Energy storage configuration optimization method and system for flexible interconnection power distribution network

Abstract

The invention provides an energy storage configuration optimization method and system for a flexible interconnection power distribution network, and belongs to the technical field of energy storage configuration of the power distribution network, and the method comprises the steps: building a dynamic cost accounting mechanism which is coupled with a network physical state in real time; constructing a short-term two-stage robust optimization model which can fully schedule various flexible resources and takes minimization of the total cost of the system as a target based on a dynamic cost accounting mechanism, and obtaining an energy storage optimal operation strategy; a long-term two-stage robust optimization model is established, the objective of the main problem is to minimize the upper bound of the sum of investment cost and worst-case operation cost, and the sub problem is to evaluate the worst performance of the investment scheme in all possible uncertainty scenes under a given investment scheme; and a robust optimal configuration scheme is obtained through iterative solution of the main problem and the sub-problems.

Description

Technical Field

[0001] This invention belongs to the field of energy storage configuration technology for power distribution networks, and particularly relates to an energy storage configuration optimization method and system for flexible interconnected power distribution networks. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] With the construction of new power systems, the penetration rate of renewable energy sources such as distributed photovoltaics in distribution networks has increased dramatically. The strong fluctuations and uncertainties in their output pose a severe challenge to the safe and stable operation of the power grid. At the same time, new market models such as peer-to-peer (P2P) energy trading have emerged, transforming energy flow from unidirectional transmission to multidirectional interaction.

[0004] During operation, firstly, current P2P energy trading lacks deep coupling with the physical characteristics of the distribution network. Traditional fixed-rate or simplified linear rotation cost models cannot accurately reflect the impact of SOP access and network topology changes on transaction path costs, leading to distorted market signals and failing to guide the formation of a power flow distribution conducive to grid safety and economy. Secondly, at the operational optimization level, network reconfiguration, SOP adjustment, and energy storage dispatch are often considered separately, and most models use deterministic optimization or simple scenario analysis, making it difficult to effectively cope with the strong uncertainties of renewable energy and loads, resulting in insufficient system robustness. Finally, energy storage planning and operation scheduling are severely disconnected. Long-term configuration decisions are mostly based on typical days, failing to incorporate operational uncertainties and the synergistic value of multiple resources into the assessment. This leads to investment plans that are either too conservative and fail to fully tap potential, or too aggressive and fail to guarantee feasibility under extreme scenarios, resulting in poor economic performance throughout the entire life cycle. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, this invention provides a method and system for optimizing energy storage configuration in flexible interconnected distribution networks, enabling coordinated planning of energy storage and network flexibility resources.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: Firstly, a method for optimizing energy storage configuration for flexible interconnected distribution networks is disclosed, including: Establish a dynamic cost accounting mechanism that is coupled in real time with the physical state of the network; Based on the dynamic cost accounting mechanism, a short-term two-stage robust optimization model is constructed that can fully allocate various flexible resources and aims to minimize the total system cost, so as to obtain the optimal operation strategy for energy storage. A long-term two-stage robust optimization model is established. The main problem is to minimize the upper bound of the sum of investment cost and worst-case operating cost. The sub-problem is to evaluate its worst performance in all possible uncertainty scenarios under a given investment plan. By iteratively solving the main problem and its subproblems, a robust optimal configuration scheme is obtained.

[0007] This solution ensures economic efficiency in investment while guaranteeing that the system has a feasible operation scheduling strategy under any permissible uncertainty scenario in the future, and that its total cost is controllable and approximately optimal.

[0008] As a further technical solution, when establishing a dynamic cost accounting mechanism that is coupled in real time with the network physical state, the specific components include: A transaction and dynamic rotation cost model is established. Based on the model, the transaction path is decomposed into a network to obtain two physical components of independent decisions. The virtual network decomposition method is used to quantify the impact of these two physical components on the network. After quantifying the impact of these two physical components on the network, the dynamic rotation cost is calculated based on the power transmission distribution factor and electrical distance.

[0009] As a further technical solution, the two independent physical components are: the power component flowing through the non-SOP path, whose power flow is carried only by the traditional network, and the power component flowing through the SOP path, whose power flow must pass through the SOP device.

[0010] As a further technical solution, a virtual network decomposition method is used to quantify the impact of these two physical components on the network, specifically including: For those with specific SOPs network Construct two virtual, purely radial subnetworks. and ; Subnetwork :exist Disconnect SOP branch , that is to say The network uses the original main network connection point as the balancing node. In this network, the power components... Treated as a node Injection and nodes The outflow of energy follows the physical laws of this radial network. Subnetwork :exist Retain SOP branch However, the SOP device itself is regarded as a new virtual balance node. All nodes in the original network are considered PQ nodes in this subnet, and the two end nodes of the SOP are... and Through this virtual balance node Achieve power exchange.

