Active power distribution network reactive voltage optimization method and related equipment

Abstract

The invention provides an active power distribution network reactive voltage optimization method and related equipment. The method comprises the following steps: constructing a centralized optimization model; decomposing the boundary constraint conditions according to sub-partitions, and determining a sub-partition optimization problem; solving each sub-partition optimization model based on the sub-partition optimization problem, converting the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the solved first power flow solution, and determining each sub-partition matrix; extracting coupling information in adjacent sub-partition matrixes to obtain an optimization sub-problem about a Lagrange multiplier coefficient; determining the optimal search direction of each sub-partition; and correcting the first search direction by utilizing the optimal search direction, carrying out iterative updating until a preset condition is met, and ending optimization processing on the reactive voltage of the active power distribution network. According to the embodiment of the invention, in the calculation process, the partitions are mutually independent, only the coupling information of the adjacent nodes of each partition is exchanged, and the information privacy among the sub-partitions is protected.

Description

Technical Field

[0001] The present application relates to the technical field of reactive power voltage optimization, and in particular to a method for optimizing reactive power voltage in an active power distribution network. Background Art

[0002] Distributed reactive power and voltage control scheme is a promising scheme in active distribution networks (ADNs). The problem of voltage violation is becoming more and more serious. During the operation of the power grid, it is unrealistic to centrally optimize the voltage and reactive power of active distribution networks due to the privacy of information and the curse of dimensionality of calculation. Distributed voltage / reactive power control methods have broad applications in active distribution networks (ADNs), but a key issue is how to improve the convergence of distributed algorithms while reducing the communication burden. Therefore, it is necessary to study a new method for voltage / reactive power optimization of active distribution networks with high efficiency and applicable to multiple regions. Summary of the invention

[0003] In view of this, the purpose of this application is to propose an active distribution network reactive power voltage optimization method and related equipment.

[0004] Based on the above objectives, the present application provides a method for optimizing reactive power voltage in an active distribution network, comprising:

[0005] Constructing a centralized optimization model; the objective function of the centralized optimization model is constructed based on the active power and voltage costs of the active distribution network and the inverter losses; the centralized optimization model includes constraints; the constraints include boundary constraints;

[0006] Decomposing the boundary constraint conditions according to sub-partitions to determine sub-partition optimization problems;

[0007] Solving each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction;

[0008] transforming the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution;

[0009] Determine each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model;

[0010] Extracting coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition;

[0011] For the optimization sub-problem, determining the optimal search direction of each of the sub-partitions;

[0012] The first search direction is corrected by using the optimal search direction, and the variables and model parameters of each sub-partition are iteratively updated until the preset conditions are met, thereby ending the optimization process of the reactive voltage of the active distribution network.

[0013] In a possible implementation manner, decomposing the boundary constraint condition according to sub-partitions to determine the sub-partition optimization problem includes:

[0014] On the basis that the solutions of the centralized optimization model before decomposition and the sub-partition optimization model after decomposition are the same, the boundary constraints are decomposed according to sub-partitions to determine the sub-partition optimization problem; wherein, when two adjacent sub-partitions satisfy the coupled equality constraints, the optimal power flow solution of the sub-partition optimization model is the optimal power flow solution of the centralized optimization model.

[0015] In a possible implementation manner, solving each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction includes:

[0016] For each sub-partition optimization problem, a fully distributed model for independent parallel calculation of each sub-partition is obtained by using augmented Lagrangian function decomposition;

[0017] Solving each of the sub-partition optimization models in the fully distributed model to obtain the first power flow solution;

[0018] Based on the optimization principle of the line search algorithm, the first search direction is obtained by calculating the difference in variable values ​​between two adjacent steps.

[0019] In a possible implementation manner, converting the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution includes:

[0020] Based on the first power flow solution and the Lagrangian dual algorithm, adding the weighted sum of the constraints in the sub-partition optimization model to obtain an augmented objective function;

[0021] Based on the augmented objective function, Lagrange multiplier coefficients are set to transform the sub-partition optimization model into the sequential quadratic programming sub-partition optimization model.

[0022] In a possible implementation, the determining each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model includes:

[0023] Determine an approximate Hessian matrix at a solution of the sequential quadratic programming subpartition optimization model, an approximate gradient matrix of the objective function, and a Jacobian matrix in the constraint conditions;

[0024] reducing the size of the approximate gradient matrix and the Jacobian matrix using a null space algorithm;

[0025] The approximate Hessian matrix, the approximate gradient matrix and the Jacobian matrix are converted so that the approximate Hessian matrix is ​​a positive definite matrix;

[0026] Each sub-partition matrix is ​​determined based on the converted approximate Hessian matrix, the converted approximate gradient matrix and the converted Jacobian matrix.

[0027] In a possible implementation manner, determining the optimal search direction of each of the sub-partitions for the optimization sub-problem includes:

[0028] For the optimization subproblem, calculating a step factor for each of the subpartitions;

[0029] Based on the step size factor, the Lagrangian multiplier coefficient of each sub-partition is updated until the gradient is zero, so as to obtain the optimal Lagrangian multiplier coefficient;

[0030] Based on the optimal Lagrange multiplier coefficient, the optimal search direction of each sub-partition is determined.

