constraint; Establish a zoned accuracy control mechanism, which involves determining the maximum permissible linearization error through error tolerance parameters and power system sensitivity analysis. : ； In the formula, The preset error tolerance parameter has a range of values. , is the condition number of the Jacobian matrix; The spectral norm of the generator matrix represents the precision of the current partition. A multi-strategy fusion approach is adopted to select the optimal partitioning direction by weighting and combining different scoring strategies. , means as follows: ； In the formula, The three scoring functions correspond to: the direction of the right singular vector of the minimum singular value of the Jacobian matrix calculated based on the power flow equation; the direction of analysis based on equipment capacity margin analysis and unit output margin; and the direction of principal component analysis results based on violation sample points. These are the weight coefficients corresponding to each scoring function, used to adjust the relative importance of different strategies in the optimal partitioning direction; they satisfy the following relationship: and ; By adaptive step size Generate candidate partitioning schemes: ； In the formula, For adaptive step size, it is dynamically adjusted based on the current partition depth, including: ； In the formula, The initial step size, As the attenuation factor, The current partitioning depth is set; each candidate partitioning scheme undergoes rigorous feasibility verification, including physical constraint satisfaction checks, linearization error boundary verification, and power system operability assessment.

4. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 3, characterized in that, In step S3, a special polyhedral partition is explicitly created for extreme operating scenarios, including: Based on the characteristics of the power output combination of new energy units, extreme scenarios are divided into: boundary extreme scenarios where all units are simultaneously at the upper or lower limit of output; unit component failure scenarios where a single line, a single unit, or a single transformer is damaged; correlation extreme scenarios where the output of all units is highly correlated or dispersed; and high-risk scenario combinations identified based on historical operating data. Different polyhedron range ratio parameters are adopted for different types of extreme scenarios to ensure that the extreme scenario partitioning fully covers the risk situation while avoiding excessive conservatism. At the same time, the physical feasibility of each extreme scenario polyhedron is verified by power flow calculation to ensure that the partitioning is within the feasible operating domain of the power system.

5. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 4, characterized in that, In step S4, the process of constructing the linearized mapping model includes: Optimal reference point weighting based on voltage stability The linearization reference point is selected by evaluating the voltage stability margin and constraint satisfaction of the candidate points: ； In the formula, This refers to the voltage stability margin, specifically the width between the initial reference point and the voltage amplitude boundary. The weighting factor is the minimum voltage stability margin. The weighting coefficients for the average voltage margin are used to select the operating point with a high stability score as the linearization benchmark point through the above weighting scoring method. Tikhonov regularization is used to treat ill-conditioned Jacobian matrices to ensure computational and numerical stability. ； In the formula, The Jacobian matrix is ​​calculated based on the power flow at the reference point. It is the minimum singular value of the Jacobian matrix; It is a small constant with a range of values. ; It is an unit diagonal matrix with the same row and column dimensions as the Jacobian matrix; For Tikhonov regularization parameters, Let be the inverse of the Jacobian matrix after Tikhonov regularization; Evaluate the accuracy of the linearized model at each partition vertex, boundary point, and interior point: ； In the formula, The maximum error calculated for linearization, This is the set of test points selected from vertices, boundary points, and interior points on each partition. To linearize the calculated voltage amplitude, To obtain the voltage amplitude at the current test point from the AC power flow, the linearization model error is checked by calculating the maximum voltage amplitude; through partitioned linearization quality assessment, partitions with errors exceeding the threshold are marked or further partitioned.

6. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 5, characterized in that, Step S5 includes: Adaptive safety margin adjustment strategy based on partition quality Taking into account the influence of the generator matrix norm and the maximum error of linearization calculation, it is expressed as follows: ； In the formula, The norm of the generator matrix The maximum error of linearization calculation The corresponding weighting coefficients; Calculate robust voltage boundaries by combining standard interval arithmetic and vertex sampling methods: ； In the formula, The minimum and maximum voltage boundaries are calculated using the standard interval arithmetic and vertex sampling methods. For the robust voltage minimum and maximum boundaries of the standard interval arithmetic, The minimum and maximum robust voltage boundaries are calculated for vertex sampling; Transform the above linearization constraints into robust linear constraints. : ； In the formula, This is the voltage sensitivity matrix.

7. The economic dispatch method for considering the uncertainty of new energy access to the power system according to claim 6, characterized in that, In step S6, the specific details of the two-stage minimax robust optimization framework are as follows: The first stage consists of decision variables, including unit start-up and shutdown status and basic output plan; the second stage consists of adjustable variables, including real-time rescheduling and renewable energy curtailment; the objective function is to minimize the total cost, which includes the cost related to the decision variables in the first stage, and the maximum cost of the adjustable variables in the second stage under the worst-case uncertainty scenario.

Images