Power distribution network and optical storage and charging system collaborative planning method considering N-1 safety criterion

Abstract

The invention discloses a power distribution network and optical storage and charging system collaborative planning method considering an N-1 safety criterion, and belongs to the field of novel power systems. The method comprises the following steps: constructing a power distribution network computable power flow model based on second-order cone relaxation and linearization, and constructing a light storage and charging energy hub model based on EH multi-energy flow coupling; constructing a two-stage stochastic programming framework, performing joint optimization on investment and operation, and performing N-1 fault recovery simulation; maximizing an uncertainty interval of photovoltaic output and charging load fluctuation at a fault moment by adopting an information gap decision theory; and based on the double-stage stochastic programming framework, a fault problem is decomposed into a main problem and a corresponding sub-problem, key element identification is carried out, and fault verification is completed. According to the method, the total cost of the whole life cycle of the system is minimized on the premise that the N-1 safety criterion is met, the power supply safety in an extreme scene is ensured, and the calculation efficiency of actual engineering application is ensured.

Description

Technical Field

[0001] This invention relates to the field of novel power systems, and more specifically to a collaborative planning method for distribution networks and photovoltaic-storage-charging systems that considers the N-1 safety criterion. Background Technology

[0002] Traditional distribution network planning methods typically separate the planning of grid expansion (lines, substations) from that of distributed power sources, energy storage, and charging facilities, or simply superimpose them under simplistic assumptions, failing to achieve globally optimal coordination of "grid-source-storage-load" resources. Furthermore, directly performing full-scenario N-1 security checks on large-scale, complex distribution networks involves enormous computational demands, becoming a major bottleneck in engineering applications.

[0003] The paper "Joint Planning of Distribution Network and Distributed Energy Storage Considering the N-1 Criterion" published by Lin Zhe, Hu Zechun, and Song Yonghua is based on the characteristics of "closed-loop design and open-loop operation" of actual distribution networks. Its physical structure includes the distribution network structure to be planned and several distributed energy storage systems to be located and sized. Photovoltaics and charging piles are typically included in the model as known or predicted boundary conditions, but their configuration is not the core optimization decision variable.

[0004] The core contribution of this most similar scheme lies in incorporating energy storage as a flexible resource into the planning framework and designing an efficient iterative verification strategy to overcome computational challenges. However, it still has limitations when dealing with the "photovoltaic-energy storage-charging" deeply integrated system targeted by this invention:

[0005] Treating photovoltaic (PV) systems and charging stations as passive, fixed boundary conditions, rather than as active decision variables for coordinated optimization with the grid and energy storage, fails to achieve complementarity between charging station deployment and weak points in the grid, and also fails to maximize the utilization of the synergistic charging and discharging characteristics of PV and energy storage to improve economic efficiency and reliability.

[0006] The handling of uncertainties in photovoltaic output and charging demand (especially fluctuations during fault periods) is relatively simplified, and typical scenarios or simple conservative ranges may be adopted. It is difficult to accurately balance investment costs and power supply risks under extreme conditions in the planning, and the robustness of the resulting solutions needs to be improved.

[0007] It mainly relies on the "discharge" of energy storage to support fault recovery, without fully considering the controllability of photovoltaics during faults and the "vehicle-grid interaction" potential of electric vehicle charging loads, and fails to build an integrated "photovoltaic-storage-charging" comprehensive response and fault support strategy.

[0008] Existing technologies have the following problems when coordinating distribution network planning for high-proportion distributed photovoltaic, energy storage, and electric vehicle charging load integration:

[0009] (1) There is a disconnect between the planning of the "grid-source-load" system and the high cost of safety redundancy. Traditional distribution network planning methods usually regard grid expansion and the configuration of user-side distributed resources (such as photovoltaics, energy storage, and charging piles) as two independent or sequential processes. This fragmented planning model makes it impossible to comprehensively consider the potential of distributed resources as backup power sources during failures in the planning stage. In order to meet the strict N-1 safety criterion, planning schemes often tend to over-invest in the construction of physical lines and substation capacity to provide sufficient redundancy, resulting in low utilization of grid assets and poor overall economic efficiency.

[0010] (2) In terms of dealing with uncertainty, there is insufficient consideration of the dynamic risks at the moment of failure. Existing planning methods are mostly based on certain typical scenarios and fail to fully consider the combined effects of various uncertainties such as the randomness of photovoltaic output, the spatiotemporal fluctuations of electric vehicle charging load, and the deviation of conventional load forecasting at the moment of N-1 failure. This results in insufficient robustness of the planning scheme in real failure scenarios, which may lead to problems such as the failure of the preset failure recovery strategy, a surge in curtailment rate, or load loss exceeding expectations, and cannot guarantee the reliability of power supply under extreme conditions.