[0031] Based on the same inventive concept, the embodiment of the present application also provides an active distribution network reactive voltage optimization device, including:

[0032] A construction module is configured to construct a centralized optimization model; the objective function of the centralized optimization model is constructed according to the active power and voltage cost of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints;

[0033] A decomposition module is configured to decompose the boundary constraint condition according to sub-partitions to determine sub-partition optimization problems;

[0034] A solution module is configured to solve each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction;

[0035] A conversion module, configured to convert the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution;

[0036] A determination module is configured to determine each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model;

[0037] An extraction module is configured to extract coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition;

[0038] A search direction module, configured to determine the optimal search direction of each of the sub-partitions for the optimization sub-problem;

[0039] The correction optimization module is configured to use the optimal search direction to correct the first search direction, iteratively update the variables and model parameters of each sub-partition until the preset conditions are met, and end the optimization processing of the reactive voltage of the active distribution network.

[0040] Based on the same inventive concept, an embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, an active distribution network reactive voltage optimization method as described in any one of the above items is implemented.

[0041] Based on the same inventive concept, an embodiment of the present application also provides a non-transitory computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable the computer to execute any of the above-mentioned active distribution network reactive voltage optimization methods.

[0042] Based on the same inventive concept, an embodiment of the present application further provides a computer program product, which includes computer program instructions, and the computer instructions are used to enable the computer program product to execute any of the above-mentioned active distribution network reactive voltage optimization methods.

[0043] As can be seen from the above, the active distribution network reactive voltage optimization method and related equipment provided by the present application are constructed by constructing a centralized optimization model; the objective function of the centralized optimization model is constructed according to the active power and voltage costs and inverter losses of the active distribution network; the centralized optimization model includes constraints; the constraints include boundary constraints; the boundary constraints are decomposed according to sub-partitions to determine the sub-partition optimization problem; each sub-partition optimization model is solved based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction; based on the first power flow solution, the sub-partition optimization model is converted into a sequential quadratic programming sub-partition optimization model; based on the sequential quadratic programming sub-partition optimization model, each sub-partition matrix is ​​determined; the coupling information in the adjacent sub-partition matrices is extracted to obtain the optimization sub-problem of the Lagrange multiplier coefficient of each sub-partition; for the optimization sub-problem, the optimal search direction of each sub-partition is determined; the first search direction is corrected using the optimal search direction, and the variables and model parameters of each sub-partition are iteratively updated until the preset conditions are met, and the optimization processing of the reactive voltage of the active distribution network is terminated. The embodiment of the present application completes the optimization problem of each sub-partition, realizes parallel independent calculation of each partition based on the consistency alternating direction multiplier method, and solves the global optimal search direction model of each partition. Secondly, based on the null space method (Null-method), the computational dimension of the problem is reduced. In the iterative process, no global communication is required, and only the information of neighboring nodes in each region is exchanged to achieve global optimization solution. The calculation results enable the multi-region active distribution network to maximize the use of renewable energy, greatly reducing energy loss. In the calculation process, the partitions are independent of each other, no central coordinator is required, and only the coupling information of adjacent nodes in each partition is exchanged. No global communication is required, which protects the information privacy between each sub-partition. In addition, based on the null space method, the objective function and constraint conditions of large-scale problems are matrix and vector dimension reduction, which reduces the complexity of the calculation method, and proposes a solution for the singularity of the matrix in the solution process, ensuring that the second layer of the algorithm obtains the global optimal search direction of the quadratic programming model (Quadratic Programming, QP) of the original problem, and then corrects the search direction of the first layer of the algorithm, ensuring the accuracy of the algorithm, and improving the convergence speed and robustness of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present application or related technologies, the drawings required for use in the embodiments or related technical descriptions are briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 This is a flow chart of a method for optimizing reactive power voltage in an active power distribution network according to an embodiment of the present application;

[0046] Figure 2 A detailed flow chart of the method for optimizing reactive power voltage in an active power distribution network according to an embodiment of the present application;

[0047] Figure 3 A schematic diagram of a method for dividing multiple regions according to an embodiment of the present application;

[0048] Figure 4 This is a schematic diagram of the distributed and fully distributed frameworks of the algorithm of the embodiment of the present application;

[0049] Figure 5 (a) and (b) are schematic diagrams showing changes in the dual residual during the iteration process of the embodiment of the present application;

[0050] Figure 6 A schematic diagram of information exchange analysis between partitions in an embodiment of the present application;

[0051] Figure 7 This is a schematic diagram of the change of matrix dimension before and after matrix dimensionality reduction based on the null space method in an embodiment of the present application;

[0052] Figure 8 This is a schematic structural diagram of an active power distribution network reactive voltage optimization device according to an embodiment of the present application;

[0053] Fig. 9 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to make the objectives, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in combination with specific embodiments and with reference to the accompanying drawings.

[0055] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the embodiments of the present application should be the usual meanings understood by people with ordinary skills in the field to which the present application belongs. The "first", "second" and similar words used in the embodiments of the present application do not represent any order, quantity or importance, but are only used to distinguish different components. "Including" or "comprising" and similar words mean that the elements or objects appearing in front of the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0056] It is understandable that before using the technical solutions of each embodiment of the present disclosure, the type, scope of use, usage scenarios, etc. of the personal information involved will be informed to the user in an appropriate manner, and the user's authorization will be obtained.

