Power distribution network distribution robust optimization method and system considering multiple correlation of new energy
Through the data-driven distribution robust optimization method, considering the multiple correlations of new energy output power, a two-stage distribution robust optimization model for flexible distribution network is constructed, which solves the uncertainty brought about by new energy power generation in the distribution network. The problem improves the stability and adaptability of the system.
Patent Information
- Application Number
- CN202411065288.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-05
- Publication Date
- 2024-11-26
AI Technical Summary
In distribution networks containing new energy power generation, the application of flexible power electronic converter technology brings challenges in system optimization and regulation. Traditional methods are difficult to effectively deal with uncertainty, resulting in bottlenecks in FDN optimization operation.
Using a data-driven distribution robust optimization method, considering the conditional correlation between the time, space and predicted value of the output power of new energy and the prediction error, generate correlation samples, and construct a two-stage distribution robust optimization model of flexible distribution networks, combined with Coordinated control measures for multiple sources, network loads, and optimal flow model are established, and optimized scheduling scheme is solved through mixed integer second-order cone planning.
Improve the stability and reliability of the distribution network, effectively deal with the imbalance of the time and space distribution of new energy output and the complexity of the adjustment of massive devices, enhance the adaptability and flexibility of the system, and better cope with uncertainty and change.
Smart Images

Figure CN119029844A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of distribution network distributed robust optimization, and in particular to a distribution network distributed robust optimization method and system taking into account multiple correlations of new energy sources. Background Art
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] At present, the rapid growth of renewable energy generation has brought many problems to the safe operation of the power system, and the widespread application of flexible power electronic commutation technology has also brought challenges to system optimization and regulation. With the rapid development of flexible power electronic technology, flexible multi-state switch (FMSS) is applied to distribution networks, which can achieve load balancing and improve voltage distribution by providing continuous reactive power compensation, and can achieve effective balance between load and renewable energy through power flow transfer, solving some bottleneck problems in the development of traditional distribution networks.
[0004] Flexible distribution network (FDN) realizes closed-loop power supply through FMSS interconnection, which has the advantages of promoting the consumption of new energy, improving power distribution, and rapid power regulation. It will become the development trend of smart distribution network in the future. At present, the investment and operation cost issues limit the large-scale application of FMSS, and it is impossible to completely replace other primary voltage regulation equipment in the short term. This makes the operation optimization of FDN need to consider the coordination and cooperation of FMSS and traditional control measures from the overall perspective.
[0005] At present, the main methods for dealing with uncertainty problems are stochastic programming and robust optimization. Distributed robust optimization (DRO) combines the advantages of both methods, dealing with uncertainty factors by searching for the worst probability distribution under uncertain conditions, and finally obtaining the optimal solution. It is currently widely used in the fields of scheduling optimization, integrated energy, and transmission planning, but less involved in the optimization operation of FDN. Therefore, exploring the application of data-driven DRO in FDN has important research significance and value. Summary of the invention
[0006] In order to solve at least one technical problem existing in the above-mentioned background technology, the present invention provides a distribution network distributed rod optimization method and system considering the multiple correlations of new energy. The flexible distribution network distributed rod optimization strategy considering the multiple correlations of new energy takes the minimization of active loss and voltage deviation of the distribution network as the goal, takes into account the multiple correlations of new energy in time, space and power dimensions, establishes a two-stage distributed rod optimization model of flexible distribution network based on data-driven, and derives and constructs the optimal power flow model of flexible distribution network containing multiple source-grid-load coordinated control measures.
[0007] In order to achieve the above object, the present invention adopts the following technical solution:
[0008] A first aspect of the present invention provides a method for optimizing the distribution network distribution robustness considering multiple correlations of new energy sources, comprising the following steps:
[0009] Considering the time and space correlation between the output power of new energy sources and the conditional correlation between the predicted value and the prediction error, a correlation sample is generated;
[0010] Based on the correlation sample reduction, typical scenarios and the probability of their occurrence are obtained;
[0011] With the goal of minimizing the active power loss and voltage deviation of the distribution network, a flexible distribution network optimal power flow model is constructed by taking various source-grid-load coordinated control measures as constraints.
[0012] Based on typical scenarios and the probability of their occurrence, the uncertainty set of sources and loads is described, and a two-stage distributed robust optimization model of flexible distribution networks based on data-driven is established in combination with the optimal power flow model of flexible distribution networks.
[0013] The distributed robust optimization model is transformed into a mixed integer second-order cone programming model and the optimal scheduling solution is obtained by solving it.