[0011] (3) At the level of model solving and engineering application, there is a bottleneck problem of high computational complexity in N-1 security verification. In order to achieve reliable planning, the planning scheme must be verified by N-1, that is, the system state after the failure of each single component. The traditional method adopts the "full scenario enumeration" strategy. As the scale of the distribution network expands, the number of scenarios to be verified increases sharply. At the same time, the AC power flow model of the distribution network itself has non-convex and nonlinear characteristics, which makes the joint planning model containing a large number of N-1 scenarios extremely large in scale, extremely difficult to solve, computationally time-consuming, and even unable to converge, which seriously restricts the application of this technology in large-scale actual power grids. Summary of the Invention

[0012] To address the shortcomings of existing technologies, the present invention aims to provide a collaborative planning method for distribution networks and photovoltaic-storage-charging (PV-SSD-Charging) systems that considers the N-1 safety criterion.

[0013] The objective of this invention is achieved through the following technical solution: a collaborative planning method for distribution networks and photovoltaic-storage-charging systems considering the N-1 safety criterion, comprising the following steps:

[0014] Phase angle relaxation is applied to the standard branch power flow model to obtain a real equation model suitable for optimization calculation. The nonlinear constraints in the real equation model are relaxed by second-order cone relaxation and linearized to construct a computable power flow model for the distribution network.

[0015] Each photovoltaic power generation device-energy storage device-charging station is abstracted as an energy hub, and the power from the distribution network and the power generated by photovoltaic power generation are used as inputs to construct a photovoltaic-storage-charging energy hub model;

[0016] The power distribution network model is non-convex and nonlinear, making direct application to an optimization problem extremely difficult. Therefore, this invention employs a series of mathematical transformations, including phase angle relaxation, second-order cone relaxation, and linearization, to transform it into a convex, efficiently solvable optimization problem while maintaining engineering accuracy. Although the photovoltaic-storage-charging model also contains nonlinear relationships, its core constraints are relatively easier to linearize or piecewise linearize. Decoupling it from the complex power grid model and modeling it separately makes the model structure clearer.

[0017] For the two states of the photovoltaic-storage-charging system, namely normal operation or N-1 fault, a two-stage stochastic programming framework is constructed. With the goal of minimizing the annualized total cost of the photovoltaic-storage-charging system, investment and operation are jointly optimized, and N-1 fault recovery simulation is performed.

[0018] The information gap decision theory is adopted to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure;

[0019] Based on the aforementioned two-stage stochastic programming framework, the fault problem is decomposed into a main problem and corresponding sub-problems. The main problem does not consider N-1 constraints and is used to solve for an economic solution. The sub-problems are used to perform N-1 fault verification on the economic solution after solving the main problem and identify faulty components. If a fault is found, the fault is used as a constraint of the main problem and the solution is re-solved. The solution process is iterated until the verification is error-free.

[0020] Furthermore, the linearization is achieved through circumscribed polygon linearization techniques.

[0021] Furthermore, the photovoltaic-storage-charging energy hub model includes:

[0022] A photovoltaic model, based on a typical solar radiation curve, represents the output characteristics of a photovoltaic system;

[0023] An electrochemical energy storage model is used to track changes in the energy state of energy storage devices, and is constrained by charge and discharge efficiency, power, energy state, and upper and lower limits of capacity.

[0024] The electric vehicle charging load model is used to aggregate distributed electric vehicle charging loads into a controllable overall load, which is constrained by the number of charging piles, power level, and user charging demand curve.

[0025] Furthermore, the joint optimization of investment and operation, with the goal of minimizing the annualized total cost of the photovoltaic-storage-charging system, includes:

[0026] An investment and operation model is constructed with minimizing the annualized comprehensive cost of the photovoltaic-storage-charging system as the objective function. The annualized comprehensive cost includes the investment cost and the expected operating cost.

[0027] The constraints include convex power flow constraints of the distribution network, operational constraints of the photovoltaic-storage-charging system, and power balance constraints of the photovoltaic-storage-charging system.

[0028] Furthermore, the N-1 fault recovery simulation includes:

[0029] The response of a simulated photovoltaic energy storage and charging system when it encounters any preset N-1 fault;

[0030] The decision variables are the network reconfiguration switch status after a fault, the emergency discharge power of the energy storage device, the active and reactive power support power of photovoltaics, and the adjustment amount of interruptible or transferable charging load.

