A power distribution network power loss area fast power supply recovery and load balancing control method coordinated with energy storage

Abstract

This invention discloses a method for rapid power restoration and load balancing control in power outage areas of distribution networks using energy storage collaboration. Relating to the fields of power system automation and distribution network fault recovery, this method first deploys distributed sensing nodes to comprehensively collect multi-dimensional parameters of the power grid topology, node loads, energy storage operation, and environment. Then, based on topological correlation and load characteristics, it clusters and partitions the data, constructing a power supply priority sequence based on multiple factors. Subsequently, it establishes a dual-objective optimization model that minimizes power restoration time and maximizes load coverage, using an algorithm to solve for the optimal energy storage-power supply combination. Next, supported by high-frequency data acquisition, it dynamically adjusts the energy storage power allocation coefficient to achieve load balancing. Finally, it monitors the power supply status and restoration completion rate through a closed-loop system; if the target is met, it moves to the next partition; otherwise, it adaptively adjusts. This method significantly improves the speed of power restoration, load balancing, and power supply reliability in power outage areas, prioritizing the power supply needs of critical loads.

Description

Technical Field

[0001] This invention relates to the fields of power system automation and distribution network fault recovery, specifically to a method for rapid power restoration and load balancing control in power outage areas of a distribution network in conjunction with energy storage. Background Technology

[0002] As a crucial hub connecting the main grid and end users, the power distribution network is a core infrastructure for ensuring a stable energy supply. Its power supply reliability directly affects many sectors, including industrial production, public services, and people's livelihoods. However, extreme weather, equipment failures, and aging lines can easily cause partial or complete power outages in the distribution network, resulting not only in huge economic losses but also potentially severely impacting the normal operation of critical loads such as medical emergencies and public transportation, threatening public safety and social stability. Therefore, rapid power restoration and load balancing control in areas experiencing power outages have become core technical requirements in the operation and management of distribution networks.

[0003] Traditional power distribution network recovery mainly relies on main grid switching operations or emergency power supply from diesel generators, but this has significant limitations: main grid switching operations require waiting for the faulty lines to be repaired, resulting in a long recovery cycle and difficulty in covering remote or complex topology areas; diesel generators have poor mobility, slow start-up, and also cause noise and environmental pollution problems, and can only meet the temporary power supply needs of small-scale, low-priority loads, and cannot adapt to large-scale, multi-type load collaborative power supply scenarios.

[0004] With the rapid development of energy storage technology, energy storage units, due to their advantages such as flexible charging and discharging and rapid response, have gradually become an important support for emergency power supply in distribution networks. However, existing energy storage-coordinated power supply control technologies still have many shortcomings: First, parameter acquisition dimensions are limited, often focusing only on load power or remaining energy storage capacity, lacking comprehensive perception of key parameters such as grid topology, line impedance, and environmental conditions. Furthermore, inconsistent data transmission delays lead to insufficient decision-making basis. Second, the zoning of power outage areas is simplistic, failing to consider line connection correlation, load density, and energy storage distribution characteristics, easily resulting in overly large zoning areas or unreasonable power supply radii, affecting power supply efficiency. Third, power supply priority ranking lacks multi-factor collaborative consideration, often prioritizing only based on load type, ignoring the influence of energy storage distance, remaining energy capacity, and line impedance. The following factors contribute to the problem: First, the lack of a robust closed-loop monitoring and adaptive adjustment mechanism leads to a single evaluation index for power restoration, which is prone to misjudgment due to instantaneous fluctuations and delayed anomaly handling, thus failing to guarantee the quality of power restoration. Second, the lack of a robust closed-loop monitoring and adaptive adjustment mechanism results in a single evaluation index for power restoration effectiveness, which is prone to misjudgment due to instantaneous fluctuations and delayed anomaly handling, thus failing to guarantee the quality of power restoration.