[0014] Furthermore, the conditional correlation between the time and space between the output power of new energy sources and the predicted value and the predicted error is considered to generate correlation samples, including:
[0015] The spatial correlation coefficient matrix of new energy variables is represented based on the correlation coefficients between time-correlated variables within new energy;
[0016] Based on the variables and transformation coefficients of the set standard normal distribution, the conditional correlation of new energy is obtained;
[0017] The time correlation variables within the new energy are represented by combining the conditional correlation of the new energy and the mean and standard deviation of the time correlation variables within the new energy;
[0018] According to the relationship between the spatial correlation coefficient matrix of new energy variables and the spatial correlation coefficient matrix of standard normal distribution variables, singular value decomposition is performed on the spatial correlation coefficient matrix of standard normal distribution variables;
[0019] Latin hypercube sampling is used to obtain independent standard normal distribution random variables, and then to obtain correlated random variables;
[0020] Based on the solution of random variables, the matrix of new energy variables is obtained, thereby generating correlation samples.
[0021] Furthermore, after generating the correlation samples, the Kmedoids clustering algorithm is used to reduce the correlation samples to obtain typical scenes and the probability of scene occurrence.
[0022] Furthermore, the constraints of the optimal power flow model of the flexible distribution network include: Distflow equation constraints considering network reconstruction, upper and lower limit constraints of substation output, upper and lower limit constraints of new energy output, system safety constraints, hydrogen energy storage model constraints, flexible multi-state switch model constraints with multi-terminal hydrogen energy storage, static VAR compensator model constraints, capacitor bank model constraints, flexible load model and network reconstruction model constraints.
[0023] Furthermore, the hydrogen energy storage model is constrained as follows:
[0024] τ EH =Δtδ EH / k ved ,
[0025] τ HE =k ved δ HE / Δt,
[0026]
[0027] In the formula, δ EH , δ HE are the energy conversion efficiencies of the electrolysis device and the fuel cell respectively; Δt is the unit operating time, k ved is the volume energy density of hydrogen in the hydrogen storage tank; are the hydrogen storage capacity of the hydrogen storage tank at time t+1 and time t respectively; are the hydrogen production power and hydrogen combustion power generation power at time t respectively; τ EH , τ HE They are the efficiency of hydrogen production by water electrolysis and electricity generation by hydrogen combustion, respectively.
[0028] The constraints of the flexible multi-state switch model with multi-terminal hydrogen energy storage are:
[0029]
[0030]
[0031]
[0032] In the formula, ψ p It is the port number set for the three-terminal FMSS to access the distribution network; are the active power and reactive power injected by FMSS at node i at time t, and are the upper and lower limits of reactive power injected by FMSS at node i at time t; S FMSS is the apparent power of FMSS.
[0033] Static VAR compensator model constraints:
[0034]
[0035] In the formula, and are the upper and lower limits of reactive power output of SVC at node i and the actual reactive power output at time t respectively;
[0036] Capacitor Bank Model Constraints:
[0037]
[0038] In the formula, and are the total capacity and the capacity of a single group of CBs actually put into use at node i at time t; is the number of CB groups actually put into operation at node i at time t, is the number of CB groups actually put into operation at node i at time t+1; t is the working status identifier at time t; The maximum number of actions in a day.
[0039] Flexible load model constraints:
[0040]
[0041] In the formula, They represent the active power increased and decreased by FL on node i at time t respectively; and They represent the active and reactive power presented to the outside by the FL on node i at time t; β FL,up , β FL,down are the proportions of active power increased and decreased by FL at time t to the predicted load; It is a 0-1 variable, which serves as an identifier for the added power; are the predicted active and reactive power at time t respectively; β FLis the proportion of the power generated by FL at time t to the predicted load.
[0042] Network reconstruction model constraints:
[0043]
[0044]
[0045]
[0046]
[0047]
[0048]
[0049] In the formula, cj and dj are the child node set and parent node set of node j respectively, and j does not contain the root node; c1 is the root node set, F jk,t is the virtual power flowing from node j to node k at time t, F ij,t is the virtual power flowing from node i to node j at time t, F j,t is the virtual load of node j; F 1k,t is the virtual power flowing from node 1 to node k at time t; v is an unrestricted real number, λ v is an infinite number; is a binary variable used to represent the on / off status of line ij at time t. It takes 1 when the line is connected and 0 when it is disconnected. The upper limit of the number of times a single contact switch operates during the operation cycle. is the upper limit of the number of times all contact switches operate in a cycle, Ω b ,Ω L They are node and branch sets respectively.
[0050] Furthermore, based on typical scenarios and the probability of their occurrence, the probability allowable deviation limit under the corresponding constraints of 1-norm and ∞-norm is used to describe the source-load uncertainty set, so that the probability of occurrence of the constructed typical scenarios matches the probability of the real scenarios.
