Layered and partitioned multi-source cooperative adaptive optimization management and control method, system and equipment for flexible power distribution network and medium

Abstract

The invention discloses a layered and partitioned multi-source cooperative adaptive optimization management and control method, system and device for a flexible power distribution network and a medium, and belongs to the technical field of electric power system operation and control, and the method comprises the steps: dynamically generating and reconstructing an elastic autonomous region based on the real-time electrical coupling degree; establishing a bidirectional constraint interaction mechanism between an intra-day layer and a real-time rolling optimization layer, and introducing forward-looking equipment action cost; a distributed collaborative algorithm based on self-adaptive virtual impedance is adopted to realize autonomous power distribution of equipment in the region; screening high-reliability user side resources through multi-dimensional credit assessment and differential excitation; and generating a robust scheduling plan by using hybrid modeling and two-stage random optimization. According to the method, the problem of cooperative management and control of strong uncertainty of the source and load sides of the power distribution network under the access of high-proportion new energy is solved, the transformation from passive response to active prevention and from centralized control to distributed cooperation is realized, and the absorption capability, the operation safety and the economical efficiency of the power distribution network are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of power system operation and control technology, specifically to a method, system, equipment, and medium for hierarchical and zoned multi-source collaborative adaptive optimization management and control of flexible distribution networks. Background Technology

[0002] As the penetration rate of distributed energy, flexible loads, and energy storage devices in distribution networks continues to rise, looking back at the path of technological development, the early focus was mainly on the grid connection and maximum power point tracking of individual distributed power sources. Later, with centralized distribution network management systems at their core, complex mathematical optimization models were constructed to attempt to coordinate overall economic efficiency and security. In recent years, to cope with the computing and communication pressure brought by the massive number of connected devices, a layered and partitioned collaborative architecture has become the mainstream research approach, such as multi-agent systems, virtual power plant aggregation technology, and "cloud-edge-device" collaborative computing frameworks. At the same time, methods such as model predictive control, which can handle uncertainties in a rolling manner, have also been widely introduced. It can be said that these progressive technological explorations have jointly formed the cornerstone of the current optimization and management of flexible distribution networks, with the goal of improving economic efficiency, resilience, and the ability to absorb new energy sources.

[0003] However, when existing technological frameworks are tested in real-world, highly dynamic, and highly uncertain environments, thorny problems emerge. First, there's the contradiction between rigid partitioning and dynamic operation. Most current partitioning methods are based on static topology or historical data. If the source load power fluctuates drastically within a region, these fixed boundaries cannot be adjusted in time, easily leading to the spread of local instability. Ultimately, the upper-level system must step in, incurring high costs. Second, there's the disconnect between optimization decisions at different time scales. Traditional day-day-intraday-real-time scheduling often involves unidirectional, layer-by-layer instruction. When the intraday optimization layer makes plans, it lacks the latest accurate operational status feedback from the real-time layer, forcing it to use conservative constraints. Meanwhile, the real-time layer, in its pursuit of rapid target tracking, may ignore the cumulative wear and tear from frequent device actions, potentially shortening device lifespan. Furthermore, there's the challenge of enabling massive distributed devices to work collaboratively. Centralized optimization has a heavy computational burden, while simple distributed peer-to-peer collaboration ignores differences in the adjustment capabilities and health status of devices, resulting in either unrealistic calculations or unreasonable allocation. Therefore, an adaptive collaborative management and control method that can dynamically reconstruct autonomous regions, optimize inter-layer bidirectional support, take into account equipment differences and health status, and accurately incentivize users based on credit has become the key to breaking through current bottlenecks and unleashing the full potential of flexible power distribution networks. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, this invention aims to solve the problems of low operating efficiency, high safety risks, and difficulty in coordination caused by the strong uncertainty on both the source and load sides, complex and variable network power flow, and massive dispersion of control objects when traditional distribution network operation and control methods face the high proportion of distributed power sources and flexible load access.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks, comprising, Based on real-time measurement data, the electrical coupling strength between distribution network nodes is dynamically calculated to generate flexible autonomous regions and determine control boundaries. Within these flexible autonomous regions, a rolling optimization decision-making mechanism is established to generate collaborative optimization instructions across the entire time scale. These collaborative optimization instructions are then used to allocate and execute power on the equipment side via a distributed collaborative algorithm. The collaborative optimization instructions are then used to filter high-user aggregates through an interactive mechanism and issue differentiated incentive strategies. In the rolling optimization decision-making mechanism, hybrid modeling of uncertain decisions is performed to generate a control plan.

[0007] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, wherein: the generation of the elastic autonomous region includes: Real-time data of the entire distribution network is collected at all nodes, the sensitivity matrix of voltage to injected power of each node is calculated, and after traversing all nodes, the electrical coupling degree matrix of the entire network at the current moment is generated. Divide the currently active initial elastic autonomous regions and obtain the actual net power of all distributed power sources within the regions; When the absolute value of the real-time net power fluctuation in any region is consistently less than the preset threshold, it is determined that all region monitoring is stable, and the existing zoning is maintained.

[0008] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, the generation of the flexible autonomous region further includes: When the absolute value of the real-time net power fluctuation value in any region is greater than or equal to a preset threshold, it is determined that the current region is triggered to reconstruct. Select the node that contributes the most to the net power fluctuation within the trigger reconfiguration area, and based on the network-wide electrical coupling matrix, aggregate nodes with electrical coupling degrees higher than the set strong coupling threshold to form a new node set; A security check is performed on the new node set. When the check passes, a new elastic autonomous region is generated, and the logical control partition map and regional attribute information of the entire network are updated.

[0009] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, the generation of collaborative optimization instructions across the entire time scale includes: Establish an intraday optimization layer with the goal of minimizing the total system operating cost; establish a real-time control layer with the goal of minimizing the total adjustment deviation cost. The intraday optimization layer outputs a resource scheduling plan to the real-time control layer; The real-time control layer performs rolling optimization based on the intraday optimization layer's output plan, and outputs collaborative optimization instructions to the device side.