[0005] In summary, existing technologies are insufficient to achieve rapid, accurate, and balanced power restoration in power distribution network loss areas. There is an urgent need for an energy storage-based collaborative control method that integrates multi-dimensional sensing, scientific zone scheduling, dual-objective optimization, dynamic equilibrium control, and closed-loop monitoring to improve the reliability and efficiency of emergency power supply. Summary of the Invention

[0006] The purpose of this invention is to provide a method for rapid power restoration and load balancing control in power outage areas of a distribution network that integrates energy storage. This method involves collecting multi-dimensional parameters through distributed sensing, performing topological clustering and power supply priority ranking, solving for the optimal energy storage and power supply combination using a dual-objective optimization model, dynamically adjusting the energy storage power allocation coefficient to achieve load balancing, and relying on closed-loop monitoring to switch to the next zone for power supply if the target is met, and adjusting if the target is not met, ultimately achieving rapid and balanced power restoration in power outage areas of the distribution network.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for rapid power restoration and load balancing control in power outage areas of a distribution network with energy storage coordination, characterized by the following steps:

[0009] S1: Conduct power outage state perception and information collection, deploy distributed sensing nodes to implement real-time monitoring of the entire distribution network, and comprehensively collect the power grid topology parameters, load parameters of each node, energy storage unit operating parameters and environmental parameters of the power outage area;

[0010] S2: Based on the topology parameters and load importance collected in the previous step, the power outage area is partitioned, a power supply priority evaluation model is constructed to determine the power supply order of each partition, and a power supply priority sequence is formed;

[0011] S3: Based on the power supply priority sequence, construct an energy storage and power supply combination optimization model with the dual objectives of the shortest power supply recovery time and the largest power supply coverage load. After introducing constraints, solve the optimal energy storage and power supply combination scheme through the algorithm.

[0012] S4: Based on the optimal energy storage power supply combination scheme, the load power and energy storage output power of each node are collected in real time during the power supply process, a dynamic load balancing control model is constructed, and load balancing is achieved by adjusting the output power distribution coefficient of the energy storage unit.

[0013] S5: Real-time monitoring of power supply voltage, current, frequency and load operation status of each node in the power outage area, calculation of power supply restoration completion rate. If both the completion rate and load balance rate meet the standards, the current zone power supply restoration is determined to be complete and the process is transferred to the next priority zone power supply restoration process. If the standards are not met, the process returns to the previous step to readjust the energy storage power allocation coefficient until the requirements are met.

[0014] Distributed sensing nodes refer to basic units in a distributed network architecture that have the ability to collect (sensing) environmental information, process data, and communicate over short distances. Each node has built-in sensors (such as temperature, pressure, current, vibration sensors, etc.) and can independently collect physical / electrical signals at its location.

[0015] Step S1 specifically includes: uniformly deploying distributed sensing nodes at distribution network line nodes, energy storage unit interfaces, and load ends to synchronously monitor power outage areas of the distribution network in real time and collect multi-dimensional core parameters; among them, the power grid topology parameters specifically include a line impedance matrix and a node connection relationship matrix. In the node connection relationship matrix, a value of 1 indicates that the corresponding two nodes are directly connected through a distribution line, and a value of 0 indicates that the two nodes have no direct connection or the line is disconnected due to a fault; the load parameters specifically include the real-time load power, load impedance, and load importance level of each node. The load importance level is divided according to the load type: level 1 is general civil load, level 2 is commercial load, level 3 is ordinary industrial load, level 4 is public service load, and level 5 is medical emergency load; the energy storage unit operating parameters specifically include energy storage capacity, real-time remaining power, upper limit of charging and discharging power, and charging and discharging efficiency; the environmental parameters specifically include real-time wind speed and light intensity. All collected data are preprocessed locally by the sensing nodes and then synchronously transmitted to the control center to ensure data transmission delay consistency.

[0016] Step S2 specifically includes: Based on the power grid topology parameters and load importance levels collected in step S1, the power outage area is first partitioned using topology clustering. Specifically, the power outage area is partitioned based on line connectivity, combined with load density and distance to energy storage units for clustering. The clustering termination condition is that the number of nodes in the partition is ≤30 or the maximum coverage radius of the partition is ≤3km; and then the topology connectivity is used to further analyze the data. Load density similarity Correlation with energy storage The clustering partitions are determined jointly, including the topological correlation. ;when When, impedance The smaller the value, the higher the topological correlation. When the topological correlation is 0, the nodes must belong to different partitions;

