Power system frequency modulation method based on joint moment optimal control under Irai random disturbance

Abstract

The invention discloses a power system frequency modulation method based on joint moment optimal control under Italian random disturbance, and the method comprises the following steps: a, inputting model parameters, including a power system parameter, a frequency modulation index, and a new energy drift / diffusion coefficient; step b, establishing a random optimal frequency modulation model of the power system under random disturbance of the Italian; c, reconstructing the random optimal frequency modulation model into a deterministic joint moment optimal control model based on a joint moment evolution equation; step d, projecting the deterministic joint moment optimal control model into a finite-dimensional Bernstein polynomial space, and converting the model into a finite-dimensional optimization problem easy to solve; and f, solving a finite dimension optimization problem to obtain an approximate optimal control strategy of the original joint moment optimal control model.

Description

Technical Field

[0001] This invention belongs to the field of frequency regulation technology, specifically relating to a power system frequency regulation method based on joint moment optimal control under Iton random disturbance. Background Technology

[0002] Frequency is a key indicator for power system operation and control, and ensuring stable power system frequency operation is of great significance to its safety and economy. However, with the widespread integration of new energy generating units, the random fluctuation characteristics of new energy output are becoming increasingly apparent, posing a serious threat to power system frequency stability. Therefore, it is necessary to consider the power system frequency regulation problem under the random fluctuations of new energy output.

[0003] The stochastic fluctuations in renewable energy output constitute a complex, time-varying stochastic distribution, also known as a stochastic process. Existing research primarily employs discrete-time stochastic distributions, such as Gaussian, Beta, and Weibull distributions, to describe renewable energy output. However, these methods neglect both the stochastic evolution of randomness within time periods and its autocorrelation over time. This makes it difficult for frequency regulation methods to obtain sufficient information to adequately address the stochastic fluctuations in renewable energy output. Therefore, it is necessary to evolve the traditional discrete-time stochastic distribution into a continuous-time Iton stochastic differential equation to more comprehensively describe the stochastic evolution characteristics of renewable energy output during power system frequency regulation.

[0004] On the other hand, the frequency regulation problem of power systems considering the stochasticity of new energy sources is a stochastic optimal control problem that is difficult to solve directly. It contains infinite-dimensional variables and constraints, and is essentially an infinite-dimensional optimization problem. Existing classical optimization methods, such as interior-point methods, simplex methods, particle swarm optimization, genetic algorithms, and the use of mature commercial solvers, are the mainstream solutions for deterministic finite-dimensional optimization problems, but they are not applicable to this stochastic infinite-dimensional optimization problem. To solve the stochastic optimal control problem, the mainstream methods include the optimality condition method based on the maximum principle and Hamilton-Jacobi-Bellman equations, which obtains the optimal strategy by solving the optimality equations. However, in the frequency regulation of power systems with complex constraints, the complexity of the optimality equations will increase sharply and become difficult to solve (Wang Yayu, Peng Yonggang, Sun Jing. A review of research on automatic generation control technology and strategies for wind farms [J / OL]. Power System Technology, 1-22 [2025-08-05]. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, the purpose of this invention is to provide a power system frequency regulation method based on joint moment optimal control under Iton random disturbances. This method improves existing stochastic optimal control theory from the perspectives of moment reconstruction and polynomial projection, providing a more refined power system frequency regulation modeling method and an efficient control strategy solution method. This is beneficial for improving the solution speed and accuracy of power system frequency regulation control, aiming to provide a fast and effective solution method for power system frequency regulation under strong random disturbances. It can improve the control accuracy and decision-making efficiency of power system frequency regulation under the widespread penetration of new energy sources, aligning with my country's goal of building a new power system that vigorously develops new energy power generation, and is expected to bring good social and economic benefits.

[0006] The present invention is achieved by at least one of the following technical solutions.

[0007] A power system frequency regulation method based on joint moment optimal control under Iton random disturbances includes the following steps: Step a: Based on the power system parameters, frequency regulation index parameters, and new energy drift / diffusion coefficient, establish a stochastic optimal frequency regulation model for the power system under Ito's stochastic disturbance; Step b: Reconstruct the stochastic optimal frequency modulation model into a deterministic joint moment optimal control model based on the joint moment evolution equation; Step c: Project the deterministic joint moment optimal control model into a finite-dimensional Bernstein polynomial space, transforming it into a finite-dimensional optimization problem that is easy to solve; Step d: Obtain the approximate optimal control strategy of the original joint moment optimal control model by solving the finite-dimensional optimization problem.