[0051] Furthermore, when solving the distributed robust optimization model, the C&CG algorithm is used to decompose the distributed robust optimization model into a main problem and sub-problems for cyclic iteration, first solving the inner minimum problem of the sub-problem, and then solving the outer maximum problem.
[0052] A second aspect of the present invention provides a distribution network distributed robust optimization system considering multiple correlations of new energy sources, comprising:
[0053] A correlation sample generation module is used to generate correlation samples by taking into account the time and space between the output power of new energy sources and the conditional correlation between the predicted value and the prediction error; based on the correlation sample reduction, typical scenarios and the probability of scenario occurrence are obtained;
[0054] The optimal power flow model building module is used to build the optimal power flow model of the flexible distribution network with the goal of minimizing the active power loss and voltage deviation of the distribution network and taking various source-grid-load coordinated control measures as constraints;
[0055] The distributed robust optimization model building module is used to describe the source-load uncertainty set based on typical scenarios and the probability of scenario occurrence, and to establish a two-stage distributed robust optimization model for flexible distribution networks based on data-driven combined with the optimal power flow model of flexible distribution networks;
[0056] The optimization scheduling module is used to transform the distributed robust optimization model into a mixed integer second-order cone programming model and solve it to obtain the optimal scheduling solution.
[0057] A third aspect of the present invention provides a computer-readable storage medium.
[0058] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in the above-mentioned robust optimization method for distribution network considering multiple correlations of new energy sources.
[0059] A fourth aspect of the present invention provides a computer device.
[0060] A computer device comprises 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 steps in the above-mentioned robust optimization method for distribution network considering multiple correlations of new energy sources are implemented.
[0061] Compared with the prior art, the present invention has the following beneficial effects:
[0062] 1. Compared with the traditional method, the present invention takes into account the time and space between the output power of new energy and the conditional correlation between the predicted value and the prediction error, generates correlation samples, and makes the optimization strategy more accurate and comprehensive.
[0063] 2. By introducing the distributed rod optimization strategy, the present invention can effectively deal with the uneven temporal and spatial distribution of renewable energy output and the complexity of regulating massive devices, thereby improving the stability and reliability of the distribution network.
[0064] 3. The present invention adopts a data-driven optimization model to make the system more adaptable and flexible, and better able to cope with uncertainty and changes.
[0065] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0066] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0067] Figure 1 It is a flow chart of a method for optimizing the distribution network distribution based on multiple correlations of new energy sources provided by an embodiment of the present invention;
[0068] Figure 2 Schematic diagram of H-FMSS access to a distribution network provided by an embodiment of the present invention;
[0069] Figure 3 It is a diagram of an improved IEEE33 node test system provided by an embodiment of the present invention;
[0070] Figure 4 is a source-load prediction output curve diagram provided by an embodiment of the present invention;
[0071] Figure 5 is a comparison diagram of optimization results when the β / α value is changed according to an embodiment of the present invention;
[0072] Figure 6 It is a reactive power output diagram of CB under three schemes provided in the embodiments of the present invention. DETAILED DESCRIPTION
[0073] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0074] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0075] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0076] Embodiment 1
[0077] like Figure 1As shown, this embodiment provides a robust optimization method for distribution network considering multiple correlations of new energy sources, including the following steps:
[0078] Step 1: Consider the time and space correlation between the output power of renewable energy and the conditional correlation between the predicted value and the prediction error, and generate correlation samples;
[0079] In this embodiment, the multiple correlation of new energy refers to the conditional correlation between the time and space of the output power of new energy and the predicted value and the predicted error. At the same time, the distance of the geographical location of the installation of new energy will also have a certain impact on each other's output, that is, there is spatial correlation at the same time. There is also a certain conditional correlation between the predicted value and the predicted error of the new energy.
[0080] Assume that the new energy variable composition M=(M1,M2,...,M i ,...,M z ) T Matrix, where z is the number of new energy sources. i It is a T×T dimensional matrix, which is used to represent the time correlation within the i-th new energy source.
[0081] The spatial correlation coefficient matrix C of M M It can be expressed as:
[0082]
[0083] In the formula, ρ Mi,Mj Indicates M i and M j The correlation coefficient between i and M j The size of the correlation, then ρ Mi,Mj =ρ Mj,Mi .
[0084] Set the standard normal distribution variable N=(N1,N2,...,N i ,...,N z ) T , the transformation coefficient is C i,0 , C i,1 , C i,2 and C i,3 According to the polynomial normal transformation method, M i Indicated as N i A third-order polynomial of :
[0085]
[0086] In the formula, μ Mi and σ Mi M iThe conditional relevance of new energy is reflected in the selection of standard deviation; M si M i standardized form.