[0031] The objective function is to minimize the total load reduction under this specific N-1 fault scenario. The optimization solution is then implemented, and power flow constraints of the distribution network and operation constraints of the photovoltaic-storage-charging system based on the new topology after the fault are introduced. Non-critical loads are also allowed to be reduced.

[0032] The specific constraints are as follows:

[0033] ;

[0034] The meanings of the symbols for each parameter are shown in Table 1 below:

[0035] Table 1: Meaning of each parameter ;

[0036] If the total load reduction exceeds a preset value after verification and optimization, the current planning framework is determined not to meet the N-1 safety criterion, and feedback information is generated.

[0037] Furthermore, the method of employing information gap decision theory to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure includes:

[0038] Based on the information gap decision theory, an uncertainty model is constructed to address the uncertainty of photovoltaic power output and charging load fluctuations at the time of failure, and the uncertainty of photovoltaic power output and charging load fluctuations is represented by the fluctuation deviation coefficient.

[0039] The objective function is to replace minimizing the annualized total cost of the photovoltaic-storage-charging system with maximizing the fluctuation deviation coefficient, while adding cost constraints.

[0040] Furthermore, the decomposition of the fault problem into a main problem and corresponding sub-problems includes:

[0041] The main problem includes a set of key N-1 fault scenarios. The set of key N-1 fault scenarios is initially empty. Whenever a subproblem is solved and an N-1 fault scenario is obtained, the N-1 fault scenario is added to the set of key N-1 fault scenarios.

[0042] The present invention also provides a collaborative planning device for distribution network and photovoltaic-storage-charging system considering the N-1 safety criterion, comprising:

[0043] The computable power flow model construction module for distribution networks is used to relax the phase angle of the standard branch power flow model to obtain a real equation model suitable for optimization calculation. The nonlinear constraints in the real equation model are relaxed by second-order cone and linearized to construct a computable power flow model for distribution networks.

[0044] The photovoltaic-storage-charging energy hub model construction module is used to abstract each photovoltaic power generation device-energy storage device-charging station into an energy hub, and uses the power from the distribution network and the power generated by photovoltaic power generation as inputs to construct the photovoltaic-storage-charging energy hub model;

[0045] The two-stage stochastic programming framework construction module constructs a two-stage stochastic programming framework for two states of the photovoltaic storage and charging system: normal operation or N-1 fault. The framework aims to minimize the annualized total cost of the photovoltaic storage and charging system, jointly optimizes investment and operation, and performs N-1 fault recovery simulation.

[0046] The risk avoidance module is used to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure, based on the information gap decision theory.

[0047] The problem decomposition and iterative verification module, based on the two-stage stochastic programming framework, decomposes the fault problem into a main problem and corresponding sub-problems. The main problem does not consider N-1 constraints and is used to solve the economic solution. The sub-problems are used to perform N-1 fault verification on the economic solution after solving the main problem and identify faulty components. If there is a fault, the fault is used as a constraint of the main problem and the solution is re-solved. The solution process is iterated until the verification is error-free.

[0048] The present invention also provides an electronic device, including a memory and a processor, wherein the memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to implement the method.

[0049] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described thereon.

[0050] The beneficial effects of this invention are as follows:

[0051] (1) For the integrated collaborative planning of "grid-source-storage-load", the economic efficiency is significantly improved while ensuring N-1 safety. This invention constructs a unified optimization framework to jointly optimize the investment decisions of distribution network lines / substations with the site selection and capacity determination decisions of photovoltaic, energy storage and charging facilities. The core objective is to utilize the flexibility of distributed energy storage and controllable charging load, as well as the local power generation capacity of photovoltaics, to shape them into reliable backup resources for the system in the planning stage, thereby partially replacing expensive grid redundancy investments and minimizing the total cost of the system's entire life cycle while meeting the N-1 safety criterion.

[0052] (2) To ensure the robustness of the planning scheme in the face of multiple uncertainties during faults and to guarantee power supply security in extreme scenarios, a risk avoidance mechanism is embedded in the planning model to quantify and handle the uncertainties of load and renewable energy output during faults. The goal is to enable the final planning scheme to tolerate a wider range of uncertainty fluctuations under given investment cost constraints, ensuring that even in adverse fault scenarios, the system can still control load losses within an acceptable range by mobilizing resources such as photovoltaics, energy storage, and charging, thereby obtaining an economical and robust planning scheme.