[0057] For example, in response to receiving an active request from a user, a prompt message is sent to the user to clearly remind the user that the operation requested to be performed will require obtaining and using the user's personal information. Thus, the user can independently choose whether to provide personal information to software or hardware such as an electronic device, application, server, or storage medium that performs the operation of the technical solution of the present disclosure according to the prompt message.

[0058] As an optional but non-limiting implementation, in response to receiving the user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. In addition, the pop-up window may also carry a selection control for the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0059] It is understandable that the above notification and the process of obtaining user authorization are merely illustrative and do not constitute a limitation on the implementation of the present disclosure. Other methods that meet relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0060] As mentioned in the background technology section, the distributed reactive power and voltage control scheme is a very promising scheme in active distribution networks (ADNs). The problem of voltage violation is becoming increasingly serious. During the operation of the power grid, due to the privacy of information and the curse of dimensionality of calculation, it is unrealistic to centrally optimize the voltage and reactive power of the active distribution network. Distributed voltage / reactive power control methods have broad applications in active distribution networks (ADNs), but a key issue is how to improve the convergence of distributed algorithms while reducing the communication burden. Therefore, it is necessary to study a new method for active distribution network voltage / reactive power optimization that is efficient and suitable for multiple regions.

[0061] Based on the above considerations, the embodiment of the present application proposes a method for optimizing reactive voltage of an active distribution network, by constructing a centralized optimization model; the objective function of the centralized optimization model is constructed according to the active power and voltage costs and inverter losses of the active distribution network; the centralized optimization model includes constraints; the constraints include boundary constraints; the boundary constraints are decomposed according to sub-partitions to determine the sub-partition optimization problem; each sub-partition optimization model is solved based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction; based on the first power flow solution, the sub-partition optimization model is converted into a sequential quadratic programming sub-partition optimization model; based on the sequential quadratic programming sub-partition optimization model, each sub-partition matrix is ​​determined; the coupling information in the adjacent sub-partition matrices is extracted to obtain the optimization sub-problem of the Lagrange multiplier coefficient of each sub-partition; for the optimization sub-problem, the optimal search direction of each sub-partition is determined; the first search direction is corrected by using the optimal search direction, and the variables and model parameters of each sub-partition are iteratively updated until the preset conditions are met, and the optimization processing of the reactive voltage of the active distribution network is terminated. The embodiment of the present application completes the optimization problem of each sub-partition, realizes parallel independent calculation of each partition based on the consistency alternating direction multiplier method, and solves the global optimal search direction model of each partition. Secondly, based on the null space method (Null-method), the computational dimension of the problem is reduced. In the iterative process, no global communication is required, and only the information of neighboring nodes in each region is exchanged to achieve global optimization solution. The calculation results enable the multi-region active distribution network to maximize the use of renewable energy, greatly reducing energy loss. In the calculation process, the partitions are independent of each other, no central coordinator is required, and only the coupling information of adjacent nodes in each partition is exchanged. No global communication is required, which protects the information privacy between each sub-partition. In addition, based on the null space method, the objective function and constraint conditions of large-scale problems are matrix and vector dimension reduction, which reduces the complexity of the calculation method, and proposes a solution for the singularity of the matrix in the solution process, ensuring that the second layer of the algorithm obtains the global optimal search direction of the quadratic programming model (Quadratic Programming, QP) of the original problem, and then corrects the search direction of the first layer of the algorithm, ensuring the accuracy of the algorithm, and improving the convergence speed and robustness of the algorithm.

[0062] The technical solutions of the embodiments of the present application are described in detail below through specific examples.

[0063] refer to Figure 1 The method for optimizing reactive power voltage of an active distribution network according to an embodiment of the present application comprises the following steps:

[0064] Step S101, constructing a centralized optimization model; the objective function of the centralized optimization model is constructed based on the active power and voltage costs of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints;

[0065] Step S102, decomposing the boundary constraint conditions according to sub-partitions to determine sub-partition optimization problems;

[0066] Step S103, solving each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction;

[0067] Step S104, converting the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution;

[0068] Step S105, determining each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model;

[0069] Step S106, extracting coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition;

[0070] Step S107, determining the optimal search direction of each of the sub-partitions for the optimization sub-problem;

[0071] Step S108, using the optimal search direction to correct the first search direction, iteratively updating the variables and model parameters of each sub-partition until the preset conditions are met, and ending the optimization process of the reactive voltage of the active distribution network.

[0072] refer to Figure 2 , is a detailed flow chart of the method for optimizing reactive power voltage in active distribution network according to an embodiment of the present application. Figure 1 and Figure 2 The embodiments of the present application are described in detail.

[0073] refer to Figure 3 , which is a schematic diagram of the method for multi-region division according to an embodiment of the present application.

[0074] refer to Figure 4 , which is a schematic diagram of the distributed and fully distributed frameworks of the algorithm of the embodiment of the present application.

[0075] The embodiment of this application uses a 5-node and 33-node distribution network example to verify the accuracy and robustness of the analysis algorithm, as well as the information exchange analysis between the various partitions of the algorithm. First, a partition decoupling method is proposed. Figure 3 shown.