[0087] Due to C M =C N Because C M is a symmetric matrix, then we can N Perform singular value decomposition, that is:
[0088]
[0089] Where U N is an orthogonal matrix; C N0 C N The diagonal matrix is composed of the singular values of , and the diagonal elements are arranged in descending order of numerical value.
[0090] Latin hypercube sampling is used to obtain independent standard normal distribution random variables L, and then a correlated random variable N is obtained, that is:
[0091]
[0092] Substituting N into formula (2) can obtain M, thereby generating correlation samples.
[0093] Compared with traditional methods, the present invention takes into account the time and space between the output power of new energy and the conditional correlation between the predicted value and the prediction error, generates correlation samples, and makes the optimization strategy more accurate and comprehensive.
[0094] Step 2: Reduce the relevant samples to obtain typical scenarios;
[0095] Since there are too many samples generated, the dimension of calculation is too large and the calculation efficiency is slow. Therefore, the Kmedoids clustering algorithm is used to reduce a large number of related samples to obtain the desired typical scenes and the probability of scene occurrence.
[0096] Step 3: Construct the optimal power flow model of flexible distribution network;
[0097] Among them, the objective function is set to minimize the network loss value and voltage deviation of the entire network.
[0098] minf(5),
[0099]
[0100] In order to take into account both the economy and safety of FDN, f is set as a multi-objective function, which consists of the network loss value f1 and the voltage deviation f2; α and β are the weight coefficients of each objective respectively; μ is the voltage deviation coefficient; r ijis the resistance of branch ij; I ij,t is the current amplitude of branch ij at time t; U N is the rated voltage of the system.
[0101] The constraints include:
[0102] (1) The Distflow equation constraints considering network reconstruction are as follows:
[0103]
[0104]
[0105]
[0106] Where P ij,t , P jk,t , Q ij,t , Q jk,t are the active power and reactive power transmitted by line ij and line jk at time t respectively; P j,t , Q j,t is the net injected active and reactive power of node j at time t; r ij 、x ij are the resistance and reactance of line ij respectively; U i,t , U j,t are the voltages of nodes i and j at time t respectively. They are the active / reactive power injected by the predicted load of node j at time t, the active / reactive power injected by FL, the active / reactive power injected by renewable energy generation, the active / reactive power injected by the substation, and the active / reactive power of the soft switch; Inject active power into the hydrogen energy storage at node j at time t; They are the injected reactive power of the static VAR compensation device and the capacitor bank respectively.
[0107] (2) Substation output upper and lower limit constraints:
[0108]
[0109] In the formula, and and They respectively represent the upper and lower limits of the active power and reactive power injected into the power grid by the substation.
[0110] (3) Constraints on upper and lower limits of new energy output:
[0111]
[0112] In the formula, Indicates the predicted active power output of new energy; the new energy type adopts PQ type, in this embodiment, China, the power factor adopts cosθ=0.90. The reactive power is calculated by the power factor.
[0113] (4) System security constraints:
[0114]
[0115] Where U i,max and U i,min are the upper and lower limits of the voltage at node i respectively; I ij,max is the upper limit of the current of line ij; P ij,max , P ij,min and Q ij,max , Q ij,min They are the upper and lower limits of active and reactive power flowing through line ij respectively.
[0116] (5) Hydrogen energy storage system (HESS) model:
[0117] In order to promote the sharing of hydrogen energy within the region, this embodiment proposes the concept of "hydrogen storage routing" for the first time, that is, encouraging cross-time and cross-space energy regulation between HESSs.
[0118] τ EH =Δtδ EH / k ved (13)
[0119] τ HE =k ved δ HE / Δt(14),
[0120]
[0121] In the formula, δ EH , δ HE are the energy conversion efficiencies of the electrolysis device and the fuel cell, respectively; Δt is the unit operating time, which is 1h in this embodiment; k ved is the volume energy density of hydrogen in the hydrogen storage tank. are the hydrogen storage capacity of the hydrogen storage tank at time t+1 and time t respectively; are the hydrogen production power and hydrogen combustion power generation power at time t respectively; τ EH , τ HE They are the efficiency of hydrogen production by water electrolysis and electricity generation by hydrogen combustion, respectively.