[0053] (3) To overcome the computational bottleneck of N-1 security verification in large-scale distribution networks and reduce the computational load of complex operations, this invention designs an efficient decomposition and iterative verification strategy and a key component identification algorithm. By decomposing the complex overall problem into sub-problems for iterative solution, the direct handling of ultra-large-scale problems is avoided. The key lies in intelligently identifying the critical lines and nodes that have the most significant impact on system reliability, prioritizing their verification and reinforcement, thereby significantly reducing redundant scenarios requiring precise calculations. While ensuring verification accuracy, the computation time is shortened to an acceptable range for engineering, making this method applicable to the planning of actual large-scale distribution networks. Attached Figure Description

[0054] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0055] Figure 1 This is a flowchart of the present invention;

[0056] Figure 2 This is a schematic diagram of linearization constraints in an embodiment of the present invention;

[0057] Figure 3 This is a schematic diagram illustrating the energy flow within the integrated energy system and its interaction with the power distribution network in an embodiment of the present invention.

[0058] Figure 4 This is a complete flowchart of the improved N-1 verification algorithm based on key component identification in an embodiment of the present invention. Detailed Implementation

[0059] To enable those skilled in the art to better understand the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art should fall within the protection scope of the embodiments of this application.

[0060] like Figure 1 As shown, this invention provides a collaborative planning method for distribution networks and photovoltaic-storage-charging (PV-SGC-Charging) systems that considers the N-1 safety criterion. By constructing an integrated planning and verification framework that deeply integrates "grid, source, storage, and load," the method reduces the overall cost of the system's entire lifecycle and improves its robustness to multiple uncertainties at the time of failure, while ensuring that the distribution network meets the strict N-1 power supply safety requirement.

[0061] To achieve the above objectives, this method sequentially executes the following four core stages, and the technical means, implementation conditions, step relationships and functions of each stage are explained in detail below.

[0062] Phase 1: Phase angle relaxation is performed on the standard branch power flow model to obtain a real number equation model suitable for optimization calculation. The nonlinear constraints in the real number equation model are relaxed by second-order cone and linearized to construct a computable power flow model of the distribution network. Each photovoltaic power generation device-energy storage device-charging station is abstracted as an energy hub. The power from the distribution network and the power generated by photovoltaic power generation are used as inputs to construct a photovoltaic-storage-charging energy hub model.

[0063] The second stage involves constructing a two-stage stochastic programming framework for the two states of the photovoltaic-storage-charging system: normal operation or N-1 fault. The goal is to minimize the annualized total cost of the photovoltaic-storage-charging system, jointly optimizing investment and operation, and simulating N-1 fault recovery.

[0064] The third stage employs information gap decision theory to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure.

[0065] In the fourth stage, based on the aforementioned two-stage stochastic programming framework, the fault problem is decomposed into a main problem and corresponding sub-problems. The main problem does not consider N-1 constraints and is used to solve for an economic solution. The sub-problems are used to perform N-1 fault verification on the economic solution after solving the main problem and identify faulty components. If a fault is found, the fault is used as a constraint of the main problem and the solution is re-solved. The solution process is iterated until the verification is error-free.

[0066] In a preferred embodiment, the first stage is integrated collaborative planning mathematical modeling, which aims to establish an accurate and efficient mathematical model for subsequent optimization, including a computable power flow model of the distribution network and a photovoltaic-storage-charging energy hub model. Specifically, it includes the following processes:

[0067] 1.1. Construction of Computable Power Flow Model for Distribution Network

[0068] To address the challenges of non-convex and nonlinear AC power flow equations in distribution networks, this method employs an improved branch power flow model and implements a series of convex relaxation and linearization techniques.

[0069] Step 1.1.1: Establish a power flow model for relaxed phase branches

[0070] Conditions and Operations: Based on the standard AC branch power flow equations, neglecting voltage phase angle differences, the complex form of the power and voltage equations is transformed into real number equations containing only amplitude values. A new variable v is defined. i =|V i | 2 (Square of node voltage magnitude) and l ij =|I ij | 2 (Square of branch current amplitude). Eliminate complex variables and phase angle information in the model.

[0071] By relaxing the phase angle of the standard branch power flow model, a real-number equation model suitable for optimization calculation is obtained, as shown in Equation (1).

[0072] (1);

[0073] In the formula: This indicates the active power transmitted in the downstream branch; This indicates the active power transmitted by the upstream branch; This indicates the active power loss of the upstream branch; This refers to the reactive power loss of the upstream branch line.

