A power distribution network distributed mutual aid collaborative optimization method for feeder autonomy

Abstract

The present application belongs to the field of optimal operation of power distribution network, and provides a kind of power distribution network distributed mutual aid collaborative optimization control method for feeder autonomy, comprising: (1) determining the autonomy of feeder and the power shortage of feeder autonomy;(2) construct the Hausdorff distance-based feeder autonomy power shortage uncertainty set;(3) construct a multi-objective control optimization model for power distribution network oriented to feeder autonomy;(4) construct a two-stage distributed robust optimization model considering the power shortage of feeder autonomy;(5) construct a distributed solution framework, solve the distributed coordination optimization model between the feeders of power distribution network and the autonomous robust optimization model within the feeder, and obtain the distributed mutual aid collaborative optimization operation scheme of power distribution network.The method of the present application can effectively realize the distributed coordination of power coordination and mutual aid between feeders and the optimal operation within the feeder, and further achieve the collaborative optimization of minimizing the distributed mutual aid control cost and feeder operation cost of power distribution network.

Description

Technical Field

[0001] This invention belongs to the field of distribution network optimization operation, specifically relating to a distributed mutual assistance and collaborative optimization control method for distribution networks oriented towards feeder autonomy. Background Technology

[0002] In recent years, with the large-scale integration of high-proportion distributed photovoltaic (PV) power and the influx of emerging and diversified carriers such as adjustable loads, electricity substitution, electric vehicles, and energy storage, the distribution network has gradually transformed from a traditional unidirectional power receiving network to an active network with bidirectional power flow. This transformation places increasingly higher demands on the coordinated regulation and flexible operation of the distribution network. However, both distributed PV output and user-side load forecasting exhibit significant randomness and volatility. This dual uncertainty of source and load can easily lead to localized power deficits in the distribution network, resulting in daytime curtailment of solar power or insufficient power during specific periods.

[0003] Furthermore, when dealing with source-load uncertainty, traditional distribution network optimization operation models often rely on accurate forecast data or overly conservative robust models. In actual operation, due to the bias in the scenario probability distribution obtained after clustering historical scenarios, existing technologies struggle to accurately and effectively characterize the true differences in the source-load uncertainty set, resulting in insufficient robustness of the optimization model when considering technical indicators.

[0004] Currently, research on novel load control strategies for different granularities, including 10kV feeders, substations, county-level areas, and the entire region, is insufficient. Regarding the refined autonomous operation of distribution networks, most control strategies lack quantitative analysis techniques for the autonomous capabilities and mutual assistance control at the individual feeder level, making it difficult to accurately quantify the gap between the feeder's power regulation capacity and autonomous power demand.

[0005] Regarding multi-objective coordination and solution mechanisms, existing optimization strategies for local heavy-load distribution areas or lines in the power grid are often limited to a single economic indicator, neglecting the consideration of the autonomous capabilities of local feeders in the distribution network. Research on the economics and market mechanisms of mutual assistance regulation is still lacking.

[0006] In summary, existing technologies still have significant shortcomings in quantitative analysis of feeder-level mutual assistance and control capabilities, construction of highly robust uncertainty sets based on historical data, and distributed multi-objective collaborative optimization that balances economy and feeder autonomy, making it difficult to meet the needs of high-level autonomy and mutual assistance optimization in distribution networks. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and propose a distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy. This method can effectively cope with the source-load uncertainty caused by high proportion of distributed photovoltaic and multi-load access, realize distributed coordination and mutual assistance between feeders and optimized operation within feeders, and thus achieve coordinated optimization to minimize the distributed mutual assistance control cost and feeder operation cost of the distribution network.

[0008] To achieve the above objectives, the present invention adopts the following technical solution.

[0009] A distributed mutual assistance and collaborative optimization control method for distribution networks oriented towards feeder autonomy includes the following steps:

[0010] (1) Calculate the mutual assistance and regulation capability for autonomous feeder in distribution network, and determine the autonomous capability and power deficit of feeder;

[0011] (2) Using historical source-load data of the distribution network, construct an uncertain set of autonomous power deficit for feeders based on Hausdorff distance, which will be used to establish a sub-Bruker optimization model in the future;

[0012] (3) With the goal of minimizing the cost of 24-hour distributed mutual assistance regulation and control of the distribution network and the cost of feeder operation, and introducing autonomous power deficit cost into the feeder operation cost, a multi-objective regulation optimization model for distribution network oriented towards feeder autonomy is constructed.