[0011] As a further technical solution, dynamic rotation costs are calculated based on power transmission distribution factor and electrical distance, specifically including: For any line in a subnetwork, its electrical distance is defined as the line distance. The ratio of the resistance to the system reference resistance; The power transfer distribution factor matrix is ​​used to describe the linear sensitivity relationship between power transfer between nodes and branch power flow in a network.

[0012] As a further technical solution, the short-term two-stage robust optimization model specifically includes: The objective of robust operation optimization is to select the first-stage decision so that the total expected operating cost caused by the second-stage adaptive decision is minimized when facing the worst uncertainty scenario. The multi-resource collaborative constraint system includes: establishing a refined model of energy storage operation and lifetime loss based on unified modeling of network topology flexibility; linearization and convex relaxation of the DistFlow model considering uncertainties.

[0013] Secondly, a system for optimizing the configuration of energy storage for flexible interconnected distribution networks is disclosed, including: The dynamic cost accounting mechanism establishment module is configured to: establish a dynamic cost accounting mechanism that is coupled in real time with the network physical state; The short-term two-stage robust optimization module is configured to: construct a short-term two-stage robust optimization model based on a dynamic cost accounting mechanism that can fully schedule various flexible resources and aims to minimize the total system cost, thereby obtaining the optimal energy storage operation strategy; The long-term two-stage robust optimization module is configured to: establish a long-term two-stage robust optimization model, the main problem being to minimize the upper bound of the sum of investment cost and worst-case operating cost, and the sub-problem being to evaluate its worst performance in all possible uncertainty scenarios under a given investment plan. By iteratively solving the main problem and its subproblems, a robust optimal configuration scheme is obtained.

[0014] The above one or more technical solutions have the following beneficial effects: The technical solution of this invention constructs an optimization framework that deeply integrates the physical characteristics of the distribution network with the market transaction mechanism. This framework should be able to accurately quantify the impact of P2P transactions on the network and generate dynamic cost signals. Based on this, it collaboratively optimizes flexible resources across multiple time scales, such as SOPs, network reconfiguration, and energy storage, to address source-load uncertainties. Ultimately, it embeds a robust operation model into energy storage planning decisions to achieve collaborative configuration based on optimal life-cycle costs.

[0015] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0016] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0017] Figure 1 This is a network decomposition diagram of the P2P transaction path according to an embodiment of the present invention; Figure 2 This is a general framework diagram of an embodiment of the present invention. Detailed Implementation

[0018] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0020] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0021] Terminology Explanation: P2P: Peer-to-Peer Energy Trading. P2P energy trading is a decentralized electricity market model that allows electricity users to trade electricity directly without the need for traditional power companies as intermediaries. Transactions are based on agreements between the parties or platform matching, and are typically implemented using digital technologies such as blockchain and smart contracts.

[0022] SOP: Also known as Soft Open Point, it is a new type of intelligent power electronic device installed at the location of traditional tie switches. It belongs to the category of intelligent soft switches and can replace traditional tie switches to achieve flexible closed-loop operation of the distribution network.

[0023] Example 1 See appendix Figure 2 As shown in the figure, this embodiment discloses an energy storage configuration optimization method for flexible interconnected distribution networks, including: Step 1: P2P transaction path decomposition and dynamic modeling of rotation fees.

[0024] In a flexible, interconnected, and active distribution network, peer-to-peer (P2P) energy trading must rely on the physical network. Traditional fixed network access fees or simple linear rates cannot reflect the diversity of trading paths and the time-varying nature of network conditions. Inaccurate cost signals can lead to irrational trading flows, potentially causing line overload, increased losses, and even negating the flexibility benefits of standard operating procedures (SOPs). Therefore, establishing a dynamic cost accounting mechanism that is coupled in real-time with the network's physical state is a primary prerequisite for guiding the healthy operation of the market and achieving optimal resource allocation.

[0025] (1-1) Mathematical formalization of the transaction basic model and network decomposition: The transaction basic model receives the original transaction data and network state, and decomposes the transaction power into components of different paths.

[0026] Consider a distribution network, whose graphical structure is represented as follows: ,in For a set of nodes, It is a system set, including a set of traditional lines. Connecting soft open points Suppose there exists a smart soft switch SOP connected to node. and Between, that is .

[0027] For any trading pair During the period Trading power Let i represent the network node where the seller (energy injector) is located, and j represent the network node where the buyer (energy absorber) is located. This can be decomposed into two independent physical components: (1) in, This indicates the power component flowing through a non-SOP path, whose power flow is carried only by the traditional network; This indicates the power component flowing through the SOP path, and its power flow must pass through the SOP device.