[0008] Furthermore, the stochastic optimal frequency regulation model of the power system under Iton stochastic disturbances includes endogenous states, exogenous states, control strategies, constraints, and objective functions, wherein: Endogenous state represents the internal operating state of a power system, including frequency, mechanical power, and branch power flow; endogenous state is a dynamic process. Exogenous states refer to external information that may affect the operation of the power system, namely, the randomness of the output power of new energy sources and the power of loads; Control decisions are the controllable decision variables of a power system, namely the frequency regulation power of generator units; Constraints: During frequency regulation, by setting inequality constraints between the upper and lower limits of the generator set's frequency regulation power, the frequency regulation power is kept within the set boundaries. Objective function: The objective of power system frequency regulation is to minimize the stochastic integral of the square of the regional control error and the control cost.

[0009] Furthermore, during frequency modulation, the exogenous state follows a stochastic dynamic process described by Itō's stochastic differential equation.

[0010] Furthermore, the constraints also include opportunities to keep the power system state under random disturbances within a safe operating range.

[0011] Furthermore, the area control error is defined as the weighted sum of frequency deviation and line transmission power deviation.

[0012] Further, in step b, the stochastic optimal frequency modulation model is reconstructed into a deterministic joint moment optimal control model based on the joint moment evolution equation, including: The stochastic and deterministic dynamic equations in the stochastic optimal frequency modulation model are uniformly described as abstract stochastic differential equations. The expectation of this abstract stochastic differential equation is taken to obtain the first-order moment evolution equation obeyed by the first-order moment function. The second-order moment functions of the endogenous and exogenous state vectors are calculated, and the second-order moment functions are differentiated according to Itoh's calculus theory to obtain the second-order moment evolution equation. Based on the above first-order and second-order moment functions, the objective function of the stochastic optimal frequency modulation model is reconstructed. For the chance constraints, a Gaussian distribution approximation is first used for transformation, and then the transformed chance constraints are further transformed into deterministic constraints based on the first-order and second-order moment evolution equations. Finally, a deterministic joint moment optimal control model is obtained.

[0013] Further, in step c, the deterministic joint moment optimal control model is projected into a finite-dimensional Bernstein polynomial space, transforming it into an easily solvable finite-dimensional optimization problem, including the following steps: First, inequality auxiliary variables are introduced to describe the joint moment optimal control model in an abstract form. Second, based on the Bellman optimality principle, the joint moment optimal control model is transformed into an optimality condition equation, i.e., a partial differential equation containing an optimization problem. Third, a multidimensional Bernstein polynomial basis and its corresponding finite-dimensional Bernstein space are constructed based on multiple one-dimensional Bernstein polynomial bases, and the variables in the optimality condition are expressed as elements in the finite-dimensional Bernstein space. The objective of the optimality condition is integrated, transforming the optimality condition into an infinite-dimensional optimization problem. The equality constraints of the infinite-dimensional optimization problem are projected onto each Bernstein polynomial basis to obtain the corresponding finite-dimensional equality constraints. The inequality constraints of the infinite-dimensional optimization problem are transformed into finite-dimensional inequality constraints about the polynomial coefficients based on the convex hull property of polynomials. Finally, the finite-dimensional optimization problem corresponding to the original joint moment optimal control model is obtained.

[0014] Furthermore, an optimization solver is used to solve the finite-dimensional optimization problem.

[0015] A computer device according to the present invention includes a memory and a processor, the memory being electrically connected to the processor, the memory storing a computer program, characterized in that: when the computer program is executed by the processor, the processor causes the processor to implement the method described thereon.

[0016] The present invention provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor implements the method described herein.

[0017] Compared with existing technologies, the beneficial effects of the present invention are as follows: This invention addresses the frequency regulation problem in power systems by proposing a stochastic optimal frequency regulation method based on Itō's stochastic differential equations. The research content is closely related to practical needs, and since there is no mature experience to draw upon, it is of significant importance. Compared with existing power system frequency regulation methods, this invention mainly has the following advantages: (1) This invention establishes for the first time a stochastic optimal frequency regulation model based on Iton's stochastic differential equation, which can accurately describe the stochastic disturbance characteristics of new energy output and the impact of such stochastic disturbance on the frequency dynamics of the power system.