[0010] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, the power allocation and execution on the equipment side includes: The device side located within the same flexible autonomous region will receive the collaborative optimization command; Based on the real-time adjustable power range and health status of the device, the adaptive virtual impedance value is calculated. The device exchanges its proposed power allocation rate information with the corresponding communication neighbor and iteratively updates its own power allocation rate by combining the corresponding adaptive virtual impedance value. Once the iterative process converges, the power adjustment amount that should be performed is calculated based on the final determined power allocation rate, and the status after execution is fed back to the real-time control layer.

[0011] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, the step of screening high-user aggregates and issuing differentiated incentive strategies includes: Based on the load adjustment needs generated by the optimization decision, an interactive invitation is generated and published to the user-side resource aggregator; Receive the adjustment capacity and expected price declared by the aggregator, retrieve the dynamic credit profile of the declaring aggregator, and classify the credit rating of the aggregator based on the performance data in the dynamic credit profile. Aggregators are selected based on credit rating from high to low, and within the same rating, based on the declared price from low to high, until the cumulative capacity meets the load regulation requirements. Incentive coefficients are set for aggregators with different credit ratings. Settlement unit price is determined based on the incentive coefficients and the bid price, and final scheduling instructions are issued to the selected aggregators.

[0012] As a preferred embodiment of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks described in this invention, the generation of the control plan includes: Extract the credit profile data of resource aggregators on the user side, map it to adjustment coefficients, and decompose the adjustment capacity declared by aggregators into credible capacity and uncertain capacity; Two-stage stochastic optimization includes a first-stage decision and a second-stage decision. In the second stage of the two-stage decision stochastic optimization, the objective function is to minimize the expected total cost and constraints are set. The objective function of the second stage is solved, and a robust control plan is output.

[0013] Another objective of this invention is to provide a hierarchical, zoned, multi-source collaborative adaptive optimization and control system for flexible power distribution networks.

[0014] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a flexible distribution network hierarchical and zoned multi-source collaborative adaptive optimization and control system, comprising: a boundary determination module, an instruction generation module, an equipment-side execution module, a user-side execution module, and a control module; The boundary determination module dynamically calculates the electrical coupling strength between distribution network nodes based on real-time measurement data, generates a flexible autonomous region, and determines the control boundary. The instruction generation module establishes a rolling optimization decision-making mechanism within the elastic autonomous region to generate collaborative optimization instructions across the entire time scale. The device-side execution module performs power allocation and execution on the device side using a distributed collaborative algorithm to implement the collaborative optimization instructions. The user-side execution module uses an interactive mechanism to filter high-user aggregates and issue differentiated incentive strategies to the collaborative optimization instructions. The control module performs hybrid modeling of uncertain decisions within the rolling optimization decision-making mechanism to generate a control plan.

[0015] The present invention provides a computer device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the aforementioned hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible power distribution networks.

[0016] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the aforementioned hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible power distribution networks.

[0017] The beneficial effects of this invention are as follows: This invention provides a hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks. Its core is to construct a complete closed-loop control system that combines perception, decision-making, execution, incentive and robust decision-making through five steps.

[0018] Specifically, based on the dynamic generation and reconstruction of flexible autonomous regions using real-time electrical coupling strength, an adaptive organizational structure is provided for management and control. Secondly, a two-way interaction mechanism between intraday and real-time optimization layers is established. Through dynamic feasible domain corridors and forward-looking equipment action cost models, the coordination of long-term and short-term goals and the executability of plans are achieved. Subsequently, a distributed collaborative algorithm based on adaptive virtual impedance is used to autonomously and rationally allocate and execute instructions on the equipment side, improving the flexibility and reliability of execution. In addition, through a dynamic incentive interaction mechanism based on multi-dimensional credit assessment, high-reliability user-side resources are selected and incentivized, transforming massive loads into high-quality and controllable resources. Hybrid modeling and two-stage stochastic optimization are adopted to classify resources according to credibility and generate robust scheduling plans that consider various uncertain scenarios. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of 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.

[0020] Figure 1 This is a flowchart illustrating the overall process of a hierarchical, zoned, multi-source collaborative adaptive optimization and control method for flexible power distribution networks, as provided in one embodiment of the present invention.

[0021] Figure 2 This is a flowchart illustrating the dynamic region generation of a hierarchical, multi-source collaborative adaptive optimization and control method for flexible power distribution networks, as provided in one embodiment of the present invention.

[0022] Figure 3 This is a flowchart illustrating the user-side differentiated incentive strategy of a hierarchical, zoned, multi-source collaborative adaptive optimization and control method for flexible distribution networks, provided as an embodiment of the present invention. Detailed Implementation

[0023] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0024] Example 1, referring to Figures 1-3 This is one embodiment of the present invention, which provides a hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks, including: S100. Based on real-time measurement data, dynamically calculate the electrical coupling strength between distribution network nodes, generate flexible autonomous regions, and determine control boundaries; S200. Within the flexible autonomous region, establish a rolling optimization decision-making mechanism to generate collaborative optimization instructions across the entire time scale; S300: The collaborative optimization instructions are distributed and executed on the device side through a distributed collaborative algorithm. S400: The collaborative optimization instructions will be used to screen high-user aggregates through an interactive mechanism and issue differentiated incentive strategies. S500: In the rolling optimization decision-making mechanism, hybrid modeling of uncertain decisions is performed to generate control plans; It should be noted that existing traditional partitioning methods, based on fixed topology or coarse electrical distances, are ill-suited to the dynamic scenarios of fluctuating power flow directions and time-varying electrical coupling relationships under high-proportion renewable energy access, leading to a disconnect between partitioning results and actual electrical connections. Traditional multi-layer optimization is mostly unidirectional and open-loop. The intraday planning layer ignores real-time operational fluctuations, making plans frequently infeasible; the real-time control layer pursues short-term tracking while ignoring long-term equipment wear and tear and economic efficiency; reliance on centralized controllers for global calculations and command issuance introduces single-point failure risks. Traditional demand response mechanisms are mostly based on price ranking, lacking quantitative assessment and incentives for the reliability of users' historical response accuracy and latency, resulting in low actual response rates and large deviations of winning resources, failing to form stable and reliable control capabilities; traditional deterministic optimization or simple interval robust optimization cannot accurately characterize the probabilistic characteristics and reliability differences of renewable energy output and flexible load response, leading to scheduling schemes that are either too conservative or too aggressive.