[0122] (6) H-FMSS model:
[0123] This embodiment adopts a three-terminal FMSS and connects the DC end to the HESS to form a multi-terminal FMSS with HESS (FMSS with HESS, H-FMSS) to provide power support for the distribution network. Figure 1 As shown, the mathematical model can be expressed as:
[0124]
[0125]
[0126]
[0127] In the formula, ψ p It is the port number set for the three-terminal FMSS to access the distribution network; are the active power and reactive power injected by FMSS at node i at time t; S FMSS is the apparent power of FMSS.
[0128] (7) Static var compensator (SVC) model:
[0129]
[0130] In the formula, and They are the upper and lower limits of the reactive power output of the SVC at node i and the actual reactive power output at time t.
[0131] (8) Capacitor banks (CB) model:
[0132]
[0133] In the formula, and are the total capacity and the capacity of a single group of CBs actually put into use at node i at time t; is the number of CB groups actually put into operation at node i at time t; t is the working status identifier at time t; The maximum number of actions in a day.
[0134] (9) Flexible loads (FL) model:
[0135] The generalized FL includes transferable loads, translatable loads and reducible loads. The constructed FL model is as follows:
[0136]
[0137] In the formula, They represent the active power increased and decreased by FL on node i at time t respectively; and They represent the active and reactive power presented to the outside by the FL on node i at time t. FL,up , β FL,down are the proportions of active power increased and decreased by FL at time t to the predicted load; It is a 0-1 variable, which serves as an identifier for the added power; are the predicted active and reactive power at time t respectively; β FL is the proportion of the power generated by FL at time t to the predicted load.
[0138] (10) Network Reconstruction:
[0139] Assume that node 1 in the distribution network is the root node, and the system contains only one root node. Construct a network reconstruction model:
[0140]
[0141]
[0142]
[0143]
[0144]
[0145]
[0146] In the formula, cj and dj are the child node set and parent node set of node j respectively, and j does not contain the root node; c1 is the root node set, F jk,t is the virtual power flowing from node j to node k at time t, F ij,t is the virtual power flowing from node i to node j at time t, F j,t is the virtual load of node j; F 1k,t is the virtual power flowing from node 1 to node k at time t; v is an unrestricted real number, λ v is an infinite number; is a binary variable used to represent the on / off status of line ij at time t. It takes 1 when the line is connected and 0 when it is disconnected. The upper limit of the number of times a single contact switch operates during the operation cycle. is the upper limit of the number of times all contact switches operate in a cycle, Ω b ,Ω L They are node and branch sets respectively.
[0147] Step 4: Construct a data-driven FDN distribution robust optimization model;
[0148] When multiple scenarios are used to describe the uncertainty of source load, the above deterministic models (1) to (27) are expressed in the following uncertainty form.
[0149]
[0150]
[0151]
[0152]
[0153] ||Qy s +P||2≤q T y s +p (32),
[0154] In the formula, p s is the probability of scene s appearing; D s is the total number of scenarios. A, C, D, E, F, G, H, K, and M are all constant coefficient matrices used in the robust optimization model; Q and P represent the reactive power and active power respectively; c, d, x, e, u, p, and q are the maximum values of all feasible domains, and A T y s is the objective function value for each scenario.
[0155] Apply step 1 to generate C correlation samples, i.e., actual scenarios. In order to ensure that the probability of occurrence of the constructed typical scenario matches the probability of the actual scenario when the number of actual scenarios is too large, a confidence set under different values of the probability allowable deviation limit θ is established, as shown in formula (33):
[0156]
[0157] Where, θ1 and θ∞ are the probability allowable deviation limits under the corresponding constraints of 1-norm and ∞-norm respectively; Ω p For p s Feasible domain, is the initial probability distribution of the scene.
[0158] The scenario probability satisfies the following confidence constraints:
[0159]
[0160] Equation (34) equals the right side of the inequality to a1 and a ∞ , and we get formula (35).
[0161]
[0162] It can be seen from the above formula that as the number of relevant samples C increases, θ becomes smaller, that is, the confidence set at the same confidence level decreases as the number of actual scenarios mastered increases.
[0163] The absolute value constraints of the 1-norm and ∞-norm in formula (33) cannot be calculated directly and need to be transformed. The 1-norm constraint after transformation is as follows:
[0164]
[0165]
[0166] Where m is an auxiliary variable.
[0167] After transformation, the ∞-norm constraint is as follows:
[0168]
[0169] The uncertainty of “source load” is handled by using the methods of equations (33) to (35), and then a data-driven distributed robust optimization model is constructed as shown below:
[0170]
[0171] Step 5: Convert the distributed robust optimization model into a mixed integer second-order cone programming model and solve it to obtain the optimal scheduling solution, which specifically includes the following steps:
[0172] Step 501: converting the distributed robust optimization model into a mixed integer second-order cone programming model;
[0173] The above distributed robust optimization model contains many nonlinear and non-convex terms, which makes it extremely difficult to solve the model and it is impossible to determine whether the solution is convergent. Therefore, the second-order cone relaxation is used to transform the above model into a mixed integer second-order cone programming model, so that the solver can be called for solution.