[0013] (4) Based on the multi-objective control optimization model of the distribution network oriented towards feeder autonomy, a two-stage sub-Blu-ray bar optimization model considering the power deficit of feeder autonomy is constructed;

[0014] (5) Construct a distributed solution framework to solve the distributed coordination optimization model between feeders and the autonomous robust optimization model within feeders of the distribution network, and obtain the distributed mutual assistance and coordinated optimization operation scheme of the distribution network.

[0015] Compared with existing technologies, this invention has better performance. In the distribution network optimization operation model, this invention considers both operational economy and feeder autonomy. At the same time, it constructs an uncertain set of feeder autonomous power deficit based on Hausdorff distance based on historical data of the distribution network. While taking into account technical indicators, it effectively improves the robustness of the model. The final optimized operation scheme achieves economical operation of the distribution network while effectively improving the feeder autonomy of the distribution network, thus realizing multi-objective optimization of the distribution network. Attached Figure Description

[0016] Figure 1 This is a flowchart of a method according to an embodiment of the present invention.

[0017] Figure 2 This is a schematic diagram of an autonomous unit based on a distribution network feeder.

[0018] Figure 3 These are the photovoltaic prediction curves for each feeder of the distribution network in this embodiment.

[0019] Figure 4 These are the load prediction curves for each feeder of the distribution network in this embodiment.

[0020] Figure 5 This is a power interaction diagram between feeders after feeder mutual assistance in this embodiment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0022] like Figure 1 As shown, this embodiment provides a distributed mutual assistance and collaborative optimization control method for distribution networks oriented towards feeder autonomy, including the following steps:

[0023] Step 1: Calculate the mutual assistance and control capability for autonomous feeder in distribution network, and determine the autonomous capability and power deficit of feeder.

[0024] Based on the distribution network feeder structure, and considering the power mutual assistance among distribution network feeders to enhance the self-governance capability of the distribution network, each feeder is considered an autonomous unit. A schematic diagram of an autonomous unit based on the distribution network feeder is shown below. Figure 2 As shown.

[0025] The resources within each feeder in the above zoning diagram include energy storage and controllable distributed power sources. The feeder power supply capacity is as follows:

[0026]

[0027] In the formula: Let t be the upward power supply capability of the feeder. This refers to the downward power supply capability. and This indicates the upward and downward power supply capacity of the energy storage at time t. , This represents the upward and downward power supply capacity of a controllable distributed power source at time t.

[0028] The feeder autonomous power demand generated by photovoltaics and loads within the feeder is shown below:

[0029]

[0030] In the formula: and These represent the power demands for upward and downward loads, respectively. , These represent the upward and downward power demands of photovoltaic power, respectively. Let t be the upward power demand generated by the source load fluctuation at time t. This represents a downward power demand.

[0031] Based on the above feeder power regulation capability and feeder autonomous power requirements, the upper and lower bounds of the feeder power autonomous range can be derived as follows:

[0032]

[0033] The gap between feeder autonomous power demand and feeder power supply capacity is defined as the feeder autonomous power deficit, and its upward and downward feeder autonomous power deficits are shown below:

[0034]

[0035] In the formula: , This is for the autonomous power deficit of the upward and downward feeders.

[0036] In this embodiment, the ratio of the feeder autonomous power deficit to the power balance interval of each feeder autonomously within the scheduling period is defined as the feeder power autonomy capability, and the calculation is as follows:

[0037]

[0038] In the formula: The autonomous capability of the i-th feeder is calculated between [0,1]. The closer it is to 1, the stronger the power autonomy of the feeder. T is the scheduling period.

[0039] Step 2: Construct an uncertain set of autonomous power deficit for feeders based on Hausdorff distance using historical source-load data of the distribution network, which will be used to establish a distributed bar optimization model in the future.

[0040] With a high proportion of distributed photovoltaic (PV) power connected to the distribution network, both PV output and load forecasting exhibit significant randomness. This dual uncertainty of source and load leads to a power deficit in the autonomous distribution network. This embodiment uses a probability distribution fuzzy set to describe the power deficit in feeder autonomy caused by source and load uncertainty. First, M sets of feeder autonomous power deficit scenarios are extracted from the historical data of PV and load for each feeder. The scenarios are then reduced using the K-means clustering algorithm to obtain N typical feeder autonomous power deficit scenarios. By permuting and combining these scenarios, a set of combined typical scenarios is generated, and its probability distribution is determined.