[0025] Therefore, to address the aforementioned issues, the steps S100-S500 are implemented to break the limitations of fixed partitions by calculating electrical coupling in real time and dynamically generating flexible autonomous regions. This allows the control architecture to adaptively adjust to the power grid's operating status, laying the foundation for local power balance and coordinated control. A two-way interaction mechanism between intraday and real-time optimization layers is established to resolve the conflict between long-term economic efficiency and short-term security, as well as between frequent equipment operation and lifespan loss, generating a globally coordinated and feasible full-time-scale scheduling plan. A distributed collaborative algorithm based on adaptive virtual impedance enables numerous distributed devices within the region to autonomously, quickly, and rationally allocate regulation commands without the need for a centralized controller. A dynamic interaction and incentive strategy based on multi-dimensional credit assessment is established to accurately select high-reliability user aggregates. By hierarchically modeling the uncertainties of different resources and employing two-stage stochastic optimization, a robust scheduling plan capable of withstanding various possible fluctuation scenarios is generated, significantly enhancing the resilience and security of the power grid operation.

[0026] Example 2, refer to Figure 1This is one embodiment of the present invention, which provides a hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks, including: In this embodiment of the invention, step S100 dynamically calculates the electrical coupling strength between distribution network nodes based on real-time measurement data, generates a flexible autonomous region, and determines the control boundary, including the following steps S101-S104: S101. Collect real-time data of the entire distribution network at all nodes, calculate the voltage sensitivity matrix of each node to injected power, and after traversing all nodes, generate the electrical coupling degree matrix of the entire network at the current moment. The specific operation steps are as follows: Smart measurement units are placed at all nodes of the distribution network to collect real-time data of the entire network synchronously at a fixed period. It should be noted that the fixed period can be once every 5 minutes. The real-time data to be collected across the entire network includes the voltage amplitude, voltage phase angle, injected active power, and injected reactive power of all nodes; and the active power flow and reactive power flow of all branches. After real-time data acquisition is completed, the comprehensive electrical coupling degree between nodes is calculated based on the data at the current moment. Specifically, the comprehensive electrical coupling degree is calculated based on Kirchhoff's laws and network topology parameters, combined with the current power flow solution, to calculate the sensitivity matrix of each node's voltage to injected power. Select node i and node j, and sum the three normalized indices by weighting them according to the absolute value of the voltage phase angle difference between nodes, the reciprocal of the electrical distance between nodes obtained based on the sensitivity matrix, and the weighted value of the apparent power flow from node i to node j at the current moment as a percentage of the branch capacity, to obtain the comprehensive electrical coupling degree between 0 and 1. It should be noted that the absolute value of the voltage phase angle difference between two nodes indicates a tighter coupling; the reciprocal of the electrical distance between nodes obtained based on the sensitivity matrix indicates a tighter coupling; and the weighted value of the proportion of apparent power flow from node i to node j to the branch capacity at the current moment indicates a more active power exchange and a tighter coupling. Among them, the closer the comprehensive electrical coupling degree is to 1, the stronger the coupling. After traversing all node pairs, the electrical coupling degree matrix of the entire network at the current moment is formed.

[0027] It should be noted that by changing the traditional static partitioning based on fixed topology or coarse electrical distance, the partitioning criteria can reflect the real-time operating status of the power grid, capture the strength of real-time electrical connections, and achieve dynamic and precise partitioning criteria.

[0028] S102. Divide the currently effective initial elastic autonomous region and obtain the actual net power of all distributed power sources within the region. The specific operation steps are as follows: Maintain a list of currently active flexible autonomous regions, and perform monitoring and judgment for each region in the list; S1021. Obtain the current actual output of all distributed power sources in the region, the current actual value of all loads, and the actual power of the region's connection line to the outside. Calculate the actual net power of the region by subtracting the current actual value of all loads from the sum of the current actual output of all distributed power sources in the region and the actual power of the region's connection line to the outside. At the same time, the predicted net power of the current region at the beginning of the current period is obtained, and the absolute value of the difference between the actual net power of the region and the predicted net power is the real-time net power fluctuation value of the current region.

[0029] S1022. Obtain the comprehensive regulation capability of the current region based on the sum of the absolute values ​​of the maximum real-time power that can be increased and the maximum power that can be decreased reported by all controllable distributed power sources and energy storage devices in the region. It should be noted that the comprehensive adjustment capability is dynamically updated according to the equipment status, and the adjustment capability threshold ratio is set to 70%. It should also be noted that the adjustment capability threshold ratio is a 30% adjustment margin reserved for the region to cope with local fluctuations in the region. When fluctuations consume 70% of the adjustment capability, the region is considered to be in a critical equilibrium state.

[0030] When the absolute value of the real-time net power fluctuation value of any region is continuously exceeded by the preset threshold, that is, the duration of the comprehensive adjustment capability corresponding to the preset threshold multiplied by the adjustment capability threshold ratio reaches three consecutive monitoring cycles, i.e. 15 minutes, it is determined that the current region has lost its autonomous balance capability, and a region reconstruction trigger command is generated. The command contains the unique identifier of the trigger region. Conversely, if all areas are deemed to be stable, the current cycle ends, and the existing zoning is maintained.