[0174] Define a new variable l ij,t 、V i,t Replace the voltage and current square terms in the above model, as shown in equation (45).
[0175]
[0176] After the second-order cone transformation, the corresponding constraints become:
[0177]
[0178] Perform second-order cone relaxation on equation (47), and the constraints after relaxation are as follows:
[0179]
[0180] Step 502: Distributed robust solution method For the proposed distributed robust model, the C&CG algorithm is used to decompose equation (42) into a master problem (MP) and a sub-problem (SP) for cyclic iteration until convergence.
[0181]
[0182] Where u is the number of iterations.
[0183] Since the probability of occurrence of each typical scenario in the sub-problem and the variable y in the second stage are independent of each other, the inner minimum problem of the sub-problem can be solved first, and then the outer maximum problem can be solved. The present invention introduces a distributed robust optimization strategy to deal with uncertainty factors by searching for the worst probability distribution under uncertain conditions, and solves the model to obtain the optimal output strategy of the equipment under the premise of meeting the system operation constraints, including substations, new energy, hydrogen energy storage, multi-terminal flexible multi-state switches with hydrogen energy storage, static VAR compensator model constraints, capacitor banks and other working states.
[0184] The present invention can effectively deal with the uneven temporal and spatial distribution of renewable energy output and the complexity of mass device regulation, and improve the stability and reliability of the distribution network. The use of a data-driven optimization model makes the system more adaptable and flexible, and can better deal with uncertainty and changes.
[0185] This embodiment provides a flexible distribution network distributed robust optimization strategy considering multiple correlations of new energy sources, and establishes a H-FMSS access distribution network model, such as Figure 2 The focus of the present invention is to verify the effectiveness of the proposed robust optimization of the new energy distribution network. The method proposed in this embodiment adopts an improved IEEE33 node test system, such as Figure 3 The predicted output curve of new energy and load is shown in Figure 4 shown.
[0186] In order to verify the necessity and effectiveness of considering the multiple correlations of new energy, the present invention sets up the following three different new energy processing schemes and performs optimization analysis respectively:
[0187] Option 1: Only consider the independent randomness of new energy;
[0188] Option 2: Only consider the spatiotemporal correlation of new energy sources;
[0189] Solution 3: Considering the multiple correlations of new energy sources (the method proposed in the present invention).
[0190] This embodiment generates 100 scenarios according to the process proposed in Section 3. The original scenarios are clustered using the SBR method, and finally 5 typical scenarios are obtained. For simplicity, this invention only discusses the photovoltaic output scenario, and the wind power output scenario is exactly the same. Figure 5 As shown in Table 1, the overall trend of the simplified scenarios is similar to that of the original scenario set. The probability of each scenario after reduction is shown in Table 1.
[0191] Table 1 Comparison of optimization results under three schemes
[0192]
[0193]
[0194] As can be seen from Table 1, the network loss value and voltage offset under the processing method proposed in the present invention are the smallest, which proves that the proposed processing method can more accurately track the random output of new energy sources and better coordinate various fast and slow control equipment, thereby reducing system network losses and improving voltage quality.
[0195] The above three processing schemes generate 5000 scenarios in each operation cycle and cluster them into 4 typical scenarios and their probability distributions. α1 and α∞ are both 0.8.
[0196] observe Figure 6 It can be seen that from Scheme 1 to Scheme 3, the number of CB switching groups is 3, 2 and 1 respectively, which is decreasing successively. However, the average value of CB output reactive power is increasing successively, which is 337kVar, 387kVar and 437kVar respectively. Therefore, it can be inferred that Scheme 3, i.e. the method proposed by the present invention, can significantly reduce the number of CB switching groups and make full use of CB.
[0197] In order to highlight the superiority of the distributed robust optimization method used in the present invention, a simulation experiment is conducted to compare with the traditional robust optimization method and the random programming method. 5000 historical scenarios are generated. The robust optimization method takes a fluctuation range of ±10%. The random programming rule makes the probability of the four typical scenarios obtained by the distributed robust model all 0.25. The comparison of the optimization results is shown in Table 2.