[0028] In a flexible interconnected distribution network, the electrical energy in a P2P transaction can be transmitted from the seller to the buyer through different physical paths. Introducing a Standard Operating Procedure (SOP) creates a controllable and rapid power exchange path parallel to the traditional radial network. Without decomposition, issues such as the occupancy of SOP equipment and the additional power flow on traditional lines cannot be accurately quantified. The two decomposed components become a bridge connecting the market layer and the physical layer (dispatch operation). Market participants can autonomously and flexibly decide on the combination of transaction paths based on the dynamic cost differences between the two. Their decision results can then be directly input into the operation optimization model through the components, naturally guiding market behavior to align with the physical economic operation goals of the power grid. Through decomposition, cost calculation can respond in real-time to network topology changes and SOP operating status, dynamically adjusting the rates of each path, overcoming the problems of traditional fixed or linear rate models.

[0029] To quantify the impact of these two power components on the network, this embodiment proposes a virtual network decomposition method. This method is applicable to networks with specific SOPs. network Create two virtual, purely radial subnetworks. and .

[0030] Subnetwork :exist Disconnect SOP branch , that is to say The network uses the original main network connection point as the balancing node. In this network, the power components... Treated as a node Injection and nodes The outflow of energy follows the physical laws of this radial network.

[0031] Subnetwork :exist Retain SOP branch However, the SOP device itself is regarded as a new virtual balance node. All nodes in the original network (including the original balanced node) are considered PQ nodes in this subnet. The two end nodes of the SOP... and Through this virtual balance node Power exchange is achieved. In this network, power components... The power flow is calculated. The significance of this processing is that, through transactions along the SOP path, the power essentially relies on the SOP's fast power transfer capability, electrically equivalent to that at the node... and A controllable power source and load pair were added between them.

[0032] This step, using a virtual network decomposition method, obtains the power components flowing through the traditional network and the power components flowing through specific SOP devices. It also yields a pure radial network for carrying and analyzing the power components of non-SOP paths and their impact, and a pure radial network with SOPs as virtual balancing nodes for carrying and analyzing the power components of SOP paths and their impact. Subsequent dynamic rotation cost calculation... This reflects the principles of path-based decomposition and component-based pricing. The calculated dynamic rotation cost and total P2P cost for users are... .

[0033] (1-2) Calculation of dynamic rotation cost based on power transmission distribution factor (PTDF) and electrical distance.

[0034] To accurately allocate network physical losses and asset occupancy costs to each P2P transaction, the following concepts and calculations need to be introduced.

[0035] For subnetworks any line in Its electrical distance Defined as: (2) in, For the line The resistance, This is the system reference resistor. When k=0, it represents the network corresponding to the non-SOP path; when k=s, it represents the network corresponding to the path flowing through a specific SOP (the s-th SOP).

[0036] Electrical distance is dimensionless and directly reflects the relative value of energy loss when power flows through the line.

[0037] The Power Transfer Distribution Factor (PTDF) matrix describes the linear sensitivity relationship between power transfer between nodes and branch power flow in a network, particularly for radial subnetworks. Its node-branch correlation matrix is The branch impedance matrix is Selecting the equilibrium node ,for For the original equilibrium node, For virtual nodes Then the PTDF matrix It can be calculated using the following formula: (3) in, Let be the imaginary part of the network node admittance matrix. and To and The correlation coefficient matrix. Matrix elements. Indicates in subnet In the middle, node Injecting 1 unit of active power and balancing the node When absorbing 1 unit of active power, in the line The change in active power generated above.

[0038] The network node admittance matrix is ​​constructed from the physical characteristics of the network, and different networks exhibit different values. Parameters f and b are the node-branch correlation matrices. The coefficient matrix. Branch impedance matrix. It corresponds to the virtual subnetwork It is a lumped mathematical representation of the electrical parameters of all branches in the distribution network. It is a diagonal matrix, which is obtained directly from known data on the physical line and equipment parameters of the distribution network.

[0039] Node-branch association matrix It is a specific subnetwork constructed based on the virtual network decomposition method. The topological connections are obtained directly through graph theory methods.

[0040] For a radial distribution network with n nodes and m branches Its node-branch correlation matrix It is an n×m matrix, and each element of it... The connection relationship between node i and branch l is defined: =+1 indicates that the positive direction of branch l is from node i. =-1 indicates that the positive direction of branch l is the flow into node i. =0 indicates that node i and branch l are not directly connected.

[0041] Node-branch association matrix The acquisition process specifically includes: Step 1: Determine the subnetwork For the specific topology, (Non-SOP path subnet): Disconnect the SOP branch SOP(m,n) in the original network G to form a pure radial network. Use the original main network connection point as the balancing node.

[0042] for (SOP Path Subnet): The SOP branch is retained in the original network G, but the SOP device itself is regarded as a new virtual balancer node, which can be numbered as node v. At this time, the two end nodes m and n of the SOP are connected through this virtual node v, and the network is still radial.

[0043] Step 2: List the set of nodes and the set of branches.

[0044] Node set : Define subnetwork All nodes and their numbers, for Special attention should be paid to the inclusion of the virtual balance node v.

[0045] Branch set : Define subnetwork All branches and their numbers are listed, and a positive direction is pre-assigned to each branch.