[0031] Reference Figure 2 S103. When the absolute value of the real-time net power fluctuation in any region is greater than or equal to a preset threshold, it is determined that the current region has triggered reconfiguration. The node with the largest net power fluctuation contribution in the region that has triggered reconfiguration is selected. Based on the electrical coupling degree matrix of the entire network, nodes with electrical coupling degree higher than the set strong coupling threshold are aggregated to form a new node set. The specific operation steps are as follows: Upon receiving a region reconstruction trigger command, the dynamic region generation algorithm is initiated. S1031. The node with the largest net power fluctuation contribution within the region identified in the region reconstruction trigger instruction is used as the initial seed node. It should be noted that the largest net power fluctuation contribution means that the absolute value of the deviation between the actual injected power of the current node and the corresponding predicted value is the largest. The current initial seed node is added to the temporary set of the new region node set. S1032. Enter the loop expansion process. In each loop, take all nodes in the new regional node set as the source nodes. In the whole network electrical coupling degree matrix, find all nodes whose coupling degree with the source nodes is higher than the strong coupling threshold and add them to the candidate node set. It should be noted that the strong coupling threshold is based on the statistical analysis of historical power grid stability partition cases. The 90th percentile of the coupling degree between nodes is taken to ensure that the electrical connection of the included nodes is close enough. Therefore, it is set to 0.85. The real-time net power deviation of each node in the candidate node set is calculated by subtracting the predicted value from the actual value of the node power, and compared with the total net power fluctuation value of the current new region node set; nodes with opposite signs of power deviation (i.e., one is positive and the other is negative) are selected first to form a priority candidate list. Furthermore, a security check is performed on the new node set. The specific steps are as follows: Nodes are selected sequentially from the priority candidate list. The simulation adds the current node to the new regional node set. The power flow calculation engine is invoked, and a safety check is performed to determine whether the key safety constraints are met. The key safety constraints include that the node voltage deviation does not exceed ±5% of the rated value and the critical line load rate does not exceed 90%. If the verification passes, the current node is moved from the candidate node set to the new regional node set, and this expansion is recorded; if the verification fails or the priority list is empty, the remaining candidate nodes are considered to be verified in order of comprehensive electrical coupling degree until all nodes have been verified. Calculate the overall adjustment capability of the current new region node set and determine whether the current capability is sufficient to cover the absolute value of the net power fluctuation of the original triggering region; if so, terminate the expansion loop; if the expansion has caused the new region size to exceed the preset maximum number of nodes limit, also terminate the loop, where the maximum number of nodes limit can be set to 50 nodes.

[0032] S1033. After the loop terminates, the final new set of regional nodes is defined as the new elastic autonomous region; and a unique logical identifier (ID) is assigned to the current new elastic autonomous region, and a regional coordination agent is assigned.

[0033] It should be noted that guiding the generation process of new regions, ensuring that new regions have internal balancing capabilities and operate safely, and achieving self-balancing-oriented intelligent construction of regions, prioritizing the absorption of nodes with opposite power deviations, means combining power-deficient regions and power-rich regions under the premise of electrical tightness, maximizing the power mutual assistance capability within the region.

[0034] S104. When the verification passes, a new flexible autonomous region is generated, and the entire network logical control partition map and region attribute information are updated. The specific operation steps are as follows: Based on the new flexible autonomous regions, the logical control partition map of the entire network is updated. The logical control partition map is a mapping table that records which flexible autonomous region each power grid node belongs to. At the same time, an attribute information package is generated for each region, which includes the region ID, the list of nodes contained therein, the currently calculated regional comprehensive net load forecast curve, the upper and lower limits of the regional comprehensive regulation capacity, and the communication address of the regional coordination agent. The current attribute information package is encapsulated as a standardized data object.

[0035] In this embodiment of the invention, in step S200, a rolling optimization decision-making mechanism is established within the elastic autonomous region to generate collaborative optimization instructions across the entire time scale, including the following steps S201-S202: Based on the aforementioned flexible autonomous region, a bidirectional interactive rolling optimization decision-making mechanism is established between the intraday optimization layer and the real-time control layer. The real-time control layer provides dynamic feasible domain corridor constraints for the intraday optimization layer, and the intraday optimization layer provides forward-looking equipment action cost assessments for the real-time layer, jointly generating collaborative optimization instructions across the entire time scale. S201. Establish an intraday optimization layer, with the total system operating cost as the target. The specific steps are as follows: Using all flexible autonomous regions as the basic scheduling unit, an intraday optimization model with a time resolution of 1 hour for the next 24 hours is established. The objective function is to minimize the total system operating cost. The total system operating cost is calculated as the sum of electricity purchase cost, equipment operation and maintenance cost, load interruption compensation cost, and forward-looking equipment operation cost. It should be noted that the electricity purchase cost is the product of tie line power and electricity price, and the initial value of forward-looking equipment operation cost is 0. Specifically, the input includes the basic scheduling unit, as well as the historical action count and health status data of the equipment. The intraday optimization layer decomposes and maps its 24-hour plan onto consecutive 15-minute real-time control cycles. For each simulated real-time cycle, based on the equipment power change points in the distribution network plan, it counts the number of simulated actions of various types of equipment within the current cycle. ; Calculate the forward-looking equipment operating costs: ; in, The total forward-looking equipment operating cost; k is the equipment index; The cost weight of a single standard action for equipment k is calculated as the ratio of the equipment purchase cost to the rated total number of actions. Let k be the sum of the number of actions performed by device k over all simulation cycles. This represents the maximum number of safe actions allowed for device k within a real-time control cycle. This represents the current health status of device k; This is a sensitivity coefficient for health status, with a positive value and a typical value of 0.5. It should be noted that, This is a dimensionless index between 0 and 1, where 1 indicates the equipment is brand new and 0 indicates the end of its lifespan. For different types of equipment in flexible distribution networks (including but not limited to photovoltaic inverters, battery storage systems, and gas turbines), a unified method based on equivalent aging is used for online calculation. That is, it is calculated by subtracting the ratio of the consumed life index to the rated life index from 1; the consumed life index (such as the number of cycles, operating hours, and number of starts) for different types of equipment in flexible distribution networks is equivalently quantified into the life operating time; the rated life index is the rated equivalent full life time, which is provided by the equipment manufacturer.