[0198] Table 2 Comparison of optimization results under different methods
[0199]
[0200] As can be seen from Table 2, the system network loss value and voltage offset obtained by robust optimization are the largest. The result obtained by random programming is the smallest, but the conservatism is not guaranteed. Compared with robust optimization and random programming, the distributed robustness used in the present invention achieves a better balance in economy, safety and conservatism, while minimizing system network loss and improving voltage quality, and has significant advantages in dealing with uncertainty planning.
[0201] Embodiment 2
[0202] This embodiment provides a distribution network distributed robust optimization system considering multiple correlations of new energy sources, including:
[0203] A correlation sample generation module is used to generate correlation samples by taking into account the time and space between the output power of new energy sources and the conditional correlation between the predicted value and the prediction error; based on the correlation sample reduction, typical scenarios and the probability of scenario occurrence are obtained;
[0204] The optimal power flow model building module is used to build the optimal power flow model of the flexible distribution network with the goal of minimizing the active power loss and voltage deviation of the distribution network and taking various source-grid-load coordinated control measures as constraints;
[0205] The distributed robust optimization model building module is used to describe the source-load uncertainty set based on typical scenarios and the probability of scenario occurrence, and to establish a two-stage distributed robust optimization model for flexible distribution networks based on data-driven combined with the optimal power flow model of flexible distribution networks;
[0206] The optimization scheduling module is used to transform the distributed robust optimization model into a mixed integer second-order cone programming model and solve it to obtain the optimal scheduling solution.
[0207] Embodiment 3
[0208] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps in the distribution robust optimization method for distribution networks considering multiple correlations of new energy sources as described above are implemented.
[0209] Embodiment 4
[0210] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the method for robust optimization of distribution network considering multiple correlations of new energy sources as described above are implemented.
[0211] Embodiment 5
[0212] This embodiment provides a program product, which is a computer program product, including a computer program, characterized in that when the computer program is executed by a processor, the steps in the distribution robust optimization method of the distribution network considering the multiple correlations of new energy as described above are implemented.
[0213] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.
[0214] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0215] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0216] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0217] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.
[0218] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A robust optimization method for distribution network considering multiple correlations of renewable energy, characterized in that: include: Considering the time and space correlation between the output power of new energy sources and the conditional correlation between the predicted value and the prediction error, a correlation sample is generated; Based on the correlation sample reduction, typical scenarios and the probability of their occurrence are obtained; With the goal of minimizing the active power loss and voltage deviation of the distribution network, a flexible distribution network optimal power flow model is constructed by taking various source-grid-load coordinated control measures as constraints. Based on typical scenarios and the probability of their occurrence, the uncertainty set of sources and loads is described, and a two-stage distributed robust optimization model of flexible distribution networks based on data-driven is established in combination with the optimal power flow model of flexible distribution networks. The distributed robust optimization model is transformed into a mixed integer second-order cone programming model and the optimal scheduling solution is obtained by solving it.
2. The method for optimizing the distribution network distribution based on multiple correlations of new energy sources according to claim 1, characterized in that: The method of considering the time and space between the output power of new energy sources and the conditional correlation between the predicted value and the predicted error to generate the correlation sample includes: The spatial correlation coefficient matrix of new energy variables is represented based on the correlation coefficients between time-correlated variables within new energy; Based on the variables and transformation coefficients of the set standard normal distribution, the conditional correlation of new energy is obtained; The time correlation variables within the new energy are represented by combining the conditional correlation of the new energy and the mean and standard deviation of the time correlation variables within the new energy; According to the relationship between the spatial correlation coefficient matrix of new energy variables and the spatial correlation coefficient matrix of standard normal distribution variables, singular value decomposition is performed on the spatial correlation coefficient matrix of standard normal distribution variables; Latin hypercube sampling is used to obtain independent standard normal distribution random variables, and then to obtain correlated random variables; Based on the solution of random variables, the matrix of new energy variables is obtained, thereby generating correlation samples.
3. The method for optimizing the distribution network distribution considering multiple correlations of new energy sources according to claim 1, characterized in that: After generating relevant samples, the Kmedoids clustering algorithm is used to reduce the relevant samples to obtain typical scenes and the probability of scene occurrence.
4. The method for optimizing the distribution network distribution considering multiple correlations of new energy sources according to claim 1, characterized in that: The constraints of the optimal power flow model of the flexible distribution network include: Distflow equation constraints considering network reconstruction, upper and lower limit constraints of substation output, upper and lower limit constraints of new energy output, system safety constraints, hydrogen energy storage model constraints, flexible multi-state switch model constraints with multi-terminal hydrogen energy storage, static VAR compensator model constraints, capacitor bank model constraints, flexible load model and network reconstruction model constraints.