[0046] Step 3: Fill the matrix according to the definition .

[0047] Create a size of × The zero matrix is ​​then used to connect nodes i and j for each branch l. Assume the positive direction is i→j: 1. In the column corresponding to branch l, fill in +1 in the row corresponding to node i.

[0048] 2. Fill in the row corresponding to node j. 1.

[0049] 3. The remaining elements in this column are 0.

[0050] Step 4: Process the balancing node.

[0051] In power flow calculations, it is necessary to select a slack node correlation matrix. This typically corresponds to a reduced-order incidence matrix, which is the matrix after removing the rows containing the slack nodes. This is done because the power of the slack nodes is automatically determined by the system equilibrium and is not treated as an independent equation.

[0052] Branch impedance matrix It corresponds to the virtual subnetwork It is a lumped mathematical representation of the electrical parameters of all branches in the distribution network. It is a diagonal matrix, which is obtained directly from known data on the physical line and equipment parameters of the distribution network.

[0053] Based on the PTDF matrix, transaction components In subnetwork The line caused by the middle The additional power flow is: (4) definition For trading pairs For the line The sensitivity coefficient.

[0054] The turnover cost generated by this transaction component is: (5) in, Price of transmission services per unit electrical distance-power. Absolute value term. This means that regardless of the power flow direction, occupying line capacity incurs costs. The formula multiplies the electrical distance by the PTDF sensitivity as a weight for cost allocation, ensuring that the cost accurately reflects the actual impact of the transaction on network losses and asset occupancy.

[0055] Among them, the dynamic rotation cost is the rotation cost for non-SOP paths. Plus SOP path transaction loss costs The sum of these. The aforementioned billing mechanism is no longer a traditional fixed rate, but is deeply coupled with the real-time physical state, topology, and adjustment behavior of SOP flexible equipment of the power network.

[0056] For the transaction components of the SOP path In addition to rotation costs, the operating losses of the SOP converter itself must also be borne. Assume the power loss rate of the SOP is... The power loss is approximately Let the cost conversion factor for power loss be... The additional cost of SOP loss is: (6) This represents the length of the scheduling period.

[0057] Therefore, trading pairs During the period The total transaction cost, which is the user's During the period The total P2P cost is the sum of the costs of all counterparties: (7) in, The cost of a single transaction for the two parties in the trading pair (i,j) during time period h. The transaction cost of the power component flowing through the s-th SOP path for the trading pair (i,j) during time period h.

[0058] (8) Formula (8) above represents the user During the period The total cost of P2P. This cost is the ultimate economic manifestation of the system's dynamic cost accounting mechanism.

[0059] When multiple Standard Operating Procedures (SOPs) exist in the network, the decomposition of transaction paths and cost calculations can be naturally expanded, and transaction power can be decomposed into: (9) in, Specifically refers to the power component flowing through the s-th SOP. It is a set of Standard Operating Procedures (SOPs). Each This corresponds to a subnetwork with this SOP as the virtual balancer node. and calculate its rotation costs independently. and loss cost The total cost is the sum of the costs of all paths, which demonstrates the model's good scalability for complex, flexible interconnected networks.

[0060] Only by decomposing the trading power can differentiated service costs be calculated for different paths. The round-trip cost in Formula 5 is calculated based on the decomposed power components. Formula (7) then sums up the costs of each path after decomposition to obtain the total cost of the trading pair.

[0061] Step 2: Robust collaborative scheduling of flexible resources with uncertainties.

[0062] After establishing the dynamic cost signaling mechanism formulas (1)-(9) that can accurately reflect physical reality, it is necessary to construct a short-term operation optimization model that can fully schedule various flexible resources and minimize the total system cost. This model must be able to effectively cope with the strong uncertainty of renewable energy output and load demand. Therefore, a robust optimization framework is adopted, the goal of which is to optimize the system's performance under the worst-case uncertainty scenario, thereby ensuring the reliability and resilience of the operation plan.

[0063] (2-1) Objective function and robust framework.

[0064] Operational optimization is oriented towards a scheduling cycle, assuming that the uncertainty mainly comes from photovoltaic power output. and fixed load They exist in a bounded uncertain set Internal fluctuations. The two-stage robust optimization formulation is as follows: Phase 1: Before uncertainty materializes, the decision-making process must determine the non-adaptive variables at this stage. Let all the decision variables in Phase 1 be a vector y, and its feasible region be... The specific decision lies in whether to build energy storage at location i and the capacity of the energy storage to be built.

[0065] Phase Two: Amidst Uncertainty Once implemented, adaptive decisions are made based on the observed specific scenarios to determine the minimum operating cost of the system.

[0066] The robust operation optimization objective is: to select the first-stage decision variable as vector y, so that it can withstand the worst uncertainty scenarios. At this point, the total expected operating cost resulting from the adaptive decision-making in the second stage is minimized. Mathematically, this can be expressed as: (10) Where y is the decision variable for the first stage. For the corresponding given The second-stage optimal decision. This is the feasible region for the first phase.