[0017] The load correlation , Let i be the load density of node i. The higher the similarity, the closer the correlation is to 1, representing the maximum load density in the power outage area; energy storage correlation. , Let be the distance from node i to the nearest energy storage unit. The maximum energy storage distance in the power outage area is considered; the smaller the distance difference, the closer the correlation is to 1. The final comprehensive correlation is the weighted sum of the three factors. The final comprehensive association is determined by prioritizing topological correlation to ensure connectivity; load correlation is secondary to ensure load balancing across zones. The initial clustering seed node selection rules are: prioritize nodes with load importance levels of 4-5; if multiple high-level load nodes exist, prioritize nodes that are "close to the energy storage unit" and have "high topological correlation"; and the distance between seed nodes must be ≥1km. The clustering termination condition is that merging stops when any one of the following is met:

[0018] 1. The number of nodes within a partition is ≤30;

[0019] 2. The maximum coverage radius of the zone is ≤3km;

[0020] 3. Overall correlation of nodes to be merged

[0021] After partitioning, a power supply priority evaluation model for the power outage area is constructed. This model comprehensively considers four core factors: load importance level, distance between the node and the nearest energy storage unit, real-time remaining power of the energy storage unit, and line impedance between nodes. Weighting coefficients are introduced to perform weighted calculations on each factor. The load importance level has a weight of no less than 0.4 to ensure priority power supply to critical loads; the distance between the node and the energy storage unit has a weight of no less than 0.2 to ensure power supply efficiency. After calculating the priority values ​​of each partition through the above evaluation model, all partitions within the power outage area are sorted in descending order of their values ​​to form a unique power supply priority sequence. This sequence is synchronously stored in the power supply dispatching system as the core basis for subsequent energy storage dispatching.

[0022] The calculation function of the power supply priority evaluation model for power outage areas is:

[0023] ;

[0024] In the formula, For the first The power supply priority value for each partition; the higher the value, the earlier the power supply order. For the first The overall weighting coefficient of each partition, This is a weighting factor representing the proportion of load power. For the first Each node at time... Real-time load power; This is the set of all load nodes within the power outage area. This indicates traversing all load nodes; For the first Each node at time... Real-time load power; This represents the weighting coefficient for the energy storage distance impact factor; For the first The straight-line distance from the core node to the nearest energy storage unit within each partition; This represents the maximum distance from all nodes within the power outage area to their nearest energy storage unit. This is a weighting factor representing the percentage of remaining energy storage capacity. For the first Each energy storage unit at time Real-time remaining battery power; For the first The maximum remaining power of each energy storage unit; This is the weighting factor for the influence of line impedance; For the first The total number of nodes contained in each partition; The elements of the node connection matrix, where a value of 1 indicates the node... The node and the first The nodes are directly connected by a power distribution line. A value of 0 indicates that the two nodes are not directly connected by a power distribution line or that the line is disconnected due to a fault. The elements of the line impedance matrix represent the first... The node and the first The impedance value of the direct connecting line between nodes; This is a correction factor, ranging from 0.05 to 0.1, used to avoid the denominator of the formula calculation being zero or the result being distorted due to the sum of the line impedances being zero or the load power being abnormal.

[0025] Step S3 specifically includes: Based on the power supply priority sequence obtained in step S2, constructing an energy storage collaborative power supply combination optimization model for each power-supply zone according to the principle of high priority first, then low priority. This model takes the shortest power supply recovery time and the largest power supply coverage load as dual optimization objectives. The power supply recovery time specifically includes the sum of line deployment time, energy storage unit start-up time, and power adjustment arrival time. The power supply coverage load specifically refers to the total load power that can be stably supplied, including the rated load and 10% of the standby load capacity. The constraints of the model specifically include: firstly, the total output power of energy storage after line loss is not less than the current power supply zone. The optimization model is designed to meet the following criteria: 1) the total load demand of the power distribution zone; 2) the real-time remaining power of the energy storage unit is maintained within a reasonable range of 20%-80%; 3) the charging and discharging power of the energy storage unit does not exceed ±10% of its rated charging and discharging power limit; and 4) the transmission power of the distribution line does not exceed 90% of the line's rated maximum carrying capacity. A non-dominated sorting genetic algorithm is used to solve the above dual-objective optimization model. During the solution process, the energy storage combination scheme population is first initialized, and then candidate schemes are generated through crossover and mutation operations. After sorting by dominance relationship and screening by congestion, the optimal energy storage power supply combination scheme is obtained. The scheme clearly defines the power supply zone, output power, and start-up sequence of each energy storage unit.