[0036] The constraints set based on conventional constraints include: the dynamic feasible domain corridor constraint is the key decision variable for the tie-line power of each hourly segment in the future, and the constraints fed back by the real-time layer are added, that is, the planned value is between the lower limit and the upper limit of the corridor. The dynamic feasible domain corridor constraint forces the intraday plan to fall within the range of real-time operational safety and feasibility. Specifically, for the next T-hour, the real-time layer calculates the upper and lower limits of the dynamic feasible region corridor for the tie-line power within T hours based on the minute-level state trajectory. Find the tie-line power from the 60 minute-level prediction points corresponding to hour T. maximum value and minimum value ; Obtain the long-term maximum allowable transmission limit for this connection. Define the safety buffer coefficient for the current time period. It should be noted that the safety buffer coefficient is dynamically adjusted based on the predicted penetration rate of new energy sources in the current period; a higher penetration rate results in a lower buffer coefficient. Increase, low penetration rate Decrease The value range is [0.75, 0.95], and it is determined by offline analysis of historical data at different penetration rates to determine the risk probability of reverse power flow or voltage problems.

[0037] The lower and upper limits of the corridors in the dynamic feasible region are: ; in, , , where represents the upper and lower limits of the dynamic feasible domain corridor for region i in time period T, and β is the scaling factor with a default value of 1.

[0038] It should be noted that the system power balance constraint in this invention can be included in the intraday optimization model. By comparing the sum of the input power of all regional tie lines, the total output of distributed power sources in the region, and the total discharge power of energy storage with the sum of the total load output value of all regions, the total charging power of energy storage, and the estimated value of conventional network loss, the two are required to be equal, ensuring that the total power generation plus injection equals the total load plus output. The tie line transmission capacity constraint is to compare the planned active power of each tie line in each time period with the long-term allowable maximum transmission power limit of that tie line (determined by the line thermal stability limit), and require that the corresponding absolute value does not exceed the limit. Energy storage energy balance and capacity constraint are achieved by adding the remaining power of the energy storage unit in the previous period to the product of the charging power in the current period and the efficiency and the time interval, and subtracting the discharge power in the current period divided by the product of the efficiency and the time interval, to obtain the power of the current period. This power is required to be between the rated minimum and maximum capacity. The upper and lower limits of equipment output are constrained by comparing the actual set power of each controllable power generation unit or energy storage converter in each time period with the rated minimum technical output and rated maximum technical output specified on the equipment nameplate, and requiring the corresponding values ​​to be within the range.

[0039] Output the resource scheduling plan for the intraday optimization layer for the next 24 hours, and send the current plan to the real-time control layer as a tracking benchmark; In summary, the intraday optimization layer outputs the regional contact line reference power plan to the real-time control layer.

[0040] S202. Establish a real-time control layer with the goal of minimizing the adjustment cost of the intraday optimization layer's plan deviation. The real-time control layer performs rolling optimization based on the intraday optimization layer's output plan and outputs collaborative optimization instructions from the device side. The specific operation steps are as follows: A real-time optimization model with a time window of 15 minutes and a resolution of 1 minute is established with a period of 5 minutes. The objective function is to minimize the total adjustment deviation cost, which is calculated by summing the equipment's own adjustment cost and the forward-looking equipment action cost. Setting constraints based on conventional constraints includes: The minute-level power balance constraint based on ultra-short-term forecasts is to compare the sum of minute-level ultra-short-term forecast output of all distributed power sources in each elastic autonomous region, the sum of minute-level planned discharge power of energy storage in the region, the sum of minute-level ultra-short-term forecast values ​​of loads in the region, the sum of minute-level planned charging power of energy storage, and the minute-level planned power of regional tie lines in the region in the real-time optimization model, and require that the difference be zero at each minute-level time point. The minute-level adjustment rate constraint of the equipment is obtained by calculating the absolute value of the difference between the current minute set power of the controllable equipment and the actual power of the previous minute, and dividing it by the time interval. This rate is required to be less than the maximum minute-level ramp rate allowed by the equipment's technical specifications. The node voltage safety constraint is to compare the actual value of each node voltage amplitude obtained from state estimation or power flow calculation with 1.07 times and 0.93 times the nominal voltage value of the node, and require the node voltage to be within this range; Line current-carrying constraints are imposed by comparing the actual current amplitude of each line, obtained from state estimation or power flow calculation, with the long-term allowable maximum current-carrying capacity of that line, determined based on conductor material and operating ambient temperature, and requiring the line current to not exceed this limit.

[0041] Based on the objective function solution results of the intraday optimization layer and the real-time control layer, the system outputs the full-time-scale collaborative optimization instructions for all controllable devices within the next 5-15 minutes, and determines the reference virtual impedance parameters according to the rated power capacity of the devices.

[0042] Furthermore, by establishing a two-way interactive rolling optimization decision-making mechanism between the intraday optimization layer and the real-time control layer, where the real-time layer provides dynamic feasible domain corridor constraints for the intraday optimization layer and the intraday optimization layer provides forward-looking equipment action costs for the real-time layer, the traditional one-way (top-down) optimization mode is broken. A two-way information and constraint interaction is established between the two time-scale optimization models to solve the problems of conflict between long-term and short-term optimization goals and the disconnect between planning and execution.

[0043] It should be noted that the dynamic feasible domain corridor constraint feeds back the minute-level fluctuation range of real-time operation to the intraday planning layer, forcing long-term plans driven by economics to fall within a safe and feasible corridor, thereby improving the feasibility and safety of the plan. In addition, the forward-looking equipment action cost model introduces the long-term equipment wear and tear costs into short-term real-time optimization, avoiding excessive wear and tear on equipment in pursuit of short-term adjustment effects.

[0044] In an embodiment of the present invention, S300 involves power allocation and execution of the cooperative optimization instruction on the device side using a distributed cooperative algorithm, including the following steps S301-S302: The control instructions for distributed power sources and energy storage in the aforementioned collaborative optimization instructions are used to allocate and execute power on the equipment side of the distribution network through a distributed collaborative algorithm based on consistency theory and adaptive virtual impedance. S301. The device side located within the same flexible autonomous region receives the collaborative optimization command, and calculates the adaptive virtual impedance value based on the device side's real-time adjustable power range and health status. The specific operation steps are as follows: Located within the same flexible autonomous region, they receive collaborative optimization instructions through the regional communication network; Real-time monitoring of local operation data of the distribution network in the flexible autonomous region, and collection of real-time adjustable power range and equipment health status index; Based on local operating data, the ratio of the real-time adjustable power range to the rated power capacity is calculated to obtain a proportional factor reflecting the current relative adjustment potential. At the same time, the equipment health status index is obtained. The reference virtual impedance parameter is divided by the product of the proportional factor and the health status index to obtain the final adaptive virtual impedance value.