[0067] Formula (10) defines the total operating cost of the system, which is the objective function of the second stage (operation and scheduling stage) in the entire optimization model. This formula determines the installation location and rated capacity of the energy storage system. Under the configuration scheme determined in the first stage, the optimal operation and scheduling strategy is found using formula (10) for the worst-case scenario of renewable energy and load.

[0068] Operating costs of user i Specifically as follows: (11) (12) (13) in, The power purchased by user i from the mainnet. For the power exchange between the power grid and users, The power sold by user i to the mainnet. and The time-of-use electricity purchase price and sales price for period h. For energy storage devices During the period The depreciation cost due to charging and discharging.

[0069] (2-2) Multi-resource collaborative constraint system. This part, together with the subsequent (2-3) energy storage model and (2-4) DistFlow model, constitutes the complete constraint conditions of the short-term two-stage robust optimization model.

[0070] The established switch state variables and SOP power variables are used as part of the first-stage decision variable y, determined before the realization of uncertainty. In the second stage, after the realization of uncertainty, these topology states and SOP adjustment capabilities determine whether the system can cope with extreme scenarios by adjusting the power flow distribution. The SOP power variables and line power flow variables in (2-2) provide the transmission path for the charging and discharging power of the energy storage in (2-3). The linearized DistFlow equation in (2-4) directly depends on the network topology (switch states) and power flow variables determined in (2-2).

[0071] 1) Unified modeling of network topology flexibility.

[0072] Traditional switch reconfiguration and SOP adjustment are two sources of flexibility with different time scales and action characteristics. This invention unifies them by using hybrid integer programming.

[0073] Switching between switch state and power flow direction: For each operable line, a binary variable is introduced. , This indicates that the line closes during time period h. Simultaneously, continuous variables are introduced. This represents the active power flow of line (i,j), with the direction being... The Big M method is used to couple discrete and continuous variables: (14) when hour, This means there is no power transmission on the line. When hour, It can vary freely within physical limits.

[0074] To ensure network security, a radial topology is required for the final network. In a flexible interconnected distribution network containing SOPs, the presence of the SOP itself forms an electrical loop. This invention treats the SOP and its two connected nodes (m,n) as a controllable power-switching pair, rather than a simple closed branch. For the network consisting of conventional switches, excluding the SOP branch, a strict radial constraint is still imposed. The power of the SOP... and As part of the injected power of nodes, they participate in power flow calculations. Their existence does not affect the radial determination of the upper-level network topology. This resolves the contradiction between allowing SOP closed-loop operation and maintaining the radial state of the network, and clarifies that SOP is a point-to-point power regulation device, rather than a connection line that changes the basic structure of the network.

[0075] (2-3) Energy storage operation and lifetime loss constraints: In short-term robust scheduling, the allocation of energy storage constitutes the feasible region of energy storage operation. In the second-stage decision-making (long-term two-stage collaborative planning), given the uncertainty scenario, the system must satisfy these constraints to find a feasible scheduling scheme.

[0076] Let the first A shared energy storage station (SES) serves a set of users. In addition to basic power balance constraints, capacity allocation weights are introduced to ensure investment fairness and prevent individual users from over-consuming resources. And impose the following constraints: (15) (16) in, User The virtual state of charge in SESm and This represents the actual stored energy and rated power of the m-th shared energy storage station SESm. and For the charging and discharging power commands assigned to user i.

[0077] Physical status of shared energy storage station (SES) and power It must be consistent with the sum of the virtual states and power allocations of all users.

[0078] To more accurately reflect the economics of battery energy storage, an improved approach is to incorporate energy storage depreciation costs. Functions related to discharge throughput and average depth of discharge: (17) in, For energy storage systems The initial total investment cost, For energy storage systems Rated capacity, For the battery at average depth of discharge The total number of charge-discharge cycles that can be completed during its lifespan. For time period The discharge energy within.

[0079] (2-4) Linearization and convex relaxation of DistFlow constraint considering uncertainty.

[0080] To efficiently handle power flow constraints in robust optimization, the DistFlow constraints are linearized and convex relaxed. The original DistFlow branch power flow equations are: (18) (19) (20) Nonlinear terms This leads to nonconvexity, and DistFlow serves as part of the constraints for short-run two-stage robust optimization.

[0081] For radial distribution networks, the relaxation is usually tight within a reasonable voltage and current range, thus providing an accurate power flow approximation while ensuring computational efficiency.

[0082] The convex relaxation step transforms the non-convex DistFlow constraint into a convex optimization problem, guaranteeing a globally optimal solution and improving computational efficiency. This provides constraints that support robust optimization iterations for robust runtime scheduling and long-term planning decisions.

[0083] 1) Define a new variable: Let , is the square of the voltage amplitude. , where is the square of the current amplitude. According to Ohm's law, we have...