[0074] Step 1.1.2: Perform second-order cone relaxation

[0075] By relaxing the nonlinear constraints in formula (1) into a second-order cone, we can transform it into a convex constraint form as shown in formula (2), thus converting the original problem into a mixed-integer second-order cone programming problem:

[0076] (2);

[0077] In the formula: Indicates a branch Reactive power transmitted upstream; This indicates the active power transmitted by the upstream branch; and These represent the squares of the complex magnitudes of the current and voltage variables, respectively.

[0078] By relaxing the non-convex constraint into a convex second-order cone constraint, the original problem is transformed into a mixed-integer second-order cone programming problem.

[0079] Step 1.1.3: Linearization of Line Capacity Constraints

[0080] The original line apparent power capacity constraint It is a circular feasible region. This method employs circumscribed polygon linearization techniques, for example, by approximating the circle with two squares rotated by 45 degrees, transforming it into a set of linear inequality constraints, such as... Figure 2 As shown.

[0081] By fully linearizing the nonlinear constraints, the model complexity is further reduced, and the stability and speed of solving large-scale problems are greatly improved.

[0082] 1.2. Construction of Energy Hub Model for Photovoltaic-Energy Storage-Charging System

[0083] To uniformly describe the coupling and conversion of multiple energy flows such as light, storage, and charging, this method adopts a static energy hub model.

[0084] Step 1.2.1: Define the system structure and coupling matrix

[0085] Conditions and Operations: Each photovoltaic-storage-charging station is abstracted as an energy hub. Its inputs include electrical energy from the distribution network and photovoltaic power generation; its outputs include electrical energy supplied to charging piles, electrical energy stored in batteries, and possible reverse grid transmission. The coupling relationship is described by a conversion efficiency matrix.

[0086] Step 1.2.2: Establish a photovoltaic-storage-charging energy hub model

[0087] Photovoltaic model: An output model based on a typical solar radiation curve is adopted. ,in, Light intensity, For the area of ​​the photovoltaic panel, For conversion efficiency.

[0088] Electrochemical energy storage model: considering charge and discharge efficiency , and energy state The constraints include: In addition, there are upper and lower limits for power and capacity.

[0089] Electric vehicle charging load model: aggregated into adjustable load, total power Constrained by the number of charging piles, power level, and user charging demand curve, it can participate in demand response.

[0090] The photovoltaic model, electrochemical energy storage model, and electric vehicle charging load model mentioned above together constitute the photovoltaic-storage-charging energy hub model.

[0091] The photovoltaic-storage-charging energy hub model clearly depicts the flow path and conversion relationship of energy between "photovoltaic power generation - battery storage - electric vehicle consumption". For example... Figure 3 As shown, this illustrates the energy flow within a user-level integrated energy system and its interaction with the power distribution network in one embodiment of the present invention.

[0092] As a preferred embodiment, the second stage is a two-stage stochastic collaborative programming framework. This stage is the core of the planning decision and adopts a two-stage stochastic programming framework, which corresponds to the two states of "normal operation" and "N-1 failure" of the system.

[0093] 2.1. "Here-and-Now" Phase: Joint Optimization of Investment and Operation

[0094] This stage involves making "one-time decisions" before the uncertainty (specific fault location) is revealed. This includes 0-1 construction variables for distribution network lines / substations to be built, and site selection (0-1 variable) and capacity determination (continuous variable) decisions for photovoltaic, energy storage, and charging equipment.

[0095] The objective function is to minimize the annualized total cost of the system, including investment cost and expected operating cost.

[0096] Investment cost: Annual investment cost of equipment such as power lines, substations, photovoltaic panels, energy storage batteries, and charging piles.

[0097] Expected operating costs: Costs of purchasing electricity from the main grid, photovoltaic operation and maintenance costs, and potential costs of purchasing gas / heat (if the system includes combined heat and power, etc.).

[0098] The constraints are the convex power flow constraints of the distribution network, the operation constraints of the photovoltaic-storage-charging system, and the system power balance constraints established in the first phase of integration. At this point, the N-1 security constraint has not yet been directly introduced.

[0099] The investment and operation model constructed in this stage (here-and-now) aims to minimize the annualized total cost of the system. The objective function is shown in formula (4), which includes investment cost and expected operating cost. The specific calculation is shown in formulas (5) and (6).

[0100] Since the integrated energy system includes photovoltaic and energy storage, the two-stage stochastic programming framework takes the minimum annualized total cost of the distribution network-user integrated energy system as the planning objective, and the objective function is shown in equation (3).

[0101] (3);

[0102] In the formula: The annualized total equipment investment cost for the entire integrated energy system of the distribution network and user level; The annualized total operating cost of the entire integrated energy system at the distribution network and user level.

[0103] The composition of annualized total equipment investment cost and annualized total operating cost is shown in equations (4) and (5).