[0045] S302. The device exchanges its proposed power allocation rate information with the corresponding communication neighbor, and iteratively updates its own power allocation rate by combining the corresponding adaptive virtual impedance value. The specific operation steps are as follows: Based on the regional communication network, identify other devices that can directly exchange data, namely communication neighbors. Each device is initialized with power allocation rate information, which represents the proportion of the power adjustment amount proposed by this device to the total regional target. The initial value can be set to zero. The iterative negotiation process is carried out at fixed intervals. In each iteration, each device sends its current power allocation rate information to all communication neighbor devices and also receives the current power allocation rate information sent by all communication neighbor devices. The device assigns a weight to each received allocation rate (including its own) based on the allocation rates of all its neighbors and its own allocation rate. Then, it calculates a new allocation rate by calculating the weighted average of all allocation rates. The weight is inversely proportional to the square of the adaptive virtual impedance value of the terminal that sent the allocation rate. Based on the updated allocation rate, the equipment calculates the corresponding temporary power adjustment command value and compares this temporary command value with its own real-time adjustable power range. In response to a temporary power adjustment command value being greater than or equal to its own real-time adjustable power range, the current device side immediately corrects it to the value at the boundary of the adjustable range. After correction, the corresponding power allocation rate is calculated in reverse based on the actual executable power value. When exchanging information in the next iteration cycle, the corrected allocation rate will be broadcast instead of the previously calculated theoretical value. The allocation rate of all devices is continuously monitored. When the change in the allocation rate of all devices is less than the allocation threshold in three consecutive iterations, it is determined that the distributed collaborative iterative process has converged and the power allocation ratio has reached a consensus. It should be noted that the allocation threshold is calibrated experimentally and is usually set to one-thousandth. Once the iteration process has converged, the power adjustment amount to be performed is calculated based on the final determined power allocation rate, and the status after execution is fed back to the real-time control layer.

[0046] In this embodiment of the invention, step S400 involves filtering high-user aggregates through an interactive mechanism and issuing differentiated incentive strategies to the collaborative optimization instruction, including the following steps S401-S404: Based on the load adjustment needs generated by the optimization decision, an interactive invitation is generated and published to the user-side resource aggregator. The specific operation steps are as follows: Reference Figure 3 The collaborative optimization instructions address the flexible load adjustment needs on the user side. Through an interactive mechanism based on multi-dimensional credit assessment and dynamic incentive mapping, highly reliable user-side resource aggregators are selected and differentiated incentive strategies are issued to guide the response.

[0047] It should be noted that this technology enables multiple devices to autonomously allocate overall control commands even without a central controller within the region.

[0048] S401, Generate an interactive invitation: To balance the power of flexible autonomous regions, during specific scheduling periods in the future In this process, load regulation requirements are obtained by aggregating resources from the user side, and at the same time, it is also necessary to collect... The time-based basic settlement price generates a structured flexible load interaction invitation, which includes... The invitation includes the time period, total load adjustment requirements, application deadline, basic compensation price, and a unique identifier for this invitation. Once generated, the invitation will be published to all registered user-side resource aggregator platforms via the communication network.

[0049] S402. Before the application deadline, each aggregator shall submit application information based on the availability of controllable load, including: aggregator identification, committed adjustment capacity and expected compensation unit price; Upon receiving the application information, the system immediately and concurrently accesses the dynamic credit profile established for the applicant aggregator. This dynamic credit profile is a real-time updated database record that stores the performance data of the current aggregator in historical interactive events. After each interactive event, the profile is updated based on the current aggregator's actual performance. The core data of the credit profile includes historical average response accuracy, historical average response latency, historical commitment fulfillment rate, and historical average capacity sustainability. S403. Receive the adjustment capacity and expected price declared by the aggregator, retrieve the dynamic credit profile of the declaring aggregator, and classify the aggregator's credit rating based on the performance data in the dynamic credit profile. The specific operation steps are as follows: After the application deadline, a real-time credit score is calculated for each aggregator that submits a valid application before this dispatch. Based on the core data of the credit profile, the historical average response delay is normalized, and weights are assigned to the four core data. The weight allocation is determined based on their importance to the reliability of power grid regulation. The real-time credit score is calculated using the weighted linear summation method. After calculating the credit scores of all the aggregators that have applied, they are classified into different levels based on credit thresholds. It should be noted that the credit thresholds are based on the statistical distribution of long-term historical credit scores. In an optional embodiment, the credit threshold can be taken from two percentiles of the credit score data of all aggregators over the past year: the upper quartile, i.e., the 75th percentile, as the lower limit threshold for Grade A, and the median, i.e., the 50th percentile, as the lower limit threshold for Grade B. It should be noted that the ranking rules are as follows: When the real-time credit score is greater than or equal to the lower limit of Grade A, the aggregator is classified as high credit. When the real-time credit score is less than the lower limit of Grade A but greater than or equal to the lower limit of Grade B, the aggregator is classified as a medium-credit merchant. When the real-time credit score is lower than the lower limit threshold of Grade B, the aggregator is classified as having low credit. S404. Based on credit rating from high to low, and within the same rating, in order of bid price from low to high, select aggregators until the cumulative capacity meets the load adjustment requirements. The specific operation steps are as follows: The declared capacity is sorted in descending order of credit rating, and within the same credit rating, it is sorted in ascending order of declared price. According to this sorting, the declared capacity is added up in sequence until the accumulated capacity first reaches or exceeds the load adjustment demand of the user-side aggregated resources; the aggregator that is added up is selected to form the winning scheduling list; the capacity of the last aggregator in the list needs to be reduced proportionally, and the reduction exceeds the limit. Furthermore, incentive coefficients are set for aggregators with different credit ratings. The settlement unit price is determined based on the incentive coefficients and the bid price. Finally, the final scheduling instruction is issued to the selected aggregator. The specific operation steps are as follows: An incentive coefficient is defined for each credit rating, and the K value is positively correlated with the credit rating. In an optional embodiment, the incentive coefficient for high-credit aggregators =1.15; Incentive coefficient for credit aggregators =1.00; Incentive coefficient for low credit aggregators =0.90.