[0084] 2) Rewrite the original equation as: (twenty one) (twenty two) (twenty three) in, and This represents the sum of all paths originating from node j and flowing to child node k; The active power flowing from node i to node j on line ij; The reactive power flowing from node i to node j on line ij; The net active power load of node j; Net reactive load at node j; and Voltage magnitudes at nodes i and j; and For the resistance and reactance of the line; The current amplitude on line ij.

[0085] 3) Non-convex constraints It is relaxed to the following second-order cone inequality: (twenty four) in, It represents the 2-norm of a vector.

[0086] 4) Apply voltage and current safety constraints: (25) Among them, and node Permissible lower and upper voltage limits, For the line The maximum effective value of the current that can be carried.

[0087] For a typical radial distribution network, under conditions where the operating voltage is close to the rated value and the line load rate is reasonable, the above-mentioned second-order cone relaxation is tight, meaning that the optimal solution after relaxation will automatically satisfy the original non-convex equation. This will provide a sufficiently accurate power flow solution while ensuring computational efficiency.

[0088] Step 3: Two-stage collaborative planning method and solution for energy storage and network flexibility resources.

[0089] The final energy storage configuration is a long-term capital investment that can adapt to various possible operating scenarios, especially extreme ones. Embedding the aforementioned robust operating model that accurately considers uncertainties into the planning and decision optimization process, forming a planning and operation closed loop, allows investment decisions to be directly assessed based on its ability to withstand long-term operating risks, which is key to achieving optimal life-cycle costs.

[0090] (3-1) The standard form of the two-stage robust programming problem.

[0091] The objective of the main problem is to minimize the upper bound of the sum of the investment cost and the worst-case operating cost: (26) (27) in, It is an auxiliary variable representing the worst-case operating cost. and For in position The unit capacity cost and unit power cost of building energy storage and For in position Maximum capacity and power that can be built In the given investment plan and a specific uncertainty scenario The minimum operating cost of the second phase running subproblem.

[0092] The subproblem is given an investment plan The following is a max-min problem: Evaluate its worst performance across all possible uncertainty scenarios. (28) Among them, the outer layer Find the worst-case uncertainty scenario that maximizes the total operating cost of the system, inner layer Given an investment plan and a specific scenario Next, we solve the aforementioned robust operation optimization problem, including the first stage of operation decision-making. Second Phase Operation Decision This yields the minimum feasible operating cost for that scenario.

[0093] (3-2) Application process of column and constraint generation (C&CG) algorithm.

[0094] The C&CG algorithm iteratively solves the main problem and subproblems, continuously adding the worst-case scenarios discovered by the subproblems, along with their corresponding operational constraints and optimality cuts, to the main problem, thereby gradually approaching the solution of the original two-stage problem. The process is as follows, and it embodies the core idea of ​​integrating long-term uncertainty into planning and decision-making: Step S1: Initialization. Set convergence tolerance. Upper Realm The lower realm Iteration counter Extreme scenario collection .

[0095] Step S2: Solve the main problem. At this point, the scenario set... It may be empty or contain scenarios discovered in previous iterations. Solving this gives the current optimal investment plan. and target value Update the upper bound: .

[0096] Step S3: Solve the subproblem, the fixed investment plan is as follows This is a max-min problem. By taking the dual of the inner minimization problem, the inner max-min problem can be transformed into a single-layer maximization problem, thus transforming the subproblem into a single-layer maximization problem. The transformed problem is usually a linear programming or cone programming problem, which can be solved directly. Solving for the worst-case scenario yields the solution. and the corresponding optimal operating cost Update the Nether: .

[0097] Step S4: Convergence check. If convergence is found, the algorithm terminates, and the current investment plan is the optimal solution. Otherwise, proceed to step S5.

[0098] Step S5: Add the Benders Cut. Based on the solutions to the subproblems, construct a linear inequality (i.e., the Benders Cut) regarding the investment variables of the main problem, and add it to the constraints of the main problem. The plane form of this cut is as follows: (29) in, Represents the dot product of vectors. It is the dual multiplier vector related to the investment variable, which quantifies the impact of small changes in the investment variable on operating costs. This cutting plane guarantees that in solving the future of the main problem, the corresponding dual multiplier vector for any investment option is the same. The value will not be lower than the scenario that this solution is designed to handle. The lower bound of the cost.

[0099] Step S6: Newly discovered extreme scenarios Add to collection ,make Return to step S2.

[0100] Through the above iterations, the investment scheme in the main problem is continuously adjusted to prevent one extreme scenario after another from being found in the subproblems. Finally, when the upper bound and the lower bound are close enough, a robust optimal configuration scheme is obtained. This scheme ensures the economic efficiency of investment while ensuring that the system has a feasible operation scheduling strategy under any allowed uncertainty scenario in the future, and that its total cost is controllable and approximately optimal.