[0018] (2) This invention proposes for the first time a method for reconstructing a stochastic optimal frequency modulation model based on the joint moment evolution equation. By constructing the joint moment evolution equation, the stochastic optimal frequency modulation model is simplified into a deterministic joint moment optimal control model, which significantly reduces the difficulty of solving it.

[0019] (3) This invention proposes for the first time a solution method for the joint moment optimal control strategy based on optimality condition projection. By projecting the optimality condition of the joint moment optimal control model onto the Bernstein polynomial space for processing, it can simplify the infinite-dimensional optimization problem that is difficult to solve into a finite-dimensional optimization problem that is easy to solve, thereby significantly improving its solution efficiency. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as limiting the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 The flowchart illustrates a power system frequency regulation method based on joint moment optimal control under Iton random disturbances, as provided in an embodiment of the present invention.

[0022] Figure 2 This is a schematic diagram of the system frequency deviation after applying the method of the present invention.

[0023] Figure 3 This is a schematic diagram of the regional control error after applying the method of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0025] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the specific embodiments described herein are only for explaining this application and are not intended to limit this application.

[0026] like Figure 1 As shown, the power system frequency regulation method based on joint moment optimal control under Iton random disturbances in this embodiment includes the following steps: Step a: Based on the power system parameters, frequency regulation index parameters, and new energy drift / diffusion coefficient, establish a stochastic optimal frequency regulation model for the power system under Ito's stochastic disturbance.

[0027] The stochastic optimal frequency modulation model includes endogenous states, exogenous states, control strategies, constraints, and an objective function, wherein: 1) Endogenous State: The endogenous state represents the internal operating state of the power system, including frequency, mechanical power, and branch power flow, and is defined as follows: (1) in, It is an endogenous state vector; It is the frequency vector of each generator set; This is the mechanical power vector of each generator set; It is the power flow vector of each branch.

[0028] These endogenous states obey the following deterministic dynamic processes: (2) (3) (4) in, It is a set of generator node nodes; It is a node The frequency; It is a node The frequency; It is a node The generator set inertia coefficient; It is a node The damping coefficient of the generator set; It is a node The time inertia constant of the generator set; It is a node The mechanical power of the generator set; It is a node The output power of the new energy generator units; It is a node The load power; and From node To the node and nodes To the node The side road; and They are branch roads and The power flow on the current; It is a node The generator set droop coefficient; Represents a node The frequency regulation power of the generator set; It is a collection of branch paths; It is a side road The equivalent impedance coefficient is defined as (5) in, It is a node at the basic running point. voltage value, It is a node at the basic running point. The voltage value; It is a node at the basic running point. The phase angle value, It is a node at the basic running point. The phase angle value; It is a side road The impedance value.

[0029] 2) Exogenous State: Exogenous state refers to external information that may affect the operation of the power system, namely the stochastic output power of new energy sources and load power, defined as follows: (6) in, It is an exogenous state vector; The power of each new energy unit ; It is the load power of each node. The vector.

[0030] As one example, during frequency modulation, the exogenous state follows a stochastic dynamic process described by Itō's stochastic differential equations: (7) in, It is the set of indices of exogenous states; It is the first exogenous state One component; It is the first exogenous state One component; yes The drift constant; yes by The drift coefficient affected; It is the first An independent Brownian motion; yes The set of subscripts; yes about The diffusion coefficient.

[0031] 3) Control Decision: Control decision refers to the controllable decision variables of the power system, namely the frequency regulation power of the generator units. : (8) in, It is the control decision vector.

[0032] 4) Constraints: During frequency modulation, the frequency modulation power cannot exceed a given boundary, which can be described by the following inequality constraint: (9) in, and It is a generator set Upper and lower limits of frequency modulation power; Furthermore, the system state under random disturbances also needs to be maintained within a given safe operating range, which can be described as the following chance constraint: (10) in, It is the set of indices of the opportunity constraint; It is the probability of an event occurring; and These are the parameters of the state-chance constraint; It's the confidence level.

[0033] 5) Objective Function: The objective of power system frequency regulation is to minimize the stochastic integral of the square of the regional control error and the control cost, as follows: (11) in, It is the expectation operator; It is the frequency modulation region, specifically a set of nodes; It is the frequency modulation range A set; It is a set of generator node sets; It is the discount factor; Indicates at time Time zone Regional control error; It is the cost control coefficient.