[0050] For each aggregator in the winning bid scheduling list, the settlement unit price is confirmed in this round of interaction based on the product of the minimum of the declared price and the basic compensation price and the incentive coefficient, the final scheduling instruction is generated and issued for execution.

[0051] In an embodiment of the present invention, in step S500, the bidirectional interactive rolling optimization decision-making mechanism performs hybrid modeling of uncertain decisions to generate a control plan, including the following steps S501-S503: S501. Extract the credit profile data of user-side resource aggregators and map it to adjustment coefficients. The specific steps are as follows: For each successfully signed user-side resource aggregator, the historical average response accuracy error rate and commitment fulfillment rate are extracted from the credit profile data. Based on the mapping rule of subtracting the product of the commitment fulfillment rate and the historical average response accuracy error rate from the commitment fulfillment rate, these two indicators are combined and converted into an adjustment coefficient, with a value between 0 and 1. It should be noted that aggregators with high fulfillment rates and small errors have an adjustment coefficient close to 1, indicating that the declared adjustment capacity is highly reliable; conversely, the adjustment coefficient closes to 0, indicating that there is a significant risk of discounting the declared capacity.

[0052] S502. Decompose the adjustment capacity declared by the aggregator into reliable capacity and uncertain capacity. The specific steps are as follows: Based on the obtained adjustment coefficients, the adjustment capacity declared by the user-side schedulable load resources is decomposed into a reliable capacity component and an uncertain capacity component: The reliable capacity portion is the product of the adjustment coefficient and the adjustment capacity declared by the user-side schedulable load resources. This portion of capacity is considered to be available under statistically uncertain conditions and directly participates in power balancing.

[0053] The uncertain capacity portion is the product of the adjustment coefficient and the adjustment capacity declared by the user-side schedulable load resources. This uncertainty will be classified as unstatistical uncertainty for management.

[0054] All forecasts for renewable energy output are considered as uncertain capacity, and the fluctuation range directly corresponds to unstatistical uncertainty.

[0055] S503. Two-stage stochastic optimization includes a first-stage decision and a second-stage decision. The specific steps are as follows: The first stage of decision variables, namely the current decision, includes the start-up and shutdown status of thermal power generating units, the daily start-up and shutdown plan of gas turbines, and the planned charging and discharging status of energy storage in each time period; The second stage decision variables, namely the optimal values ​​of all continuous variables awaiting decision, are based on the scenario of statistical uncertainty. Specifically, they are the active power output of each conventional unit, the actual charging and discharging power of energy storage, the power purchased from the main grid, and the actual power called up by the user-side trusted capacity. Furthermore, in the second stage of the two-stage decision stochastic optimization, the objective function is to minimize the expected total cost, and constraints are set. The objective function of the second stage is solved to output a robust control plan. The specific operation steps are as follows: The objective function is to minimize the expected total cost. ; in, The electricity price at time t. To determine the active power output of the generator set during time period t in scenario s, This refers to the actual charging and discharging power of the unit's energy storage during time period t in scenario s. For the user-side trusted capacity power in scenario s at time t, Let g be the cost function of unit g. For energy storage loss costs, To compensate users for costs, Let be the probability of scenario s occurring, where s is the scenario. Let g represent the main grid power purchased in time period t under scenario s, where g represents conventional generator sets and t represents the time period. The constraints include power balance constraints, equipment operation constraints, network security constraints, and user-side call constraints. Using a mathematical programming solver, and employing algorithms such as branch and bound and cutting plane, the optimal solution for the first and second stage decision variables that satisfies all constraints is searched to minimize the total expected cost including the penalty term. After solving, the core output includes: The first phase of decision-making outputs includes the unit start-up and shutdown plan and the status of the energy storage plan.

[0056] The second-stage decision serves as the main output, generating a complete robust power generation and dispatch plan for each typical load scenario, including the output curves of each unit, the main grid power purchase, the energy storage charging and discharging plan, and the user-side reliable capacity call plan. Adaptive control is implemented based on the generated control plan.

[0057] It should be noted that the decision-making model explicitly considers uncertain scenarios and their probabilities, seeks robust solutions with optimal expected costs, and generates robust scheduling plans with foresight and risk resistance. Compared with traditional deterministic optimization, it considers a variety of possible uncertain scenarios, making the final execution plan more adaptable and less prone to collapse when facing actual fluctuations.

[0058] Example 3 is an embodiment of the present invention. The above is an illustrative scheme of a hierarchical and zonal multi-source collaborative adaptive optimization control method for flexible distribution networks. It should be noted that the technical solution of a hierarchical and zonal multi-source collaborative adaptive optimization control system for flexible distribution networks and the technical solution of the above-described hierarchical and zonal multi-source collaborative adaptive optimization control method for flexible distribution networks belong to the same concept. Details not described in detail in the technical solution of the hierarchical and zonal multi-source collaborative adaptive optimization control system for flexible distribution networks in this embodiment can be found in the description of the above-described technical solution of the hierarchical and zonal multi-source collaborative adaptive optimization control method for flexible distribution networks.

[0059] This embodiment provides a hierarchical and zoned multi-source collaborative adaptive optimization and control system for flexible distribution networks, including: a boundary determination module, an instruction generation module, an equipment-side execution module, a user-side execution module, and a control module; The boundary determination module dynamically calculates the electrical coupling strength between distribution network nodes based on real-time measurement data, generates flexible autonomous regions, and determines the control boundaries. The instruction generation module establishes a two-way interactive rolling optimization decision-making mechanism within the flexible autonomous region to generate collaborative optimization instructions across the entire time scale. The device-side execution module uses a distributed collaborative algorithm to allocate and execute power on the device side for collaborative optimization instructions. The user-side execution module uses an interactive mechanism to filter high-user aggregates and issue differentiated incentive strategies for collaborative optimization instructions; The control module performs hybrid modeling of uncertain decisions in a two-way interactive rolling optimization decision-making mechanism to generate control plans.