[0101] The optimal configuration includes: energy storage location: at which nodes to install energy storage; energy storage capacity: how much electricity the energy storage can store; and energy storage power: the maximum charging and discharging power of the energy storage.

[0102] In this implementation example, P2P transaction path decomposition and dynamic modeling of rotation costs are adopted: In flexible interconnected distribution networks, factors such as P2P transaction path selection, SOP access status, and network topology changes directly affect transaction costs, and traditional fixed-rate models cannot adapt to their dynamic and uncertain nature. A dynamic cost calculation method based on PTDF and network decomposition needs to be established to accurately quantify rotation costs under different path and network conditions, providing market transactions with cost signals coordinated with economic dispatch.

[0103] In this implementation example, robust collaborative scheduling of multiple flexible resources with uncertainties is required: the strong uncertainty of photovoltaic output and load in the distribution network, as well as the differences in the response characteristics of resources such as SOP, energy storage, and adjustable loads, increase the complexity of system collaborative operation. A robust multi-resource operation and scheduling model considering uncertainties needs to be constructed to uniformly optimize various resources and ensure that the system can still operate safely and economically under extreme scenarios.

[0104] Short-run two-stage robust optimization focuses on operational optimization within a single scheduling cycle, representing a single-timescale operational scheduling problem. Long-run two-stage robust programming, on the other hand, couples the planning (long-run) and operational (short-run) timescales together.

[0105] In this implementation example, a two-stage robust collaborative planning method and solution for energy storage systems are presented. Energy storage configuration needs to address uncertainties such as photovoltaic power generation, long-term load fluctuations, and diverse operating strategies. Traditional planning methods struggle to assess its true value and risks throughout its entire lifecycle. Therefore, a closed-loop planning method based on two-stage robust optimization and the C&CG algorithm is needed to feed operational risks back to investment decisions, achieving collaborative planning of energy storage and network flexibility resources.

[0106] Short-term scheduling is a "sub-problem," addressing operational optimization on a daily or hourly basis given a specific energy storage configuration. Long-term planning is the "main problem," with decision variables being the location, capacity, and power of the energy storage. This involves capital investment decisions on an annual basis. Long-term planning first proposes an investment plan. Short-term scheduling then incorporates this plan into its model, subjecting it to various potential extreme weather or load scenarios. If an extreme scenario leads to excessively high operating costs or inoperability, short-term scheduling feeds back the scenario and its corresponding cost information to long-term planning. Long-term planning adjusts its investment plan based on this feedback, such as increasing energy storage capacity at critical points to cope with the extreme scenario. This iterative process continues until a solution is found that is both economically viable and can withstand all extreme scenarios.

[0107] The aforementioned short-term dispatch model uniformly optimizes SOPs, network reconfiguration, and energy storage. When long-term planning assesses whether energy storage needs to be built at a remote node, short-term dispatch attempts to resolve voltage issues using SOP transfers or adjusting switch states. If these flexible resources can address the problem, the planning level will not blindly invest in energy storage; only if they cannot will the planning level decide to invest. This achieves cost-effective investment at the expense of operational flexibility. It provides accurate market signals, and the dynamic cost accounting mechanism based on PTDF and electrical distance influences P2P trading decisions and power flow distribution in short-term dispatch, thereby affecting the assessment of energy storage economics in long-term planning. This ensures the entire optimization process is based on real physical costs, avoiding market signal distortion caused by fixed grid access fees, thus guiding energy storage investment to locations that best alleviate network congestion and reduce network losses.

[0108] Given the current lack of P2P transactions and dynamic cost accounting in existing technologies, as well as the lack of unified modeling for Standard Operating Procedures (SOPs) and network reconfiguration, this embodiment's sub-solution integrates P2P market transactions with a dynamic cost accounting mechanism based on the physical characteristics of the distribution network. This addresses the problem that traditional fixed-rate models cannot reflect the impact of SOP access and topology changes on transaction costs, providing accurate physical cost signals for market transactions and guiding market behavior to proactively cooperate with the safe and economical operation of the power grid. A robust collaborative planning method for the entire lifecycle of the "market-operation-planning" closed loop is incorporated, embedding a short-term robust operation model, including dynamic cost signals and multi-resource collaborative scheduling, as a sub-problem into the long-term two-stage robust planning, and solving it iteratively using the C&CG algorithm. The final energy storage configuration scheme not only considers investment costs but also quantifies its total lifecycle cost under extreme scenarios to achieve robust optimization.

[0109] Example 2 The purpose of this embodiment is to provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the above-described method.

[0110] Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0111] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the above method.