[0041]

[0042] In the formula: The number of combined scenes, Let be the probability of the combined scenario s occurring. , These represent the number of typical photovoltaic scenarios and the scenario probabilities in the combined scenario s. , These represent the typical number of load scenarios and the scenario probabilities in the combined scenario s, respectively.

[0043] Due to the uncertainty of typical scenarios of autonomous power deficit in feeders, the scenario probability distribution obtained after clustering has a deviation. In this embodiment, the Hausdorff distance is used to quantify the degree of deviation of the probability distribution. Since the Hausdorff distance is the maximum minimum distance between two point sets, it is extremely sensitive to outliers. In order to avoid interference from extreme outliers, the average value is used instead of the maximum value to calculate the improved Hausdorff distance constraint, so as to more accurately characterize the differences between sets.

[0044]

[0045] In the formula: , These are the number of elements in point set A and point set B, respectively; Given the one-way Hausdorff distance from A to B, calculate the distance from any point in A. Find the minimum distance to q points in B, obtain the q minimum values, and calculate their average; similarly, the one-way Hausdorff distance from B to A can be obtained. Improved Hausdorff distance for sets A and B That is, the larger of the two.

[0046] Based on the initial probability distribution of the generated combined typical scenario set, and with the Hausdorff distance as a constraint, an uncertainty set for feeder autonomous power deficit scenarios is constructed, the mathematical expression of which is shown below:

[0047]

[0048] In the formula: These are the probability values ​​of scene s based on the Hausdorff distance, and the Hausdorff distance constraint is... ,in This represents the probability tolerance value.

[0049] To control the fluctuation range of the probability distribution of combined scenarios, the probability value of scenario s must meet the preset confidence requirement. The constraint relationship can be expressed as:

[0050]

[0051] In the formula: K and M represent the number of typical and historical scenarios of autonomous power deficit for feeders, respectively. , where is the confidence level.

[0052] Step 3: With the goal of minimizing the 24-hour distributed mutual assistance control cost and feeder operation cost of the distribution network, and introducing autonomous power deficit cost into the feeder operation cost, construct a multi-objective control optimization model for distribution network oriented towards feeder autonomy.

[0053] Distribution network optimization aims to minimize the total cost within a given period, including power exchange costs between feeders and the operating costs of each feeder. By considering the autonomous power deficit costs within each feeder, it achieves both economic optimization of the distribution network and enhancement of feeder autonomy, thus realizing synergistic optimization of economic efficiency and feeder autonomy, ultimately reducing the total operating cost. Specifically, it is expressed as follows:

[0054]

[0055] In the formula: , These are the operating cost and power interaction cost of the i-feeder, respectively.

[0056] When optimizing the dispatching of a distribution network considering the autonomous power deficit of feeders after zoning, the power exchange between feeders is based on their initial operating state. The coordination and mutual assistance cost between feeders is:

[0057]

[0058] In the formula: Let be the power output of feeder i to other feeders at time t. Let be the power input from other feeders to feeder i at time t. This is the exchange cost coefficient.

[0059] After the initial power exchange between each feeder, each feeder solves in parallel based on the initial power exchange results to generate the optimal scheduling strategy. The internal operating cost of each feeder is:

[0060]

[0061] In the formula: For feeder autonomous power deficit costs, , , and These are the costs of controllable distributed power generation within the feeder, energy storage operation costs, transaction costs with the main grid, and curtailment penalty costs.

[0062] (1) Global constraints between feeders

[0063] When feeders interact, the interaction power between feeders must meet the balance constraint.

[0064]

[0065] In the formula: , These represent the active power transmitted and received by the i-th feeder to other feeders at time t.

[0066] (2) Feeder autonomous power deficit cost

[0067] The feeder autonomous power deficit cost is expressed as the sum of the upward and downward deficit costs.

[0068]

[0069] In the formula: This is the feeder autonomous power deficit cost coefficient.

[0070] (3) Operating costs and constraints of controllable distributed power sources

[0071]

[0072] In the formula: , and The upper and lower limits of the output power and the maximum ramp power of the controllable distributed power source; Let be the power generation at time t, and let the power generation cost coefficient be... , and .