[0076] exist Figure 4The difference between the distributed and fully distributed frameworks is compared and analyzed in the paper. The algorithm proposed in this application is mainly based on the parallel solution between sub-partitions in the fully distributed framework, and each sub-partition is solved independently without waiting.

[0077] With respect to step S101, in this embodiment, a centralized optimization model of reactive power and voltage of an active distribution network is established. The model is composed of an objective function and constraints, specifically including:

[0078] Objective function: The model aims to minimize the active power of the distribution network and the combination of voltage cost and inverter loss:

[0079]

[0080] Branch flow constraints:

[0081]

[0082] Node voltage drop constraints:

[0083]

[0084] Node active and reactive power injection flow constraints:

[0085]

[0086] Among them, P ij ,Q ij are the active power flow and reactive power flow from node i to node j, respectively, j ,b j are the shunt conductance and shunt susceptance of node j, is the numbered set of nodes in the system, ε is the numbered set of branches in the system, r ij ,x ij are the resistance and reactance of branch (i, j), P Dj ,Q Dj are the active power and reactive power requirements of load j, respectively, D is the number of the system load node, P Gj ,Q Gj are the active power and reactive power of distributed generation DG, Q cj To compensate for the reactive output of the equipment, the distributed generation node set is represented as The set of compensation equipment CD nodes is expressed as li j is the square of the current amplitude of branch (i, j), v i is the square of the voltage amplitude at node i, v j is the square of the voltage amplitude at node j, P jrepresents the active power of node j, Q j represents the reactive power of node j, P jk ,Q jk They represent the active power flow and reactive power flow from node j to node k respectively.

[0087] The inequality constraints of the active distribution network voltage / reactive power control model include:

[0088]

[0089] V1=V ref (9)

[0090]

[0091] Among them, V i , are the upper and lower limits of the voltage safety amplitude of node i, respectively. ij , are the upper and lower limits of the active power of branch (i, j), Q ij , are the upper and lower limits of the reactive power of branch (i, j), S Gi Represents apparent power, V ref Represents the voltage amplitude of the equilibrium node.

[0092] Furthermore, the non-convex constraints of the above centralized model are relaxed by a second-order cone convexity:

[0093] In this model, there are non-convex nonlinear constraints in the constraints, specifically equality constraints. Therefore, this model is a non-convex nonlinear model. Based on the second-order cone relaxation (SQCP) relaxation method, the non-convex constraints are convexly relaxed, and the relaxed model becomes:

[0094]

[0095] Among them, due to l ij ,v i are the squares of the current and voltage magnitudes, respectively, and have the property of being non-negative, so they are standard rotated second-order cone constraints.

[0096] With respect to step S102, in some embodiments, the decomposing the boundary constraints according to sub-partitions to determine the sub-partition optimization problem includes: on the basis of satisfying that the solutions of the centralized optimization model before decomposition and the sub-partition optimization model after decomposition are the same, decomposing the boundary constraints according to sub-partitions to determine the sub-partition optimization problem; wherein, when two adjacent sub-partitions satisfy the coupled equality constraint, the optimal power flow solution of the sub-partition optimization model is the optimal power flow solution of the centralized optimization model.

[0097] In this embodiment, the boundary condition constraints in the centralized model are decomposed to ensure that the centralized network solutions before and after the decomposition are the same.

[0098] The coupling variables of partition B and partition A need to satisfy the following coupling equality constraints:

[0099]

[0100] Among them, when the adjacent partitions B and A meet the above coupling conditions, the sub-partition optimal power flow (OPF) solution is the centralized OPF optimal solution, and the coupling constraint conditions of partitions B and A are simply described as:

[0101]

[0102] Among them, A A With A B Represent two partition coupling matrices respectively. The sub-partition optimization model after active distribution network decoupling, for the sub-problem partition A∈(1,...,N), the OPF modular consistency alternating direction multiplier method is expressed as:

[0103] It is expressed as:

[0104]

[0105] st:g A (x A,int )=0

[0106] h A (x A，int )<=0

[0107]

[0108] x A,cou =x B,cou (17)

[0109] Among them, x A ∈X A is the feasible domain of neutron partition A, g A (xA,int ) represents the equality constraint of partition A, h A (x A，int ) represents the inequality constraint of partition A, f A (x A，int ) represents the objective function of partition A:

[0110]

[0111] where ε A , They represent internal branches, tie lines and DG equipment respectively.

[0112] In some embodiments, the sub-partition optimization problem is based on solving each of the sub-partition optimization models to obtain a first power flow solution and a first search direction, including: for each of the sub-partition optimization problems, using augmented Lagrangian function decomposition to obtain a fully distributed model in which each sub-partition is independently calculated in parallel; solving each of the sub-partition optimization models in the fully distributed model to obtain the first power flow solution; based on the optimization principle of the line search algorithm, the first search direction is obtained by calculating the numerical difference of the variables between two adjacent steps.

[0113] In some embodiments, the sub-partition optimization model is converted into a sequential quadratic programming sub-partition optimization model based on the first power flow solution, including: based on the first power flow solution and the Lagrangian dual algorithm, adding the weighted sum of the constraints in the sub-partition optimization model to obtain an augmented objective function; based on the augmented objective function, setting the Lagrangian multiplier coefficient to convert the sub-partition optimization model into the sequential quadratic programming sub-partition optimization model.