[0034] Specifically, regional control error Can be defined as (12) in, It is the frequency modulation range The set of generator node units; It is the frequency modulation range A collection of connecting lines; It is a region The deviation coefficient.

[0035] For ease of subsequent application, the objective function can be described in the following abstract form: (13) in, and It is the second-order coefficient matrix in the objective function; It is the first-order coefficient vector in the objective function.

[0036] Step b: Based on the joint moment evolution equation, reconstruct the stochastic optimal frequency modulation model into a deterministic joint moment optimal control model.

[0037] The aforementioned stochastic optimal frequency modulation model is a stochastic optimal control problem, which is difficult to solve directly due to the complexity of the Iton stochastic process. Therefore, this invention reconstructs the stochastic optimal frequency modulation model into a deterministic optimal control model based on the joint moment evolution equation, thereby reducing the difficulty of solving the model. This mainly includes the following two steps: (1) Derivation of the joint moment evolution equation: First, to facilitate the subsequent reconstruction process, the stochastic dynamic process equation (7) and the deterministic dynamic equations (2)-(4) in the above stochastic optimal frequency modulation model are uniformly described as the following abstract form of stochastic differential equations: (14) in, , , It is the coefficient matrix of the deterministic dynamic process equations (2) to (4); , , It is the coefficient matrix of the stochastic dynamic process equation (7); It is a Brownian motion vector; and It is the coefficient vector of the control strategy, i.e. .

[0038] Taking the expectation of the stochastic differential equation (14), we can obtain the first-order moment evolution equation as follows: (15) in, It is an endogenous state The first moment; It is an exogenous state The first moment.

[0039] Endogenous state and exogenous state vectors A second-order matrix can be calculated as: (16) in, , , , It is the second-order moment matrix and the endogenous state and exogenous state The four corresponding block matrices.

[0040] Based on Itoh's calculus theory, differentiating (16) yields the second-order moment evolution equation as follows:

[0041] (17) (2) Reconstruction of the joint moment optimal control model: Based on the first and second moments mentioned above, the objective function (11) for power system frequency regulation can be reconstructed into a deterministic form as follows: (18) in, It is the trace operation of a matrix.

[0042] Furthermore, for the chance constraint (10), it can be approximated by a Gaussian distribution and transformed into: (19) in, It is a Gaussian distribution quantiles; yes The expectation.

[0043] Based on the aforementioned evolution equations of the first and second moments, equation (19) can be transformed into a deterministic form as follows: (20) Thus, the stochastic optimal frequency modulation model is transformed into a deterministic joint moment optimal control model, which includes a deterministic objective function (18), deterministic differential equations (2)-(4), (15), and (17), and deterministic inequality constraints (9) and (20).

[0044] Step c: Project the deterministic joint moment optimal control model into a finite-dimensional Bernstein polynomial space to transform it into a finite-dimensional optimization problem that is easy to solve.

[0045] The joint moment optimal control model obtained above is an infinite-dimensional optimization problem, and its optimal strategy is difficult to solve directly. Therefore, this invention further proposes a strategy solution method based on the optimality condition projection technique. This method comprises three steps: First, based on the Bellman optimality principle, the joint moment optimal control model is transformed into optimality condition equations. However, this is an infinite-dimensional problem, difficult to solve directly. Therefore, this invention further projects this infinite-dimensional problem into a finite-dimensional Bernstein polynomial space to transform it into a finite-dimensional optimization problem that is easier to solve. Finally, by solving this finite-dimensional optimization problem, an approximate optimal control strategy for the original joint moment optimal control model is obtained. The specific process is as follows: c1) Derivation of Optimality Conditions: To simplify the subsequent derivation, the aforementioned joint moment optimal control model is rewritten in the following abstract form: (twenty one) (twenty two) (twenty three) (twenty four) in, It is a joint moment vector, containing first-order moments. , and second moment , , , ; and yes At any moment and time The possible values ​​of ; and yes At any moment and time The possible values ​​of ; It is the immediate cost function in the objective function (18); It is the right-hand side function in differential equations (2)-(4), (16), and (17); It is the left-hand side function of the inequality constraint equations (9) and (20); It is a vector of auxiliary variables for inequalities.