[0026] The objective function of the optimization model is:

[0027] ;

[0028] ;

[0029] In the formula: This is the set of nodes in the currently power-receiving zone. To accelerate the deployment of energy storage power supply lines; Based on line impedance The power supply line transmission efficiency; For energy storage units Startup response time; For the first Power supply reachability coefficient of each node, =1 indicates that power is available. =0 indicates that power is not available;

[0030] The constraints of the optimization model include:

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] In the formula: For energy storage units Real-time output power; Δt represents the minimum remaining energy threshold of energy storage unit k; Δt is the power supply time step. The line transmission power between nodes i and j; Let be the maximum transmission power of the j-th line; the above bi-objective optimization model is solved using a non-dominated sorting genetic algorithm to obtain the optimal energy storage power supply combination scheme;

[0036] Step S4 specifically includes: Based on the optimal energy storage power supply combination scheme determined in step S3, the load power of each node and the output power of the energy storage unit are collected in real time at a frequency of 500ms / time through distributed sensing nodes, and a dynamic load balancing control model is constructed. Load balancing is achieved by dynamically adjusting the output power distribution coefficient of the energy storage unit. At the same time, a load balancing evaluation index is defined. This index is obtained by comparing the sum of the relative deviations between the actual load power of each node and the average load power of the entire partition before and after load balancing control. The index value ranges from 0 to 1. The closer the value is to 1, the more uniform the load power distribution of each node and the better the load balancing effect. The power distribution adjustment logic of the dynamic load balancing control model is as follows: Based on the proportion of the real-time load power of each node to the total load power of the partition, the load importance level, and the deviation value of the load balancing evaluation index, the power distribution amount of each energy storage unit to the corresponding node is dynamically adjusted. For nodes with sudden increases in load power or high importance levels, power supply is prioritized. Through real-time correction of the adjustment coefficient, the load balancing evaluation index is ensured to remain stable above 0.8.

[0037] The power allocation adjustment function of the dynamic load balancing control model is:

[0038] ;

[0039] In the formula, For the first Each energy storage unit at time To the Real-time output power allocation of each node; For the first Each energy storage unit at time Total output power; For the first Each node at time... Real-time load power; For the first The load importance weighting coefficient for each node, with values ​​corresponding to the load importance level; For the first Each energy storage unit is responsible for supplying power to all nodes. This is the load balancing adjustment factor; This is a load balancing evaluation metric, with a value ranging from 0 to 1. For the first The power supply coverage area of ​​each energy storage unit is calculated using the following formula: , For set The total number of nodes within.

[0040] The definition of load balancing evaluation metrics The calculation formula is as follows:

[0041] ;

[0042] In the formula, This is a load balancing evaluation index, with a value ranging from 0 to 1. The closer the value is to 1, the more evenly the load distribution. This represents the total number of nodes within the current evaluation partition. Number the nodes; For the first Each node at time... Real-time load power; For all nodes within the current partition at time... Average load power; For the first Each node at time... Initial load power before load balancing control is implemented.

[0043] Step S5 specifically includes: setting up monitoring points at the beginning, end, and key load nodes of the distribution network line to monitor the power supply voltage, current, frequency, and load operation status of each node in the power outage area in real time; uploading the deviation between the monitoring data and standard parameters to the control center in real time; calculating the power supply restoration completion rate, which is obtained by statistically analyzing the total load power of the nodes that are currently powered normally and comparing it with the total load power of all zones in the power outage area. The calculation uses a 1-minute time window for sliding average to avoid misjudgment caused by instantaneous fluctuations; when the power supply restoration completion rate is not less than 95% and the load is balanced... When the load balance evaluation index is not lower than 0.9, the power supply to the current partition is determined to be restored. The control center automatically calls the power supply priority sequence generated in step S2 and enters the power supply restoration process of the next priority partition. When the power supply restoration completion rate is lower than 95% or the load balance evaluation index is lower than 0.9, the problem type is first located by monitoring data. If it is caused by a sudden increase in load power, the power allocation of the corresponding node is increased. If it is caused by insufficient energy storage output, the energy storage combination is adjusted. Then, the process returns to step S4 to readjust the power allocation coefficient of the energy storage unit and repeats the adjustment monitoring process until both of the above indicators meet the requirements.