[0104] (4);

[0105] (5);

[0106] In the formula: This is the annualized factor for investment costs; This is the annualized factor for operating costs; , , , , , , , , These are the unit investment costs for transmission lines, substations, photovoltaic systems, combined heat and power units, electric boilers, electric-to-gas conversion equipment, energy storage batteries, gas storage tanks, and thermal storage tanks, respectively. For 0-1 variables, Indicates the transmission line on the branch Construction in the area This indicates that the transmission line is not on the branch. Construction in the area; , , , , , , , Substations, photovoltaic power plants, combined heat and power units, electric boilers, electric-to-gas conversion equipment, energy storage batteries, gas storage tanks, and thermal storage tanks are located at the nodes. Investment capacity at the location; For distribution networks in scenarios Time period The unit cost of purchasing electricity from the superior power grid; For user-level integrated energy systems in various scenarios Time period The unit cost of purchasing gas from energy wholesalers; For user-level integrated energy systems in various scenarios Time period The unit cost of purchasing heat from energy wholesalers; For the scene Time period node Active power of the substation; For the scene Time period node Gas purchase capacity of the user-level integrated energy system; For the scene Time period node The heat purchase capacity of the user-level integrated energy system; This represents the time interval for the random scene.

[0107] The specific calculation methods for the annualized investment cost factor and the annualized operating cost factor are shown in equations (6) and (7):

[0108] (6);

[0109] (7);

[0110] In the formula: The discount rate; The expected service life of the equipment; The number of random scenes.

[0111] 2.2. "Wait-and-See" Phase: N-1 Fault Recovery Simulation and Resilience Verification

[0112] like Figure 4 As shown, this stage simulates the response of the photovoltaic energy storage and charging system to any preset N-1 fault (such as a single line disconnection or a single substation outage) after the "Here-and-Now" stage scheme is determined.

[0113] The decision variables are the network reconfiguration switch status after a fault, the emergency discharge power of energy storage, the active / reactive power support of photovoltaics, and the adjustment amount of interruptible / transferable charging load.

[0114] The objective at this point is to minimize the total load reduction (i.e., the amount of power wasted) under this specific fault scenario.

[0115] Constraints: Based on the equipment capacity and network topology determined in the "Here-and-Now" phase, and considering the new topology after a fault, the distribution network power flow constraints and photovoltaic-storage-charging operation constraints are reapplied. The critical constraint is that non-critical loads can be reduced.

[0116] At this stage, the ability of the preliminary plan to cope with a single failure is tested. If zero load reduction cannot be achieved or the reduction exceeds the limit under a certain failure scenario, it means that the plan does not meet the N-1 safety criterion, and this information needs to be fed back to the first stage.

[0117] Based on the fault verification results, new constraints are added to the investment optimization model (first stage) to correct the core iterative mechanism of the planning scheme.

[0118] The wait-and-see model aims to minimize load loss during a failure, and its objective function is shown in equation (7):

[0119] (8);

[0120] In the formula: For distribution networks in scenarios Time period node The amount of active power wasted at the location; For user-level integrated energy systems in various scenarios Time period node The amount of active power wasted at the location.

[0121] As a preferred embodiment, the third stage is risk aversion enhancement based on information gap decision theory. To address the profound uncertainty of drastic fluctuations in photovoltaic output and charging load during fault times, this method embeds information gap decision theory into the second-stage framework. The specific steps are as follows:

[0122] 3.1. Defining the Uncertainty Model

[0123] Instead of making probability distribution assumptions about uncertainty, we define its fluctuation range. For example, we set the load at the time of failure. and photovoltaic output Fluctuating around the predicted value: ,in The fluctuation deviation coefficient is unknown.

[0124] 3.2. Constructing a risk aversion model

[0125] An IGDT risk aversion strategy is adopted. In the "Here-and-Now" phase, the objective function is changed from "minimizing cost" to "maximizing volatility deviation coefficient". "and add a cost constraint: total cost" .in The optimal cost under the uncertainty model. The percentage of additional costs that decision-makers are willing to bear to improve robustness (e.g., 3%).

[0126] During the "Wait-and-See" phase, load and photovoltaic output are substituted into its fluctuation range. The worst-case scenario (the upper limit is used for load and the lower limit is used for photovoltaic) is verified.

[0127] This mechanism transforms the planning scheme from a "point-optimal solution" into a "robust decision." It explicitly answers the question of "how much adverse fluctuation the planning scheme can withstand," ensuring that the scheme can maintain power supply capacity even under extreme failure scenarios.