[0073] (4) Operating costs and constraints of energy storage systems

[0074]

[0075] In the formula: , Let be the charging power and discharging power of the stored energy at time t, respectively. This represents the unit loss cost coefficient for energy storage. For charge and discharge efficiency; , These are the maximum charging and discharging power of the energy storage; To represent the charging and discharging states of energy storage respectively; , Let be the energy storage capacity at time t and its rated capacity, respectively. , Indicates the maximum and minimum states of charge.

[0076] (5) Transaction costs and constraints between feeder and main network

[0077]

[0078] In the formula: , These represent the power exchange between the feeder and the main grid at time t. Time-of-use pricing for purchasing and selling electricity from the grid. For electricity purchase and sale; This represents the maximum transmission power of the tie line.

[0079] (6) Distributed photovoltaic curtailment penalty costs and constraints

[0080]

[0081] In the formula: , These represent the actual and available photovoltaic output at time t, respectively. The penalty cost coefficient for abandoning light; Let t be the available distributed photovoltaic output of the i-th feeder.

[0082] (7) Power balance constraints inside the feeder

[0083]

[0084] In the formula: , They are respectively User load and distributed photovoltaic output of feeder i at any time.

[0085] Step 4: Construct a two-stage sub-Blule bar optimization model that considers the autonomous power deficit of the feeder.

[0086] Based on the multi-objective control optimization model for distribution networks oriented towards feeder autonomy constructed in step 3 above, a two-stage partial Bruker optimization model considering the power deficit of feeder autonomy is constructed, with the objective function shown below:

[0087]

[0088] In the formula: N is the number of feed lines; , The pre-control stage includes the power exchange cost between feeders and their respective operating costs. The operating costs for each feeder during the real-time control phase; This refers to the power interaction value between feeders; , The variables are divided into first-stage and second-stage variables. The first-stage variables include the power purchase and sale plan, the start-up and shutdown of generating units, and the charging and discharging status of energy storage. The second-stage variables include the output of elements inside each feeder and the amount of curtailed solar power. These are random variables, including the output values ​​of photovoltaic power and load. This is the uncertain set of autonomous power deficit for the feeder.

[0089] Constraints can be abstractly represented in the following form:

[0090]

[0091] In the formula: the constraints can be divided into two categories: internal constraints of the feeder and mutual constraints between feeders. The first six lines define the internal constraints of the feeder, and the seventh line specifies the mutual constraints between feeders.

[0092] Step 5: Construct a distributed solution framework to solve the distributed coordination optimization model between distribution network feeders and the autonomous robust optimization model within feeders, and obtain the distributed mutual assistance and coordinated optimization operation scheme of the distribution network.

[0093] A distributed framework combining the Alternating Direction Multiplier (ADMM) method and the Column Constraint Generation (CCG) algorithm is used to solve the problem.

[0094] For the distributed coordination optimization model between distribution network feeders, the ADMM algorithm is used to solve the problem, simplifying the two-stage distributed bar optimization model into the standard form of ADMM:

[0095]

[0096] In the formula: x and z are the original problem variables, and the function and The corresponding power exchange costs between feeders and feeder operating costs The constraints correspond to the global constraints of the feeder, and A, B, and C are simplified coupling matrices.

[0097] To address the coordination optimization problem between feeders, a boundary variable interaction mechanism is employed to achieve distributed solution. To effectively handle the coordination between constraints and the optimization objective, Lagrange multipliers and penalty terms are introduced, integrating the original problem's constraints into the objective function to form a unified optimization model. Its specific mathematical expression is shown below:

[0098]

[0099] In the formula: , Let be the Lagrange multipliers and the penalty parameter, and .

[0100] Finally, the threshold values ​​for the incremental changes of the dual and original variables are used as the convergence criteria after solving the problem. When the iteration update amount is lower than the set tolerance, the algorithm terminates and outputs the optimal solution. The convergence criteria are as follows:

[0101]

[0102] In the formula: , Let be the original residual and the dual residual, respectively, and k be the number of iterations. The algorithm terminates and outputs the optimal solution when both the original residual and the dual residual satisfy the preset convergence condition.

[0103] For the autonomous robust optimization model within the feeder, the optimization model within each feeder can be solved in parallel. The CCG algorithm is introduced to solve the sub-robust optimization model within each feeder. First, the original problem is decomposed into a main problem and sub-problems, as shown below.

[0104]

[0105] In the formula: n is the number of iterations of the CCG algorithm.