[0044] The power restoration calculation formula is as follows:

[0045] ;

[0046] In the formula, The power restoration completion rate is represented by a value ranging from 0 to 1. The closer the value is to 1, the better the power restoration effect. A value of 0.95 or higher is considered to indicate that the power restoration of the current zone has met the standard. It is the set of all load nodes within the power outage area; For the first Each node at time... Real-time load power; For the first Each node at time... The power supply status indicator has a value of 1, which indicates that the node has resumed normal power supply, and a value of 0 indicates that the node has not resumed power supply or the power supply status is abnormal.

[0047] This energy storage-integrated distribution network power restoration and load balancing control method is based on multi-dimensional data acquisition. By deploying distributed sensing nodes at distribution network line nodes, energy storage interfaces, and load ends, it synchronously collects core parameters of the power grid topology, load, energy storage, and environment. After local preprocessing, the data is synchronously transmitted to the control center to provide real-time and consistent data support for subsequent decision-making.

[0048] Based on topological correlation and resource distribution characteristics, the power outage area is scientifically zoned. Taking the tightness of line connection as the core, the zoning method is adopted by combining load density and energy storage distance to ensure the feasibility and efficiency of zoned power supply. Then, a priority evaluation model is constructed that integrates load importance, energy storage distance, remaining energy storage capacity and line impedance. A unique power supply priority sequence is generated through weighted calculation to ensure that critical loads such as medical emergency and public services receive priority power supply resources.

[0049] A dual-objective optimization model is constructed for each priority zone, with "shortest power restoration time" and "maximum power supply coverage load" as the core objectives. Constraints such as energy storage power, remaining power, and line carrying capacity are incorporated. A non-dominated sorting genetic algorithm is used to solve the optimal energy storage power supply combination scheme, clarifying the power supply zone, output power, and start-up sequence of each energy storage unit, thereby achieving efficient scheduling of energy storage resources.

[0050] Real-time dynamic monitoring enables precise load balancing control. The load power of nodes and the output power of energy storage are collected at a frequency of 500ms / time. A dynamic load balancing control model is constructed. The energy storage power allocation coefficient is dynamically adjusted according to the load ratio, importance level and balance deviation of nodes. Power supply is prioritized for nodes with sudden load increases or high importance, ensuring that the relative deviation between the load power of each node and the average load power is minimized.

[0051] Power supply effectiveness is ensured through closed-loop monitoring and adaptive adjustment. Monitoring points are set at the beginning and end of the line and key nodes to track voltage, current, frequency and load operation status in real time. The power supply recovery completion rate is calculated using the 1-minute moving average method. When the recovery completion rate and load balance rate both meet the requirements, the system automatically switches to the next priority zone for power supply. If the requirements are not met, the problem is located through monitoring data, and the power allocation coefficient or energy storage combination is readjusted.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] 1. This invention deploys distributed sensing nodes evenly at distribution network line nodes, energy storage interfaces, and load ends to synchronously collect core parameters of multiple dimensions, including grid topology, load, energy storage unit operation, and environment. After local preprocessing, the data is transmitted synchronously, ensuring data integrity and latency consistency. This provides comprehensive and accurate basic data for subsequent zoning, scheduling, and balance control, solving the decision-making bias problem caused by insufficient data support in existing technologies.

[0054] 2. This invention takes line connection correlation as the core, combines load density and energy storage distance to adopt topological clustering method for partitioning to ensure the feasibility of partitioning; at the same time, it constructs a priority evaluation model that integrates load importance, energy storage distance, remaining energy storage capacity and line impedance to form a unique power supply priority sequence, which not only ensures priority power supply for key loads such as medical emergency and public services, but also takes into account power supply efficiency, and greatly optimizes the power supply resource allocation logic.

[0055] 3. This invention constructs a dual-objective optimization model for each priority zone, which aims to achieve "shortest power restoration time + maximum power supply coverage load". It incorporates multiple constraints such as energy storage output power, remaining power, upper limit of charging and discharging power, and line carrying power. The optimal solution is obtained by using a non-dominated sorting genetic algorithm. The power supply zone, output power, and start-up sequence of the energy storage unit are clearly defined. Compared with the existing technology, this invention significantly improves the utilization rate of energy storage resources and achieves faster power restoration speed and a larger stable load coverage range.