[0114] In some embodiments, the determining of each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model includes: determining the approximate Hessian matrix of the sequential quadratic programming sub-partition optimization model at the solution, the approximate gradient matrix of the objective function, and the Jacobian matrix in the constraints; using a null space algorithm to reduce the size of the approximate gradient matrix and the Jacobian matrix; transforming the approximate Hessian matrix, the approximate gradient matrix, and the Jacobian matrix so that the approximate Hessian matrix is ​​a positive definite matrix; and determining each sub-partition matrix based on the transformed approximate Hessian matrix, the transformed approximate gradient matrix, and the transformed Jacobian matrix.

[0115] In the embodiment of the present application, the distributed model is solved by applying a two-layer distributed active distribution network reactive power voltage optimization method based on consistent ADMM, and the specific steps are as follows:

[0116] Set each partition variable Consistency constants and the Lagrange multiplier λ 0 , penalty parameter ρ 0 , the coefficient matrix Initial value, set convergence basis ε, maximum number of iterations k max For each of the above partition optimization problems, the augmented Lagrangian function is used to decompose the fully distributed model in which each partition can be independently calculated in parallel.

[0117]

[0118] Among them, z A is the consistency variable of the coupled variables of partition A, λ A is the Lagrange multiplier variable of the coupling constraint, nv i is the number of variables in the sub-region, ρ is the penalty factor, X A Represents the constraints of partition A.

[0119] Import parameter values Correct the penalty factor and solve the optimal power flow solution for each partition A=1,...,N independently and in parallel:

[0120]

[0121] Here, based on the line search algorithm Optimization principle, default step size factor By calculating the difference between the variables in the kth and k+1th steps, we can get the descending search direction:

[0122]

[0123] The solution of each partition variable in the first layer is used as the initial point of the second layer variable Consistency variables and Lagrange multipliers Based on the basic idea of ​​Lagrange duality, the weighted sum of constraints is added to the objective function to obtain its augmented objective function. The Lagrangian function of the original problem and the coupled equality constraints are expressed as:

[0124]

[0125]

[0126] in, are the dual multipliers of nonlinear equality and inequality constraints respectively. The coupled equality constraint of the original problem cannot completely satisfy the value of formula (17) to be 0. Therefore, the coupled equality constraint satisfaction value is set to is the Lagrange multiplier coefficient of the dual problem, and the problem is transformed into a quadratic programming (SQP) model to solve the search direction of the original problem in each partition, which is expressed as:

[0127]

[0128] in, The constant term can be omitted. In the Sequential Quadratic Programming (SQP) model for each partition, The approximate Hessian matrix at Approximate gradient matrix of the objective function Jacobian in Constraints The matrices are represented as:

[0129]

[0130] Among them, the Hessian matrix in formula (25) is solved based on the D-BFGS method to avoid derivation, and the subscript in formula (27) r The set is represented in r Inequality at a feasible point In addition, the Null-space method is used to reduce the size of the coefficient matrix in the objective function and constraint conditions to reduce the computational burden. Formula (24) is reformulated as:

[0131]

[0132] The Chinese name of the D-BFGS method is the diagonally scaled BFGS method. This is a variant of the BFGS (Broyden-Fletcher-Goldfarb-Shanno) method. The BFGS method itself is a quasi-Newton method for solving unconstrained optimization problems. It finds the minimum value of the optimization problem by iteratively updating an approximate Hessian matrix (or its inverse matrix).

[0133] Formulas (25), (26), (27) can be reformulated as ensure is a positive definite matrix. With the iteration, the above matrix is ​​constantly changing. At present, the quadratic programming simplified model of the Taylor second-order expansion of the Lagrangian function of the original problem is obtained. For formula (28) with respect to variables Δv,λ s The derivative is:

[0134]

[0135] The KKT equation is expressed as:

[0136]

[0137] Among them, the KKT equations are called KKT conditions or KKT equations in Chinese, and the full name is Karush-Kuhn-Tucker conditions. They are a set of necessary conditions for finding local optimal solutions in nonlinear programming and are applicable to optimization problems with constraints.

[0138] in, The KKT matrix expression of formula (30) is:

[0139]

[0140] For the KKT matrix The row rank is satisfied if is non-singular and can obtain the unique global optimal solution of the above matrix, which is obtained by the KKT conditional elimination equation method:

[0141]

[0142] The idea of ​​Gaussian iteration is used for the above equations, and each partition matrix is ​​set and vector Simplifying formula (33) we get:

[0143]

[0144] Based on formula (34), we can clearly see that Only the information of the local partition is involved. In order to implement the distributed solution of the above problem, the matrices and vectors corresponding to the coupling variables in the partition are extracted. Information can be used to transform the above Lagrange multiplier model (34) into the following form for distributed computing:

[0145]

[0146] Among them, for any two adjacent subproblems A,B∈N in each partition coupling constraint, there exists A A ,A B The i,jth row in is non-zero, and the definition is A A Middle c i The row is not 0, so consider extracting only the The matrix information related to the coupling variables is obtained by and vector in:

[0147]

[0148] Among them, A A (j A ) represents the first OK. For convenience, we define Finally, solving the optimal value formula (35) of the dual multiplier variable for each partition coupling constraint is equivalent to solving the following unconstrained optimization problem:

[0149]

[0150] In some embodiments, determining the optimal search direction of each of the sub-partitions for the optimization sub-problem includes: calculating the step factor of each of the sub-partitions for the optimization sub-problem; updating the Lagrangian multiplier coefficient of each of the sub-partitions based on the step factor until the gradient is zero to obtain the optimal Lagrangian multiplier coefficient; and determining the optimal search direction of each of the sub-partitions based on the optimal Lagrangian multiplier coefficient.