[0106] After decomposing the above-mentioned sub-Bluhl model, the main problem first obtains the optimal solution for operating cost based on the initial worst-case scenario probability distribution, solving for the first-stage decision variables, including the power purchase and sale plans of each feeder and the main grid, the start-up and shutdown of controllable distributed power sources, and the charging and discharging status of energy storage, and updating the lower bound LB. The sub-problems first solve the min problem to obtain the cost-optimal operating scheme under each scenario, thus obtaining the second-stage decision variables, the power output plan, including the power of controllable distributed power sources, the charging and discharging power of energy storage, and the amount of curtailed solar power. Then, the max problem is run to obtain the worst-case probability distribution corresponding to the maximum autonomous power deficit of the feeder, and the upper bound UB is updated. Finally, the solution is completed when the upper and lower bounds converge to less than the preset precision. The optimization model within each feeder is a sub-Bluhl model, and each robust model can be solved in parallel.

[0107] Photovoltaic and load forecast curves for each feeder of the distribution network are as follows: Figure 3 , Figure 4 As shown in Table 1, the overall autonomous capability before feeder mutual assistance and the autonomous capability of each feeder after feeder mutual assistance and coordination are compared. The autonomous capability of each feeder after mutual assistance and coordination is improved.

[0108] Table 1. Autonomous Capacity of Distribution Network Feeder Interconnection Before and After

[0109]

[0110] To further illustrate the impact of feeder mutual assistance and coordination on feeder autonomy in the distribution network, the power interaction of each feeder after feeder mutual assistance is analyzed, such as... Figure 5 As shown, a positive interaction power indicates that the feeder supplies power to other feeders, while a negative power indicates that it receives power from other feeders.

[0111] Based on the power interaction results of each feeder, the photovoltaic installed capacity of each feeder, and the load curve, the photovoltaic installed capacity of feeder 1 is the smallest and is insufficient to support the load during the day. The photovoltaic installed capacity of feeders 2 and 3 is larger, and curtailment will occur during the day. However, by coordinating the power coordination and mutual assistance among feeders, distributed photovoltaic and energy storage resources in the region can be effectively integrated, the autonomy of each feeder can be improved, and cross-regional power complementarity and optimized operation can be achieved.

[0112] The contents not described in detail in this specification are existing technologies known to those skilled in the art.

[0113] The above description is merely a specific embodiment of the present invention, but the scope of application of the present invention is not limited thereto. Any other implementation methods derived by those skilled in the art based on the technical solutions of the present invention are also within the scope of protection of the present invention.

Claims

1. A distributed mutual assistance and collaborative optimization control method for distribution networks oriented towards feeder autonomy, characterized in that... The method includes the following steps: (1) Calculate the mutual assistance and regulation capability for autonomous feeder in distribution network, and determine the autonomous capability and power deficit of feeder; (2) Using historical source-load data of the distribution network, construct an uncertain set of autonomous power deficit for feeders based on Hausdorff distance, which will be used to establish a sub-Bruker optimization model in the future; (3) With the goal of minimizing the cost of 24-hour distributed mutual assistance regulation and control of the distribution network and the cost of feeder operation, and introducing autonomous power deficit cost into the feeder operation cost, a multi-objective regulation optimization model for distribution network oriented towards feeder autonomy is constructed. (4) Based on the multi-objective control optimization model of the distribution network oriented towards feeder autonomy, a two-stage sub-Blu-ray bar optimization model considering the power deficit of feeder autonomy is constructed; (5) Construct a distributed solution framework to solve the distributed coordination optimization model between feeders and the autonomous robust optimization model within feeders of the distribution network, and obtain the distributed mutual assistance and coordinated optimization operation scheme of the distribution network.

2. The distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy as described in claim 1, characterized in that: In step (1), the resources within each feeder of the distribution network include energy storage and controllable distributed power sources, and the feeder power supply capacity is: In the formula: Let t be the upward power supply capability of the feeder. This refers to the downward power supply capability. and This indicates the upward and downward power supply capacity of the energy storage at time t. , This represents the upward and downward power supply capacity of a controllable distributed power source at time t. The feeder autonomous power demand generated by photovoltaics and loads within the feeder is expressed as follows: In the formula: and These represent the power demands for upward and downward loads, respectively. , These represent the upward and downward power demands of photovoltaic power, respectively. Let t be the upward power demand generated by the source load fluctuation at time t. This represents a downward power demand; Based on the above feeder power regulation capability and feeder autonomous power requirements, the upper and lower bounds of the feeder power autonomous interval can be derived as follows: The gap between the autonomous power demand of feeders and the power supply capacity of feeders is defined as the autonomous power deficit of feeders. The upward and downward autonomous power deficits of feeders are shown in the following formulas: In the formula: , Autonomous power deficit for both upward and downward feeders; The ratio of the autonomous power deficit of each feeder to the power balance interval of the feeder autonomy within the scheduling cycle is defined as the feeder power autonomy capability, as shown in the following formula. In the formula: The autonomous capability of the i-th feeder is calculated between [0,1]. The closer it is to 1, the stronger the power autonomy of the feeder. T is the scheduling period.