[0128] As a preferred embodiment, the fourth stage is a fast N-1 verification strategy based on decomposition iteration and key identification. To solve the computational bottleneck of full-scenario enumeration verification, this method designs a master-sub-problem decomposition and intelligent scene screening mechanism.

[0129] 4.1 Problem Breakdown

[0130] The problem of failure is broken down into a main problem and its corresponding sub-problems.

[0131] Main problem: Integrate a "Here-and-Now" investment operation model with a dynamically growing set of "critical N-1 failure scenario constraints" (initially empty) to generate planning solutions.

[0132] Sub-problem: The second phase of the "Wait-and-See" fault recovery model is used to verify whether the current solution to the main problem passes a specific N-1 fault scenario.

[0133] 4.2. Identification of Critical Faulty Components

[0134] Conditions and Operations: After each main problem is solved, based on the current network topology and power flow distribution, graph theory and electrical sensitivity analysis are applied to automatically identify the "critical branches" and "critical substations" most likely to cause severe power supply bottlenecks. Identification criteria include: whether it is the only tie line in the network, whether it carries significant power transmission, and whether its disconnection would cause a large number of loads to lose power.

[0135] 4.3 Iterative Solution Process

[0136] Initialization: Without considering any N-1 constraints in the main problem, the solution yields an economically optimal initial solution.

[0137] Key scenario verification: Substitute the current solution into the sub-problem and prioritize verifying the key component failure scenarios identified in step 4.2.

[0138] Constraint generation and feedback: If a critical failure scenario fails the verification (load reduction occurs), the recovery feasibility constraint (i.e. "Benders") under that scenario is added to the critical N-1 failure scenario constraint set of the main problem.

[0139] Resolving the main problem: The main problem is solved again under the new constraints to obtain a revised planning scheme that takes into account the discovered critical faults.

[0140] Looping and Convergence: Repeat steps one and two until all critical fault scenarios pass verification. Subsequently, the remaining non-critical faults can be sampled or batch verified until all N-1 scenarios meet the requirements or the computation time limit is reached.

[0141] This strategy avoids the cumbersome enumeration and verification of hundreds or thousands of failure scenarios, and concentrates computing power on the weakest link most likely to cause problems. It quickly obtains a feasible solution that meets the N-1 safety criterion through a limited number of iterations, and the computational efficiency is improved by more than an order of magnitude compared with traditional methods.

[0142] This invention forms a complete and feasible collaborative planning method for power distribution networks and photovoltaic-storage-charging systems through the synergistic effect of four major technical means: integrated modeling, two-stage planning, robustness enhancement, and efficient verification.

[0143] This invention also provides a collaborative planning device for distribution networks and photovoltaic-storage-charging systems that considers the N-1 safety criterion, including a computable power flow model construction module for distribution networks, a photovoltaic-storage-charging energy hub model construction module, a two-stage stochastic programming framework construction module, a risk avoidance module, and a problem decomposition and iterative verification module.

[0144] The specific implementation process of the functions and roles of each module in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0145] The present invention also provides an electronic device, including a memory and a processor, wherein the memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to implement the method.

[0146] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described thereon.

[0147] The computer-readable storage medium can be an internal storage unit of any data processing device described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be any data processing device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data processing device, and can also be used to temporarily store data that has been output or will be output.

[0148] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only.

[0149] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A collaborative planning method for distribution networks and photovoltaic-storage-charging systems considering the N-1 safety criterion, characterized in that, Includes the following steps: Phase angle relaxation is applied to the standard branch power flow model to obtain a real equation model suitable for optimization calculation. The nonlinear constraints in the real equation model are relaxed by second-order cone relaxation and linearized to construct a computable power flow model for the distribution network. Each photovoltaic power generation device-energy storage device-charging station is abstracted as an energy hub, and the power from the distribution network and the power generated by photovoltaic power generation are used as inputs to construct a photovoltaic-storage-charging energy hub model; For the two states of the photovoltaic-storage-charging system, namely normal operation or N-1 fault, a two-stage stochastic programming framework is constructed. With the goal of minimizing the annualized total cost of the photovoltaic-storage-charging system, investment and operation are jointly optimized, and N-1 fault recovery simulation is performed. The information gap decision theory is adopted to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure; Based on the aforementioned two-stage stochastic programming framework, the fault problem is decomposed into a main problem and corresponding sub-problems. The main problem does not consider N-1 constraints and is used to solve for an economic solution. The sub-problems are used to perform N-1 fault verification on the economic solution after solving the main problem, identify faulty components, and if a fault is found, it is used as a constraint of the main problem and the solution is re-solved. The solution process is iterated until the verification is correct.