[0151] Furthermore, the optimal search direction and gradient of the variables are solved:

[0152] After knowing the search direction in the algorithm, solve the step factor for each partition (replace x k+1 =x k +α k p k After substituting the quadratic function into α k minimize)):

[0153]

[0154] in:

[0155]

[0156] Update the Lagrange multiplier coefficients of each partition in the next step by solving the step size factor:

[0157]

[0158] To find the search direction for the next iteration, we need to use the linear combination relationship of the search direction of the previous iteration to update the search direction of the next iteration of each partition:

[0159]

[0160] Until the gradient Get the optimal Lagrange multiplier coefficients for each partition Then the calculation of the local search direction can be expressed as:

[0161]

[0162] Utilize the optimal search direction of the SQP problem for each partition Correct the direction of formula (21):

[0163]

[0164] Based on the above optimal search direction and gradient, update each partition variable, consistency parameter and Lagrange multiplier solution:

[0165]

[0166] (λ) k+1 =(λ s ) k+1 (47)

[0167] Return to the first layer of the algorithm and continue with the k+1th iteration until the algorithm meets the convergence accuracy and the maximum number of iterations.

[0168] refer to Figure 5 (a) and (b) are schematic diagrams showing the changes in the dual residual during the iteration process of an embodiment of the present application.

[0169] refer to Figure 6 , which is a schematic diagram of information exchange analysis between partitions in an embodiment of the present application.

[0170] refer to Figure 7 , which is a schematic diagram of the change in matrix dimension before and after matrix dimensionality reduction based on the null space method in an embodiment of the present application.

[0171] refer to Figure 5 (a) and (b), using the method proposed in this application, the algorithm accuracy ε = 0.0001, the maximum number of iterations K max = 100, the change trend of the search direction during the voltage square iteration of node 3 in partition A is as follows: Figure 5 As shown in the figure, when only the first layer of distributed solution is used without correction, the algorithm does not converge to the correct solution after 100 iterations. However, after using the corrected search direction, the algorithm converges to the optimal solution after 8 iterations.

[0172] First, taking the IEEE33 node three-partition as an example, when the iteration number k=1, in the process of the first layer calculation, each sub-partition is completely independent of each other, and there is no information exchange between two adjacent partitions. Therefore, here we mainly analyze the information exchange between two adjacent partitions in the second layer sub-partition calculation process of the algorithm of this application, where the Hessian matrix information exchange of each sub-partition variable is as follows: Figure 6As shown, using each partition variable obtained in the first layer, the Hessian matrix information A of each partition is obtained based on D-BFGS. In the second layer of the algorithm, the coupling information B in the Hessian matrix of each partition is obtained, and the boundary coupling information of two adjacent partitions is summed to obtain C, which is used to solve the unconstrained problem of the Lagrange multiplier coefficients of its sub-partition. Additional information exchange is avoided during the calculation process to protect the privacy between the partitions.

[0173] In addition, Figure 7 The Null-space method is used to simplify the dimensionality of the Hessian matrix (Resuced-Hessianquasi-Newton approximation). As the dimensionality decreases, the efficiency of the subsequent calculation algorithm is significantly improved. The matrix A is a full row rank matrix and guarantees is a positive definite matrix, let The eigenvalue of is eig. If eig≤=0, The matrix diagonal adds a very small value of 10 -6 .

[0174] It can be seen from the above embodiments that the method for optimizing reactive voltage of an active distribution network described in the embodiment of the present application is to construct a centralized optimization model; the objective function of the centralized optimization model is constructed according to the active power and voltage costs of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints; the boundary constraints are decomposed according to sub-partitions to determine the sub-partition optimization problem; each sub-partition optimization model is solved based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction; based on the first power flow solution, the sub-partition optimization model is converted into a sequential quadratic programming sub-partition optimization model; based on the sequential quadratic programming sub-partition optimization model, each sub-partition matrix is ​​determined; the coupling information in the adjacent sub-partition matrices is extracted to obtain the optimization sub-problem of the Lagrange multiplier coefficient of each sub-partition; for the optimization sub-problem, the optimal search direction of each sub-partition is determined; the first search direction is corrected using the optimal search direction, and the variables and model parameters of each sub-partition are iteratively updated until the preset conditions are met, and the optimization processing of the reactive voltage of the active distribution network is terminated. The embodiment of the present application completes the optimization problem of each sub-partition, realizes parallel independent calculation of each partition based on the consistency alternating direction multiplier method, and solves the global optimal search direction model of each partition. Secondly, based on the null space method (Null-method), the computational dimension of the problem is reduced. In the iterative process, no global communication is required, and only the information of neighboring nodes in each region is exchanged to achieve global optimization solution. The calculation results enable the multi-region active distribution network to maximize the use of renewable energy, greatly reducing energy loss. In the calculation process, the partitions are independent of each other, no central coordinator is required, and only the coupling information of adjacent nodes in each partition is exchanged. No global communication is required, which protects the information privacy between each sub-partition. In addition, based on the null space method, the objective function and constraint conditions of large-scale problems are matrix and vector dimension reduction, which reduces the complexity of the calculation method, and proposes a solution for the singularity of the matrix in the solution process, ensuring that the second layer of the algorithm obtains the global optimal search direction of the quadratic programming model (Quadratic Programming, QP) of the original problem, and then corrects the search direction of the first layer of the algorithm, ensuring the accuracy of the algorithm, and improving the convergence speed and robustness of the algorithm.