3. The distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy as described in claim 1, characterized in that: In step (2), the construction of the feeder autonomous power deficit uncertainty set based on the Hausdorff distance is specifically as follows: First, M sets of feeder autonomous power deficit scenarios are extracted from the historical data of photovoltaic and load of each feeder. The scenarios are reduced using the K-means clustering algorithm to obtain N typical feeder autonomous power deficit scenarios. The combination of typical scenarios is then used to generate a set of combined typical scenarios and determine their probability distribution. In the formula: The number of combined scenes, Let be the probability of the combined scenario s occurring. , These represent the number of typical photovoltaic scenarios and the scenario probabilities in the combined scenario s. , These represent the typical number of load scenarios and the scenario probability in the combined scenario s, respectively. Due to the uncertainty of typical scenarios of autonomous power deficit in feeders, the scenario probability distribution obtained after clustering has a bias. The degree of bias of the probability distribution is quantified by using Hausdorff distance, and the average value is used instead of the maximum value to calculate the improved Hausdorff distance constraint, so as to more accurately characterize the differences between sets. In the formula: , These are the number of elements in point set A and point set B, respectively; Given the one-way Hausdorff distance from A to B, calculate the distance from any point in A. Find the minimum distance to q points in B, obtain the q minimum values, and calculate their average; similarly, the one-way Hausdorff distance from B to A can be obtained. Improved Hausdorff distance for sets A and B That is, the larger of the two; Based on the initial probability distribution of the generated combined typical scenario set, and with the Hausdorff distance as a constraint, an uncertain set of feeder autonomous power deficit scenarios is constructed, the expression of which is: In the formula: These are the probability values ​​of scene s based on the Hausdorff distance, and the Hausdorff distance constraint is... ,in This is the probability tolerance value; To control the fluctuation range of the probability distribution of combined scenarios, the probability value of scenario s must meet a preset confidence requirement. The constraint relationship is expressed as follows: In the formula: K and M represent the number of typical and historical scenarios of autonomous power deficit for feeders, respectively. , where is the confidence level.

4. The distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy as described in claim 1, characterized in that... In step (3), the construction of a multi-objective control optimization model for distribution networks oriented towards feeder autonomy is specifically as follows: Total operating cost of distribution network It is expressed as follows: In the formula: , These are the operating cost and power interaction cost of the i-feeder, respectively; When optimizing the dispatching of a distribution network considering the autonomous power deficit of feeders after zoning, the power exchange between feeders is based on their initial operating state. The coordination and mutual assistance cost between feeders is: In the formula: Let be the power output of feeder i to other feeders at time t. Let be the power input from other feeders to feeder i at time t. This is the exchange cost coefficient; After the initial power exchange between each feeder, each feeder solves in parallel based on the initial power exchange results to generate the optimal scheduling strategy. The internal operating cost of each feeder is: In the formula: For feeder autonomous power deficit costs, , , and These are the costs of controllable distributed power generation within the feeder, energy storage operation costs, grid transaction costs, and curtailment penalty costs; I. Global constraints between feeders When feeders interact, the interaction power between feeders must satisfy the balance constraint. In the formula: , These represent the active power transmitted and received by the i-th feeder to other feeders at time t; II. Feeder self-management power deficit cost The feeder autonomous power deficit cost is expressed as the sum of the upward and downward deficit costs. In the formula: This refers to the feeder autonomous power deficit cost coefficient. III. Operating Costs and Constraints of Controllable Distributed Power Sources In the formula: , and The upper and lower limits of the output power and the maximum ramp power of the controllable distributed power source; Let be the power generation at time t, and let the power generation cost coefficient be... , and ; IV. Operating costs and constraints of energy storage systems In the formula: , Let be the charging power and discharging power of the stored energy at time t, respectively. This represents the unit loss cost coefficient for energy storage. For charge and discharge efficiency; , These are the maximum charging and discharging power of the energy storage; To represent the charging and discharging states of energy storage respectively; , Let be the energy storage capacity at time t and its rated capacity, respectively. , Indicates the maximum and minimum states of charge; V. Feeder and Mainnet Transaction Costs and Constraints In the formula: , These represent the power exchange between the feeder and the main grid at time t. Time-of-use pricing for purchasing and selling electricity from the grid. For electricity purchase and sale; This represents the maximum transmission power of the tie line; VI. Penalty Costs and Constraints for Distributed Photovoltaic Curtailment In the formula: , These represent the actual and available photovoltaic output at time t, respectively. The penalty cost coefficient for abandoning light; Let t be the available distributed photovoltaic output of the i-th feeder; VII. Power balance constraints within the feeder In the formula: , They are respectively User load and distributed photovoltaic output of feeder i at any time.