5. The method for optimizing the distribution network distribution considering multiple correlations of new energy sources as claimed in claim 4, characterized in that: The constraints of the hydrogen energy storage model are: t EH =Δtδ EH / k ved , t HE =k ved d HE / Δt, In the formula, δ EH , δ HE are the energy conversion efficiencies of the electrolysis device and the fuel cell respectively; Δt is the unit operating time, k ved is the volume energy density of hydrogen in the hydrogen storage tank; are the hydrogen storage capacity of the hydrogen storage tank at time t+1 and time t respectively; are the hydrogen production power and hydrogen combustion power generation power at time t respectively; τ EH , τ HE They are the efficiency of hydrogen production by water electrolysis and electricity generation by hydrogen combustion; The constraints of the flexible multi-state switch model with multi-terminal hydrogen energy storage are: In the formula, ψ p It is the port number set for the three-terminal FMSS to access the distribution network; are the active power and reactive power injected by FMSS at node i at time t, and are the upper and lower limits of reactive power injected by FMSS at node i at time t; S FMSS is the apparent power of FMSS; Static VAR compensator model constraints: In the formula, and are the upper and lower limits of reactive power output of SVC at node i and the actual reactive power output at time t respectively; Capacitor Bank Model Constraints: In the formula, and are the total capacity and the capacity of a single group of CBs actually put into use at node i at time t; is the number of CB groups actually put into operation at node i at time t, is the number of CB groups actually put into operation at node i at time t+1, O t is the working status identifier at time t; The maximum number of actions in a day is limited; Flexible load model constraints: In the formula, They represent the active power increased and decreased by FL on node i at time t respectively; and They represent the active and reactive power presented to the outside by the FL on node i at time t; β FL,up , β FL,down are the proportions of active power increased and decreased by FL at time t to the predicted load; It is a 0-1 variable, which serves as an identifier for the added power; are the predicted active and reactive power at time t respectively; β FL is the proportion of the power generated by FL at time t to the predicted load; Network reconstruction model constraints: In the formula, cj and dj are the child node set and parent node set of node j respectively, and j does not contain the root node; c1 is the root node set, F jk,t is the virtual power flowing from node j to node k at time t, F ij,t is the virtual power flowing from node i to node j at time t, F j,t is the virtual load of node j; F 1k,t is the virtual power flowing from node 1 to node k at time t; v is an unrestricted real number, λ v is an infinite number; is a binary variable used to represent the on / off status of line ij at time t. It takes 1 when the line is connected and 0 when it is disconnected. The upper limit of the number of times a single contact switch operates during the operation cycle. is the upper limit of the number of times all contact switches operate in a cycle, Ω b ,Ω L They are node and branch sets respectively.
6. The method for optimizing the distribution network distribution considering multiple correlations of new energy sources according to claim 1, characterized in that: Based on typical scenarios and the probability of scenario occurrence, the probability allowable deviation limit under the corresponding constraints of 1-norm and ∞-norm is used to describe the source-load uncertainty set, so that the probability of occurrence of the constructed typical scenario matches the probability of the real scenario.
7. The method for optimizing the distribution network distribution considering multiple correlations of new energy sources according to claim 1, characterized in that: When solving the distributed robust optimization model, the C&CG algorithm is used to decompose the distributed robust optimization model into a main problem and sub-problems for cyclic iteration. The inner minimum problem of the sub-problem is solved first, and then the outer maximum problem is solved.
8. A distributed robust optimization system for distribution networks considering multiple correlations of renewable energy, characterized in that: include: A correlation sample generation module is used to generate correlation samples by taking into account the time and space correlation between the output power of new energy sources and the conditional correlation between the predicted value and the predicted error; Based on the correlation sample reduction, typical scenarios and the probability of their occurrence are obtained; The optimal power flow model building module is used to build the optimal power flow model of the flexible distribution network with the goal of minimizing the active power loss and voltage deviation of the distribution network and taking various source-grid-load coordinated control measures as constraints; The distributed robust optimization model building module is used to describe the source-load uncertainty set based on typical scenarios and the probability of scenario occurrence, and to establish a two-stage distributed robust optimization model for flexible distribution networks based on data-driven combined with the optimal power flow model of flexible distribution networks; The optimization scheduling module is used to transform the distributed robust optimization model into a mixed integer second-order cone programming model and solve it to obtain the optimal scheduling solution.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the robust optimization method for distribution network considering multiple correlations of new energy sources as described in any one of claims 1 to 7 are implemented.
10. A computer 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 steps in the robust optimization method for distribution network considering multiple correlations of new energy sources as described in any one of claims 1-7 are implemented.
Citation Information
Cited By
Power distribution network reconstruction method, system and device, storage medium and program product
CN120237658A