[0175] It should be noted that the method of the embodiment of the present application can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario and completed by multiple devices cooperating with each other. In the case of such a distributed scenario, one of the multiple devices can only perform one or more steps in the method of the embodiment of the present application, and the multiple devices will interact with each other to complete the described method.

[0176] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the above embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0177] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides an active distribution network reactive voltage optimization device.

[0178] refer to Figure 8 , the active distribution network reactive voltage optimization device comprises:

[0179] A construction module 81 is configured to construct a centralized optimization model; the objective function of the centralized optimization model is constructed according to the active power and voltage costs of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints;

[0180] A decomposition module 82 is configured to decompose the boundary constraint condition according to sub-partitions to determine sub-partition optimization problems;

[0181] A solving module 83 is configured to solve each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction;

[0182] A conversion module 84 is configured to convert the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution;

[0183] A determination module 85 is configured to determine each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model;

[0184] An extraction module 86 is configured to extract coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition;

[0185] A search direction module 87, configured to determine the optimal search direction of each of the sub-partitions for the optimization sub-problem;

[0186] The correction optimization module 88 is configured to use the optimal search direction to correct the first search direction, iteratively update the variables and model parameters of each sub-partition until the preset conditions are met, and end the optimization processing of the reactive voltage of the active distribution network.

[0187] For the convenience of description, the above device is described in terms of functions divided into various modules. Of course, when implementing the present application, the functions of each module can be implemented in the same or multiple software and / or hardware.

[0188] The device of the above embodiment is used to implement the corresponding active distribution network reactive voltage optimization method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0189] Based on the same inventive concept, corresponding to any of the above-mentioned embodiments and methods, the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the active distribution network reactive voltage optimization method described in any of the above embodiments is implemented.

[0190] Fig. 9 A more specific schematic diagram of the hardware structure of an electronic device provided in this embodiment is shown, and the device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040 are connected to each other through the bus 1050 in the device.

[0191] The processor 1010 can be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0192] The memory 1020 may be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 may store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program codes are stored in the memory 1020 and are called and executed by the processor 1010.

[0193] The input / output interface 1030 is used to connect the input / output module to realize information input and output. The input / output module can be configured in the device as a component (not shown in the figure), or it can be externally connected to the device to provide corresponding functions. The input device may include a keyboard, a mouse, a touch screen, a microphone, various sensors, etc., and the output device may include a display, a speaker, a vibrator, an indicator light, etc.

[0194] The communication interface 1040 is used to connect a communication module (not shown) to realize communication interaction between the device and other devices. The communication module can realize communication through a wired mode (such as USB, network cable, etc.) or a wireless mode (such as mobile network, WIFI, Bluetooth, etc.).

[0195] The bus 1050 includes a path that transmits information between the various components of the device (eg, the processor 1010, the memory 1020, the input / output interface 1030, and the communication interface 1040).

[0196] It should be noted that, although the above device only shows the processor 1010, the memory 1020, the input / output interface 1030, the communication interface 1040 and the bus 1050, in the specific implementation process, the device may also include other components necessary for normal operation. In addition, it can be understood by those skilled in the art that the above device may also only include the components necessary for implementing the embodiments of the present specification, and does not necessarily include all the components shown in the figure.

[0197] The electronic device of the above embodiment is used to implement the corresponding active distribution network reactive voltage optimization method in any of the above embodiments, and has the beneficial effects of the corresponding method embodiment, which will not be repeated here.

[0198] Based on the same inventive concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the active distribution network reactive voltage optimization method as described in any of the above embodiments.

[0199] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.

[0200] The computer instructions stored in the storage medium of the above embodiment are used to enable the computer to execute the active distribution network reactive voltage optimization method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0201] Based on the same inventive concept, corresponding to the active distribution network reactive voltage optimization method described in any of the above embodiments, the present disclosure also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer so that the computer and / or the processor execute the active distribution network reactive voltage optimization method. Corresponding to the execution subject corresponding to each step in each embodiment of the active distribution network reactive voltage optimization method, the processor that executes the corresponding step may belong to the corresponding execution subject.

[0202] The computer program product of the above embodiment is used to enable the computer and / or the processor to execute the active distribution network reactive voltage optimization method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0203] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present application (including the claims) is limited to these examples. In line with the concept of the present application, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the embodiments of the present application as described above, which are not provided in detail for the sake of simplicity.