[0046] Based on the Bellman optimality principle, the optimal control strategy of the joint moment optimal control model (21)-(23) obeys the following optimality condition equations: (25) (26) (27) in, It is a vector differential operator; It is the optimal value function for the joint moment vector; It concerns the control strategy for the joint moment vector; The right-hand side term of the differential equation is about A changing function; Inequality constraints on the left-hand side about A changing function; It is a function of the auxiliary variables with respect to the joint moment vector.

[0047] c2) Finite-dimensional projection: Solving the optimality condition equation (25) yields the optimal control strategy. However, this equation is an infinite-dimensional partial differential equation containing an optimization problem, making it extremely difficult to solve. Therefore, this invention projects the optimality condition onto a finite-dimensional Bernstein polynomial space for solution. First, the variables in (25) are expressed as elements in a finite-dimensional Bernstein polynomial space: (28) (29) (30) in, , , yes , , The polynomial projection approximation; It is the set of subscripts of a multidimensional Bernstein polynomial basis; It is a multidimensional indicator; It is the first Polynomial basis subscripts of dimension; It is a joint moment vector dimensionality; yes The Polynomial coefficients; yes The Polynomial coefficients; yes The One coefficient; It is the first about the joint moment vector The multidimensional Bernstein polynomial basis is constructed by tensor product of multiple one-dimensional Bernstein polynomial bases, as follows: (31) in, yes The Components in each dimension; It is the first A one-dimensional Bernstein polynomial basis is defined as follows: (32) in, and yes The upper and lower limits; It is the order of the Bernstein polynomial.

[0048] By examining the joint moment vector in the optimality condition (25) Integrating and substituting into equations (28)-(30) and (25), we can equivalently transform them into an infinite-dimensional optimization problem as follows: (33) (34) (35) (36) in, , and The objective function, differential equation, and inequality constraints are related to... A changing function.

[0049] For the infinite-dimensional equality constraints (34) and (35), projecting them onto each Bernstein polynomial basis yields the following projection equations: (37) (38) in, The inner product of a given function and Bernstein polynomial basis functions is defined as: (39) in, Represents any function.

[0050] It is worth noting that the integral operations in the projection equations (37) and (38) can be simplified to integral constants with respect to the polynomial basis functions only, and can be pre-calculated to reduce the computational load. Therefore, equations (37) and (38) are merely finite-dimensional equality constraints with respect to the polynomial coefficients to be determined.

[0051] Furthermore, based on the convex hull property of Bernstein polynomials, the infinite-dimensional inequality constraint (36) can be transformed into: (40) This can be further simplified to the following finite-dimensional inequality constraint with respect to the coefficients: (41) Thus, the original joint moment optimal control model is transformed into a finite-dimensional optimization model, which includes equations (33), (37), (38), and (41).

[0052] Step d: Obtain the approximate optimal control strategy of the original joint moment optimal control model by solving the finite-dimensional optimization problem.

[0053] As one example, a mainstream optimization solver can be used for rapid solution. The optimal control strategy obtained from solving the finite-dimensional optimization problem can be expressed as: (42) in, It is the optimal control strategy; It is a coefficient The optimal value.

[0054] Step e, in the t At each control moment, observe the current system state. ; Step f: Calculate control decisions based on the optimal control strategy. And execute; Step g, set time t = t +1, if the current time t equal to the termination time T If the above steps are completed, the process ends; otherwise, return to step e.

[0055] The above embodiments of the present invention are merely examples for clearly illustrating the present invention and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively describe all possible implementations here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A power system frequency regulation method based on joint moment optimal control under Ito stochastic disturbance, characterized in that, The method comprises the following steps: Step a, establishing a stochastic optimal frequency regulation model of a power system under Ito stochastic disturbance according to power system parameters, frequency modulation index parameters, and new energy drift / diffusion coefficients; Step b, reconstructing the stochastic optimal frequency regulation model into a deterministic joint moment optimal control model based on joint moment evolution equations; Step c, projecting the deterministic joint moment optimal control model into a finite-dimensional Bernstein polynomial space to convert it into a finite-dimensional optimization problem that is easy to solve; Step d, obtaining an approximate optimal control strategy of the original joint moment optimal control model by solving the finite-dimensional optimization problem.