2. The method according to claim 1, characterized in that, The linearization is achieved through the circumscribed polygon linearization technique.

3. The method according to claim 1, characterized in that, The photovoltaic energy storage and charging hub model includes: A photovoltaic model, based on a typical solar radiation curve, represents the output characteristics of a photovoltaic system; An electrochemical energy storage model is used to track changes in the energy state of energy storage devices, and is constrained by charge and discharge efficiency, power, energy state, and upper and lower limits of capacity. The electric vehicle charging load model is used to aggregate distributed electric vehicle charging loads into a controllable overall load, which is constrained by the number of charging piles, power level, and user charging demand curve.

4. The method according to claim 1, characterized in that, The joint optimization of investment and operation, with the goal of minimizing the annualized total cost of the photovoltaic-storage-charging system, includes: An investment and operation model is constructed with minimizing the annualized comprehensive cost of the photovoltaic-storage-charging system as the objective function. The annualized comprehensive cost includes the investment cost and the expected operating cost. The constraints include convex power flow constraints of the distribution network, operational constraints of the photovoltaic-storage-charging system, and power balance constraints of the photovoltaic-storage-charging system.

5. The method according to claim 1, characterized in that, The N-1 fault recovery simulation includes: The response of a simulated photovoltaic energy storage and charging system when it encounters any preset N-1 fault; The decision variables are the network reconfiguration switch status after a fault, the emergency discharge power of the energy storage device, the active and reactive power support power of photovoltaics, and the adjustment amount of interruptible or transferable charging load. The objective function is to minimize the total load reduction under this specific N-1 fault scenario. The optimization solution is then implemented, and power flow constraints of the distribution network and operation constraints of the photovoltaic-storage-charging system based on the new topology after the fault are introduced. Non-critical loads are also allowed to be reduced. If the total load reduction exceeds a preset value after verification and optimization, the current planning framework is determined not to meet the N-1 safety criterion, and feedback information is generated.

6. The method according to claim 1, characterized in that, The method employing information gap decision theory to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of fault includes: Based on the information gap decision theory, an uncertainty model is constructed to address the uncertainty of photovoltaic power output and charging load fluctuations at the time of failure, and the uncertainty of photovoltaic power output and charging load fluctuations is represented by the fluctuation deviation coefficient. The objective function is to replace minimizing the annualized total cost of the photovoltaic-storage-charging system with maximizing the fluctuation deviation coefficient, while adding cost constraints.

7. The method according to claim 1, characterized in that, The process of breaking down the fault problem into a main problem and corresponding sub-problems includes: The main problem includes a set of key N-1 fault scenarios. The set of key N-1 fault scenarios is initially empty. Whenever a subproblem is solved and an N-1 fault scenario is obtained, the N-1 fault scenario is added to the set of key N-1 fault scenarios.

8. A collaborative planning device for distribution network and photovoltaic-storage-charging system considering the N-1 safety criterion, characterized in that, include: The computable power flow model construction module for distribution networks is used to relax the phase angle of the standard branch power flow model to obtain a real equation model suitable for optimization calculation. The nonlinear constraints in the real equation model are relaxed by second-order cone and linearized to construct a computable power flow model for distribution networks. The photovoltaic-storage-charging energy hub model construction module is used to abstract each photovoltaic power generation device-energy storage device-charging station into an energy hub, and uses the power from the distribution network and the power generated by photovoltaic power generation as inputs to construct the photovoltaic-storage-charging energy hub model; The two-stage stochastic programming framework construction module constructs a two-stage stochastic programming framework for two states of the photovoltaic storage and charging system: normal operation or N-1 fault. The framework aims to minimize the annualized total cost of the photovoltaic storage and charging system, jointly optimizes investment and operation, and performs N-1 fault recovery simulation. The risk avoidance module is used to maximize the uncertainty range of photovoltaic power output and charging load fluctuations at the time of failure, based on the information gap decision theory. The problem decomposition and iterative verification module, based on the two-stage stochastic programming framework, decomposes the fault problem into a main problem and corresponding sub-problems. The main problem does not consider N-1 constraints and is used to solve the economic solution. The sub-problems are used to perform N-1 fault verification on the economic solution after solving the main problem and identify faulty components. If there is a fault, the fault is used as a constraint of the main problem and the solution is re-solved. The solution process is iterated until the verification is error-free.

9. An electronic device comprising a memory and a processor, characterized in that, The memory is coupled to the processor; wherein the memory is used to store program data, and the processor is used to execute the program data to implement the method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images