5. The distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy as described in claim 4, characterized in that: In step (4), for the multi-objective control optimization model of the distribution network oriented towards feeder autonomy obtained in step (3), a two-stage sub-Blu-ray bar optimization model considering the power deficit of feeder autonomy is constructed, and the objective function is shown in the following equation: In the formula: N is the number of feed lines; , The pre-control stage includes the power exchange cost between feeders and their respective operating costs. The operating costs for each feeder during the real-time control phase; This refers to the power interaction value between feeders; , The variables are divided into first-stage and second-stage variables. The first-stage variables include the power purchase and sale plan, the start-up and shutdown of generating units, and the charging and discharging status of energy storage. The second-stage variables include the output of elements inside each feeder and the amount of curtailed solar power. These are random variables, including the output values ​​of photovoltaic power and load. For the feeder autonomous power deficit uncertainty set; The constraints are expressed in the following form: 。 6. The distributed mutual assistance and coordinated optimization control method for distribution networks oriented towards feeder autonomy as described in claim 1, characterized in that: In step (5), a distributed framework is constructed based on the alternating direction multiplier method and column constraint generation algorithm to solve the above two-stage split-bar optimization model. The algorithm is written in Matlab language, and the solution process and convergence criteria are as follows: S1, for the distributed coordination optimization model between distribution network feeders, the ADMM algorithm is used to solve the problem, simplifying the two-stage distributed bar optimization model into the standard form of ADMM: In the formula: x and z are the original problem variables, and the function and The corresponding power exchange costs between feeders and feeder operating costs The constraints correspond to the global constraints of the feeder, and A, B, and C are simplified coupling matrices. To address the coordination optimization problem between feeders, a boundary variable interaction mechanism is employed to achieve distributed solution. To effectively handle the coordination between constraints and the optimization objective, Lagrange multipliers and penalty terms are introduced, integrating the original problem's constraints into the objective function to form a unified optimization model. Its specific mathematical expression is as follows: In the formula: , Let be the Lagrange multipliers and the penalty parameter, and ; Finally, the threshold values ​​for the incremental changes of the dual and original variables are used as the convergence criteria after solving the problem. When the iteration update amount is lower than the set tolerance, the algorithm terminates and outputs the optimal solution. The convergence criteria are as follows: In the formula: , Let be the original residual and the dual residual, respectively, and k be the number of iterations. The algorithm terminates and outputs the optimal solution when both the original residual and the dual residual satisfy the preset convergence condition. S2, for the autonomous robust optimization model within the feeder, the optimization model within each feeder is solved in parallel. The sub-robust optimization model within each feeder is solved using the CCG algorithm. First, the original problem is decomposed into a main problem and sub-problems, which are expressed as follows: In the formula: n is the number of iterations of the CCG algorithm; After decomposing the above-mentioned sub-Bruker model, the main problem first obtains the optimal solution for operating cost based on the initial worst-case probability distribution, solves the first-stage decision variables, and updates the lower bound LB. The sub-problems, based on solving the min problem first, obtain the cost-optimal operating scheme for each scenario, thus obtaining the second-stage decision variable, the power output plan. Then, the max problem is run to obtain the worst-case probability distribution corresponding to the maximum autonomous power deficit of the feeder, and the upper bound UB is updated. Finally, the solution is completed when the upper and lower bounds converge to less than the preset precision. The optimization model within each feeder is a sub-Bruker model, and each robust model can be solved in parallel.

Images