[0112] Example 4 The purpose of this embodiment is to provide an energy storage configuration optimization system for flexible interconnected distribution networks, including: The dynamic cost accounting mechanism establishment module is configured to: establish a dynamic cost accounting mechanism that is coupled in real time with the network physical state; The short-term two-stage robust optimization module is configured to: construct a short-term two-stage robust optimization model based on a dynamic cost accounting mechanism that can fully schedule various flexible resources and aims to minimize the total system cost, thereby obtaining the optimal energy storage operation strategy; The long-term two-stage robust optimization module is configured to: establish a long-term two-stage robust optimization model, the main problem being to minimize the upper bound of the sum of investment cost and worst-case operating cost, and the sub-problem being to evaluate its worst performance in all possible uncertainty scenarios under a given investment plan. By iteratively solving the main problem and its subproblems, a robust optimal configuration scheme is obtained.

[0113] Example 5 The purpose of this embodiment is to provide a computer program product containing instructions that, when run on a computer, causes the computer to perform the methods and functions involved in any of the embodiments described above.

[0114] The steps and methods involved in the apparatus of the above embodiments correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0115] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0116] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for optimizing energy storage configuration in flexible interconnected distribution networks, characterized in that, include: Establish a dynamic cost accounting mechanism that is coupled in real time with the physical state of the network; Based on the dynamic cost accounting mechanism, a short-term two-stage robust optimization model is constructed that can fully allocate various flexible resources and aims to minimize the total system cost, so as to obtain the optimal operation strategy for energy storage. A long-term two-stage robust optimization model is established. The main problem is to minimize the upper bound of the sum of investment cost and worst-case operating cost. The sub-problem is to evaluate its worst performance in all possible uncertainty scenarios under a given investment plan. By iteratively solving the main problem and its subproblems, a robust optimal configuration scheme is obtained.

2. The energy storage configuration optimization method for flexible interconnected distribution networks as described in claim 1, characterized in that, When establishing a dynamic cost accounting mechanism that is coupled in real time with the physical state of the network, the specific components include: A transaction and dynamic rotation cost model is established. Based on the model, the transaction path is decomposed into a network to obtain two physical components of independent decisions. The virtual network decomposition method is used to quantify the impact of these two physical components on the network. After quantifying the impact of these two physical components on the network, the dynamic rotation cost is calculated based on the power transmission distribution factor and electrical distance.

3. The energy storage configuration optimization method for flexible interconnected distribution networks as described in claim 1, characterized in that, The two independent physical components are: the power component flowing through the non-SOP path, whose power flow is carried only by the traditional network, and the power component flowing through the SOP path, whose power flow must pass through the SOP device.

4. The energy storage configuration optimization method for flexible interconnected distribution networks as described in claim 1, characterized in that, The virtual network decomposition method is used to quantify the impact of these two physical components on the network, specifically including: For those with specific SOPs network Construct two virtual, purely radial subnetworks. and ; Subnetwork :exist Disconnect SOP branch , that is to say The network uses the original main network connection point as the balancing node. In this network, the power components... Treated as a node Injection and nodes The outflow of energy follows the physical laws of this radial network. Subnetwork :exist Retain SOP branch However, the SOP device itself is regarded as a new virtual balance node. All nodes in the original network are considered PQ nodes in this subnet, and the two end nodes of the SOP are... and Through this virtual balance node Achieve power exchange.

5. The energy storage configuration optimization method for flexible interconnected distribution networks as described in claim 1, characterized in that, Dynamic round-trip costs are calculated based on power transmission distribution factor and electrical distance, specifically including: For any line in a subnetwork, its electrical distance is defined as the line distance. The ratio of the resistance to the system reference resistance; The power transfer distribution factor matrix is ​​used to describe the linear sensitivity relationship between power transfer between nodes and branch power flow in a network.

6. The energy storage configuration optimization method for flexible interconnected distribution networks as described in claim 1, characterized in that, The short-term two-stage robust optimization model specifically includes: The objective of robust operation optimization is to select the first-stage decision so that the total expected operating cost caused by the second-stage adaptive decision is minimized when facing the worst uncertainty scenario. The multi-resource collaborative constraint system includes: establishing a refined model of energy storage operation and lifetime loss based on unified modeling of network topology flexibility; linearization and convex relaxation of the DistFlow model considering uncertainties.

7. An energy storage configuration optimization system for flexible interconnected distribution networks, characterized in that, include: The dynamic cost accounting mechanism establishment module is configured to: establish a dynamic cost accounting mechanism that is coupled in real time with the network physical state; The short-term two-stage robust optimization module is configured to: construct a short-term two-stage robust optimization model based on a dynamic cost accounting mechanism that can fully schedule various flexible resources and aims to minimize the total system cost, thereby obtaining the optimal energy storage operation strategy; The long-term two-stage robust optimization module is configured to: establish a long-term two-stage robust optimization model, the main problem being to minimize the upper bound of the sum of investment cost and worst-case operating cost, and the sub-problem being to evaluate its worst performance in all possible uncertainty scenarios under a given investment plan. By iteratively solving the main problem and its subproblems, a robust optimal configuration scheme is obtained.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 6.

9. 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, it implements the steps of the method described in any one of claims 1-6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it performs the steps of the method described in any one of claims 1-6 above.

Images