[0056] 4. This invention uses a high-frequency acquisition of node load power and energy storage output power at 500ms / time to construct a dynamic load balancing control model. Based on the node load ratio, importance level, and balance deviation, the energy storage power allocation coefficient is dynamically adjusted to prioritize power supply to nodes with sudden load increases or high importance, ensuring that the load balance is stable above 0.8, effectively solving the problems of uneven load distribution and large power supply fluctuations in existing technologies.

[0057] 5. This invention monitors voltage, current, frequency, and load status in real time at the beginning and end of the line and key nodes. It uses a 1-minute moving average method to calculate the power supply restoration completion rate and combines it with the load balance to form a dual-index evaluation system. When the standard is not met, the problem type can be accurately located, and closed-loop optimization can be achieved by returning to adjust the power distribution coefficient or energy storage combination to avoid misjudgment and adjustment lag, and ensure that the power loss area is fully, stably, and evenly restored. Attached Figure Description

[0058] Figure 1 This is a flowchart of a method for rapid power restoration and load balancing control in power loss areas of a distribution network with energy storage coordination, according to the present invention. Detailed Implementation

[0059] The technical solutions of the present invention will now be described in detail with reference to the accompanying drawings.

[0060] like Figure 1 As shown, step S1 involves power outage perception and information collection, deploying distributed sensing nodes to conduct real-time monitoring of the entire distribution network, and comprehensively collecting grid topology parameters, load parameters of each node, energy storage unit operating parameters, and environmental parameters in the power outage area. Specifically, this includes: uniformly deploying distributed sensing nodes at distribution network line nodes, energy storage unit interfaces, and load ends to synchronously monitor the power outage area of ​​the distribution network in real time and collect multi-dimensional core parameters. The grid topology parameters specifically include a line impedance matrix and a node connection relationship matrix. In the node connection relationship matrix, a value of 1 indicates that the corresponding two nodes are directly connected through a distribution line, and a value of 0 indicates... The system indicates that there is no direct connection between the two nodes or the connection is broken due to a fault. The load parameters specifically include the real-time load power, load impedance, and load importance level of each node. The load importance level is divided according to the load type: Level 1 is general civil load, Level 2 is commercial load, Level 3 is general industrial load, Level 4 is public service load, and Level 5 is medical emergency load. The energy storage unit operation parameters specifically include energy storage capacity, real-time remaining power, upper limit of charging and discharging power, and charging and discharging efficiency. The environmental parameters specifically include real-time wind speed and light intensity. All collected data are pre-processed locally by the sensing nodes and then synchronously transmitted to the control center to ensure data transmission latency consistency.

[0061] Step S2: Based on the topology parameters and load importance collected in the previous step, partition the power outage area and construct a power supply priority evaluation model to determine the power supply order of each partition, forming a power supply priority sequence. Specifically, based on the grid topology parameters and load importance levels collected in Step S1, first, use topology clustering to partition the power outage area. Specifically, partition the power outage area based on line connectivity, combined with load density and distance to energy storage units for clustering. The clustering termination condition is that the number of nodes within a partition is ≤30 or the maximum coverage radius of the partition is ≤3km. Then, use topology connectivity... Load density similarity Correlation with energy storage The clustering partitions are determined jointly, including the topological correlation. ;when When, impedance The smaller the value, the higher the topological correlation. When the topological correlation is 0, the nodes must belong to different partitions;

[0062] The load correlation , Let i be the load density of node i. The higher the similarity, the closer the correlation is to 1, representing the maximum load density in the power outage area; energy storage correlation. , Let be the distance from node i to the nearest energy storage unit. The maximum energy storage distance in the power outage area is considered; the smaller the distance difference, the closer the correlation is to 1. The final comprehensive correlation is the weighted sum of the three factors. The final comprehensive association is determined by prioritizing topological correlation to ensure connectivity; load correlation is secondary to ensure load balancing across zones. The initial clustering seed node selection rules are: prioritize nodes with load importance levels of 4-5; if multiple high-level load nodes exist, prioritize nodes that are "close to the energy storage unit" and have "high topological correlation"; and the distance between seed nodes must be ≥1km. The clustering termination condition is that merging stops when any one of the following is met:

[0063] 1. The number of nodes within a partition is ≤30;

[0064] 2. The maximum coverage radius of the zone is ≤3km;

[0065] 3. Overall correlation of nodes to be merged

[0066] After partitioning, a power supply priority evaluation model for the power outage area is constructed. This model comprehensively considers four core factors: load importance level, distance between the node and the nearest energy storage unit, real-time remaining power of the energy storage unit, and line impedance between nodes. Weighting coefficients are introduced to perform weighted calculations on each factor. The load importance level has a weight of no less than 0.4 to ensure priority power supply to critical loads; the distance between the node and the energy storage unit has a weight of no less than 0.2 to ensure power supply efficiency. After calculating the priority values ​​of each partition through the above evaluation model, all partitions within the power outage area are sorted in descending order of their values ​​to form a unique power supply priority sequence. This sequence is synchronously stored in the power supply dispatching system as the core basis for subsequent energy storage dispatching.

[0067] The calculation function of the power supply priority evaluation model for power outage areas is:

[0068] ;

[0069] In the formula, For the first The power supply priority value for each partition; the higher the value, the earlier the power supply order. For the first The overall weighting coefficient of each partition, This is a weighting factor representing the proportion of load power. For the first Each node at time... Real-time load power; This is the set of all load nodes within the power outage area. This indicates traversing all load nodes; For the first Each node at time... Real-time load power; This represents the weighting coefficient for the energy storage distance impact factor; For the first The straight-line distance from the core node to the nearest energy storage unit within each partition; This represents the maximum distance from all nodes within the power outage area to their nearest energy storage unit. This is a weighting factor representing the percentage of remaining energy storage capacity. For the first Each energy storage unit at time Real-time remaining battery power; For the first The maximum remaining power of each energy storage unit; This is the weighting factor for the influence of line impedance; For the first The total number of nodes contained in each partition; The elements of the node connection matrix, where a value of 1 indicates the node... The node and the first The nodes are directly connected by a power distribution line. A value of 0 indicates that the two nodes are not directly connected by a power distribution line or that the line is disconnected due to a fault. The elements of the line impedance matrix represent the first... The node and the first The impedance value of the direct connecting line between nodes; This is a correction factor, ranging from 0.05 to 0.1, used to avoid the denominator of the formula calculation being zero or the result being distorted due to the sum of the line impedances being zero or the load power being abnormal.

[0070] Step S3: Based on the power supply priority sequence, construct an energy storage-coordinated power supply combination optimization model with the dual objectives of minimizing power supply recovery time and maximizing power supply coverage load. After introducing constraints, solve the optimal energy storage-power supply combination scheme through an algorithm. Specifically, based on the power supply priority sequence obtained in Step S2, construct an energy storage-coordinated power supply combination optimization model for each power supply zone according to the principle of high priority first, then low priority. This model takes the minimum power supply recovery time and the maximum power supply coverage load as the dual optimization objectives. The power supply recovery time specifically includes the sum of line deployment time, energy storage unit start-up time, and power adjustment arrival time. The power supply coverage load specifically refers to the total load power that can be stably supplied, including the rated load and 10% of the reserve load. The constraints of the model include: first, the total output power of energy storage after line losses is not less than the total load demand of the current power supply zone; second, the real-time remaining power of the energy storage unit is maintained within a reasonable range of 20%-80%; third, the charging and discharging power of the energy storage unit does not exceed ±10% of its rated charging and discharging power limit; and fourth, the transmission power of the distribution line does not exceed 90% of the line's rated maximum carrying power. The above bi-objective optimization model is solved using a non-dominated sorting genetic algorithm. In the solution process, the energy storage combination scheme population is first initialized, and then candidate schemes are generated through crossover and mutation operations. After sorting by dominance relationship and screening by congestion, the optimal energy storage power supply combination scheme is obtained. The scheme clearly defines the power supply zone, output power, and start-up sequence of each energy storage unit.