[0204] In addition, to simplify the description and discussion, and in order not to make the embodiments of the present application difficult to understand, the known power supply / ground connection with the integrated circuit (IC) chip and other components may or may not be shown in the provided drawings. In addition, the device can be shown in the form of a block diagram to avoid making the embodiments of the present application difficult to understand, and this also takes into account the fact that the details of the implementation of these block diagram devices are highly dependent on the platform to be implemented in the embodiments of the present application (that is, these details should be fully within the scope of understanding of those skilled in the art). In the case of elaborating specific details (e.g., circuits) to describe exemplary embodiments of the present application, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details or when these specific details are changed. Therefore, these descriptions should be considered to be illustrative rather than restrictive.

[0205] Although the present application has been described in conjunction with specific embodiments of the present application, many replacements, modifications and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0206] The embodiments of the present application are intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included in the scope of protection of the present application.

Claims

1. A method for optimizing reactive power voltage in an active distribution network, characterized in that: include: Build a centralized optimization model; The objective function of the centralized optimization model is constructed based on the active power and voltage cost of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints; Decomposing the boundary constraint conditions according to sub-partitions to determine sub-partition optimization problems; Solving each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction; transforming the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution; Determine each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model; Extracting coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition; For the optimization sub-problem, determining the optimal search direction of each of the sub-partitions; The first search direction is corrected by using the optimal search direction, and the variables and model parameters of each sub-partition are iteratively updated until the preset conditions are met, thereby ending the optimization process of the reactive voltage of the active distribution network.

2. The method according to claim 1, characterized in that Decomposing the boundary constraint conditions according to sub-partitions to determine sub-partition optimization problems includes: On the basis that the solutions of the centralized optimization model before decomposition and the sub-partition optimization model after decomposition are the same, the boundary constraints are decomposed according to sub-partitions to determine the sub-partition optimization problem; wherein, when two adjacent sub-partitions satisfy the coupled equality constraints, the optimal power flow solution of the sub-partition optimization model is the optimal power flow solution of the centralized optimization model.

3. The method according to claim 1, characterized in that The step of solving each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction includes: For each sub-partition optimization problem, a fully distributed model for independent parallel calculation of each sub-partition is obtained by using augmented Lagrangian function decomposition; Solving each of the sub-partition optimization models in the fully distributed model to obtain the first power flow solution; Based on the optimization principle of the line search algorithm, the first search direction is obtained by calculating the difference in variable values ​​between two adjacent steps.

4. The method according to claim 1, characterized in that: The converting the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution includes: Based on the first power flow solution and the Lagrangian dual algorithm, adding the weighted sum of the constraints in the sub-partition optimization model to obtain an augmented objective function; Based on the augmented objective function, Lagrange multiplier coefficients are set to transform the sub-partition optimization model into the sequential quadratic programming sub-partition optimization model.

5. The method according to claim 1, characterized in that The determining of each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model includes: Determine an approximate Hessian matrix at a solution of the sequential quadratic programming subpartition optimization model, an approximate gradient matrix of the objective function, and a Jacobian matrix in the constraint conditions; reducing the size of the approximate gradient matrix and the Jacobian matrix using a null space algorithm; The approximate Hessian matrix, the approximate gradient matrix and the Jacobian matrix are converted so that the approximate Hessian matrix is ​​a positive definite matrix; Each sub-partition matrix is ​​determined based on the converted approximate Hessian matrix, the converted approximate gradient matrix and the converted Jacobian matrix.

6. The method according to claim 1, characterized in that The step of determining the optimal search direction for each of the sub-partitions for the optimization sub-problem includes: For the optimization subproblem, calculating a step factor for each of the subpartitions; Based on the step size factor, the Lagrangian multiplier coefficient of each sub-partition is updated until the gradient is zero, so as to obtain the optimal Lagrangian multiplier coefficient; Based on the optimal Lagrange multiplier coefficient, the optimal search direction of each sub-partition is determined.

7. An active distribution network reactive voltage optimization device, characterized in that: include: A building module, configured to build a centralized optimization model; The objective function of the centralized optimization model is constructed based on the active power and voltage cost of the active distribution network and the inverter loss; the centralized optimization model includes constraints; the constraints include boundary constraints; A decomposition module is configured to decompose the boundary constraint condition according to sub-partitions to determine sub-partition optimization problems; A solution module is configured to solve each of the sub-partition optimization models based on the sub-partition optimization problem to obtain a first power flow solution and a first search direction; A conversion module, configured to convert the sub-partition optimization model into a sequential quadratic programming sub-partition optimization model based on the first power flow solution; A determination module is configured to determine each sub-partition matrix based on the sequential quadratic programming sub-partition optimization model; An extraction module is configured to extract coupling information in adjacent sub-partition matrices to obtain an optimization sub-problem about Lagrange multiplier coefficients for each sub-partition; A search direction module, configured to determine the optimal search direction of each of the sub-partitions for the optimization sub-problem; The correction optimization module is configured to use the optimal search direction to correct the first search direction, iteratively update the variables and model parameters of each sub-partition until the preset conditions are met, and end the optimization processing of the reactive voltage of the active distribution network.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the method according to any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 6.

10. A computer program product, comprising computer program instructions, which, when executed on a computer, cause the computer to execute the method according to any one of claims 1 to 6.

Images