[0060] This embodiment also provides an electronic device applicable to a hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks as proposed in the above embodiment.

[0061] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a hierarchical, zoned, multi-source collaborative adaptive optimization and control method for flexible distribution networks as proposed in the above embodiments.

[0062] The storage medium proposed in this embodiment and the method for implementing a hierarchical and zonal multi-source collaborative adaptive optimization and control of a flexible distribution network proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0063] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0064] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks, characterized in that: include, Based on real-time measurement data, the electrical coupling strength between distribution network nodes is dynamically calculated to generate a flexible autonomous region and determine the control boundary. Within the aforementioned flexible autonomous region, a rolling optimization decision-making mechanism is established to generate collaborative optimization instructions across the entire time scale; The collaborative optimization instructions are used to allocate and execute power on the device side through a distributed collaborative algorithm; The collaborative optimization instructions will use an interactive mechanism to filter high-user aggregates and issue differentiated incentive strategies. In the rolling optimization decision-making mechanism, hybrid modeling of uncertain decisions is performed to generate a control plan.

2. The method for hierarchical and zoned multi-source collaborative adaptive optimization and control of a flexible distribution network as described in claim 1, characterized in that, The generation of the flexible autonomous region includes: Real-time data of the entire distribution network is collected at all nodes, the sensitivity matrix of voltage to injected power of each node is calculated, and after traversing all nodes, the electrical coupling degree matrix of the entire network at the current moment is generated. Divide the currently active initial elastic autonomous regions and obtain the actual net power of all distributed power sources within the regions; When the absolute value of the real-time net power fluctuation in any region is consistently less than the preset threshold, it is determined that all region monitoring is stable, and the existing zoning is maintained.

3. The method for hierarchical and zoned multi-source collaborative adaptive optimization and control of a flexible distribution network as described in claim 2, characterized in that, The generation of the flexible autonomous region also includes: When the absolute value of the real-time net power fluctuation value in any region is greater than or equal to a preset threshold, it is determined that the current region is triggered to reconstruct. Select the node that contributes the most to the net power fluctuation within the trigger reconfiguration area, and based on the network-wide electrical coupling matrix, aggregate nodes with electrical coupling degrees higher than the set strong coupling threshold to form a new node set; A security check is performed on the new node set. When the check passes, a new elastic autonomous region is generated, and the logical control partition map and regional attribute information of the entire network are updated.

4. The hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks as described in claim 3, characterized in that, The method for generating collaborative optimization instructions across the entire time scale includes: Establish an intraday optimization layer with the goal of minimizing the total system operating cost; establish a real-time control layer with the goal of minimizing the total adjustment deviation cost. The intraday optimization layer outputs a resource scheduling plan to the real-time control layer; The real-time control layer performs rolling optimization based on the intraday optimization layer's output plan, and outputs collaborative optimization instructions to the device side.

5. The method for hierarchical and zoned multi-source collaborative adaptive optimization and control of a flexible distribution network as described in claim 4, characterized in that, The power allocation and execution on the device side include: The device side located within the same flexible autonomous region will receive the collaborative optimization command; Based on the real-time adjustable power range and health status of the device, the adaptive virtual impedance value is calculated. The device exchanges its proposed power allocation rate information with the corresponding communication neighbor and iteratively updates its own power allocation rate by combining the corresponding adaptive virtual impedance value. Once the iterative process converges, the power adjustment amount that should be performed is calculated based on the final determined power allocation rate, and the status after execution is fed back to the real-time control layer.

6. The method for hierarchical and zoned multi-source collaborative adaptive optimization and control of a flexible distribution network as described in claim 5, characterized in that, The process of selecting high-user aggregates and issuing differentiated incentive strategies includes: Based on the load adjustment needs generated by the optimization decision, an interactive invitation is generated and published to the user-side resource aggregator; Receive the adjustment capacity and expected price declared by the aggregator, retrieve the dynamic credit profile of the declaring aggregator, and classify the credit rating of the aggregator based on the performance data in the dynamic credit profile. Aggregators are selected based on credit rating from high to low, and within the same rating, based on the declared price from low to high, until the cumulative capacity meets the load regulation requirements. Incentive coefficients are set for aggregators with different credit ratings. Settlement unit price is determined based on the incentive coefficients and the bid price, and final scheduling instructions are issued to the selected aggregators.

7. The method for hierarchical and zoned multi-source collaborative adaptive optimization and control of a flexible distribution network as described in claim 6, characterized in that, The generation and management plan includes: Extract the credit profile data of resource aggregators on the user side, map it to the adjustment coefficient, and decompose the adjustment capacity declared by the aggregators into credible capacity and uncertain capacity; Two-stage stochastic optimization includes a first-stage decision and a second-stage decision. In the second stage of the two-stage decision stochastic optimization, the objective function is to minimize the expected total cost and constraints are set. The objective function of the second stage is solved, and a robust control plan is output.

8. A hierarchical and zoned multi-source collaborative adaptive optimization and control system for flexible distribution networks, employing the hierarchical and zoned multi-source collaborative adaptive optimization and control method for flexible distribution networks as described in any one of claims 1 to 7, characterized in that, include: Boundary determination module, instruction generation module, device-side execution module, user-side execution module, and control module; The boundary determination module dynamically calculates the electrical coupling strength between distribution network nodes based on real-time measurement data, generates a flexible autonomous region, and determines the control boundary. The instruction generation module establishes a rolling optimization decision-making mechanism within the elastic autonomous region to generate collaborative optimization instructions across the entire time scale. The device-side execution module performs power allocation and execution on the device side using a distributed collaborative algorithm to implement the collaborative optimization instructions. The user-side execution module uses an interactive mechanism to filter high-user aggregates and issue differentiated incentive strategies to the collaborative optimization instructions. The control module performs hybrid modeling of uncertain decisions within the rolling optimization decision-making mechanism to generate a control plan.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the flexible distribution network hierarchical and partitioned multi-source collaborative adaptive optimization and control method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the hierarchical and zonal multi-source collaborative adaptive optimization and control method for flexible distribution networks as described in any one of claims 1 to 7.

Images