2. The method of claim 1, wherein the method is based on the joint-moment optimal control under the Itô stochastic perturbation. The stochastic optimal frequency regulation model of the power system under Ito stochastic disturbance comprises endogenous states, exogenous states, control strategies, constraint conditions, and an objective function, wherein: The endogenous states represent the internal operating states of the power system, including frequency, mechanical power, and branch power flow, and a dynamic process to which the endogenous states are subject; The exogenous states represent external information that may affect the operation of the power system, i.e., random new energy output power and load power; The control decision is a controllable decision variable of the power system, i.e., the frequency modulation power of a generator unit; The constraint condition is that, during frequency modulation, by setting inequality constraints of the upper and lower limits of the frequency modulation power of the generator unit, the frequency modulation power is within the set boundary; The objective function is to minimize the stochastic integral of the regional control error square and the control cost.

3. The method of claim 2, wherein the method is based on the joint-moment optimal control under the Itô stochastic perturbation. During frequency modulation, the exogenous states are subject to a random dynamic process described by an Ito stochastic differential equation.

4. The method of claim 2, wherein the method is based on the joint-moment optimal control under the Itô stochastic perturbation. The constraint condition further comprises keeping the states of the power system under random disturbance within a safe operating interval through a chance constraint.

5. The method of claim 2, wherein the method is based on the joint-moment optimal control under the Itô stochastic perturbation. The regional control error is defined as the weighted sum of the frequency deviation and the line transmission power deviation.

6. The method of claim 2, wherein the method is based on the joint-moment optimal control under the Itô stochastic perturbation. In step b, the stochastic optimal frequency regulation model is reconstructed into a deterministic joint moment optimal control model based on joint moment evolution equations, comprising: The random dynamic equation and the deterministic dynamic equation in the stochastic optimal frequency regulation model are uniformly described as an abstract form of a stochastic differential equation; the first moment evolution equation to which the first moment function is subject is obtained by taking the expectation of the abstract form of the stochastic differential equation; the second moment function of the endogenous state and the exogenous state vector is calculated, and the second moment evolution equation is obtained by differentiating the second moment function according to the Ito calculus theory; the objective function of the stochastic optimal frequency regulation model is reconstructed based on the above first moment and second moment functions; for the chance constraint, the Gaussian distribution approximation is first used for transformation, and then the transformed chance constraint is further transformed into a deterministic form of constraint based on the first moment and the second moment evolution equations, and finally the deterministic joint moment optimal control model is obtained.

7. The power system frequency modulation method based on joint moment optimal control under Ito stochastic disturbance according to claim 1, wherein, in step c, the deterministic joint moment optimal control model is projected into a finite-dimensional Bernstein polynomial space to convert it into a finite-dimensional optimization problem that is easy to solve, comprising the following steps: Firstly, the joint moment optimal control model is described in abstract form by introducing inequality auxiliary variables; secondly, the joint moment optimal control model is transformed into an optimality condition equation, i.e. a partial differential equation containing an optimization problem, based on the Bellman optimality principle; thirdly, a multi-dimensional Bernstein polynomial basis and the corresponding finite-dimensional Bernstein space are constructed based on multiple one-dimensional Bernstein polynomial bases, and the variables in the optimality condition are expressed as elements in the finite-dimensional Bernstein space; The objective of the optimality condition is integrated to equivalently transform the optimality condition into an infinite-dimensional optimization problem; the equality constraints of the infinite-dimensional optimization problem are projected onto each Bernstein polynomial basis to obtain corresponding finite-dimensional equality constraints; the inequality constraints of the infinite-dimensional optimization problem are transformed into finite-dimensional inequality constraints on polynomial coefficients based on the convex hull property of polynomials; finally, a finite-dimensional optimization problem corresponding to the original joint moment optimal control model is obtained.

8. The method of claim 7, wherein the method is based on the joint-moment optimal control of power system frequency regulation under Itô stochastic perturbation. The finite-dimensional optimization problem is solved by using an optimization solver.

9. A computer device comprising a memory and a processor, the memory and the processor being electrically connected, the memory storing a computer program, characterized in that: The computer program, when executed by the processor, causes the processor to implement the method of any one of claims 1-7.

10. A computer-readable storage medium, the computer-readable storage medium storing a computer program, characterized in that: The computer program, when executed by the processor, causes the processor to implement the method of any one of claims 1-7.

Images