[0071] The objective function of the optimization model is:

[0072] ;

[0073] ;

[0074] In the formula: This is the set of nodes in the currently power-receiving zone. To accelerate the deployment of energy storage power supply lines; Based on line impedance The power supply line transmission efficiency; For energy storage units Startup response time; For the first Power supply reachability coefficient of each node, =1 indicates that power is available. =0 indicates that power is not available;

[0075] The constraints of the optimization model include:

[0076] ;

[0077] ;

[0078] ;

[0079] ;

[0080] In the formula: For energy storage units Real-time output power; Δt represents the minimum remaining energy threshold of energy storage unit k; Δt is the power supply time step. The line transmission power between nodes i and j; Let be the maximum transmission power of the j-th line; the above bi-objective optimization model is solved using a non-dominated sorting genetic algorithm to obtain the optimal energy storage power supply combination scheme;

[0081] Step S4: Based on the optimal energy storage and power supply combination scheme, collect the load power and energy storage output power of each node in real time during the power supply process, construct a dynamic load balancing control model, and achieve load balancing by adjusting the output power allocation coefficient of the energy storage unit; specifically, based on the optimal energy storage and power supply combination scheme determined in Step S3, collect the load power and energy storage unit output power of each node in real time at a frequency of 500ms / time through distributed sensing nodes, construct a dynamic load balancing control model, and achieve load balancing by dynamically adjusting the output power allocation coefficient of the energy storage unit; at the same time, define a load balancing evaluation index, which compares the load power before and after load balancing control, and the load power of each node in real time during the power supply process. The sum of the relative deviations between the actual load power of a node and the average load power of the entire zone is used to determine the index. The index value ranges from 0 to 1. The closer the value is to 1, the more uniform the load power distribution of each node and the better the load balancing effect. The power allocation adjustment logic of the dynamic load balancing control model is as follows: based on the proportion of the real-time load power of each node to the total load power of the zone, the load importance level, and the deviation value of the load balancing evaluation index, the power allocation of each energy storage unit to the corresponding node is dynamically adjusted. For nodes with sudden increases in load power or high importance levels, power supply is prioritized. Through real-time correction of the adjustment coefficient, the load balancing evaluation index is kept stable above 0.8.

[0082] The power allocation adjustment function of the dynamic load balancing control model is:

[0083] ;

[0084] In the formula, For the first Each energy storage unit at time To the Real-time output power allocation of each node; For the first Each energy storage unit at time Total output power; For the first Each node at time... Real-time load power; For the first The load importance weighting coefficient for each node, with values ​​corresponding to the load importance level; For the first Each energy storage unit is responsible for supplying power to all nodes. This is the load balancing adjustment factor; This is a load balancing evaluation metric, with a value ranging from 0 to 1. For the first The power supply coverage area of ​​each energy storage unit is calculated using the following formula: , For set The total number of nodes within.

[0085] The definition of load balancing evaluation metrics The calculation formula is as follows:

[0086] ;

[0087] In the formula, This is a load balancing evaluation index, with a value ranging from 0 to 1. The closer the value is to 1, the more evenly the load distribution. This represents the total number of nodes within the current evaluation partition. Number the nodes; For the first Each node at time... Real-time load power; For all nodes within the current partition at time... Average load power; For the first Each node at time... Initial load power before load balancing control is implemented.

[0088] Step S5: Monitor the power supply voltage, current, frequency and load operation status of each node in the power outage area in real time, calculate the power supply restoration completion rate. If the completion rate and load balance rate both meet the standards, the power supply restoration of the current zone is determined to be completed and the process is transferred to the next priority zone power supply restoration process. If the standards are not met, return to the previous step to readjust the energy storage power allocation coefficient until the requirements are met. Specifically, this includes: setting up monitoring points at the beginning, end, and key load nodes of the distribution network lines to monitor the power supply voltage, current, frequency, and load operation status of each node in the power outage area in real time; uploading the deviation between the monitoring data and standard parameters to the control center in real time; calculating the power restoration completion rate, which is obtained by statistically analyzing the total load power of the nodes that are currently powered normally and comparing it with the total load power of all zones in the power outage area. The calculation uses a 1-minute time window for sliding average to avoid misjudgment caused by instantaneous fluctuations; when the power restoration completion rate is not less than 95% and the load balance evaluation index is not less than 0.9, the power supply of the current zone is determined to be restored, and the control center automatically calls the power supply priority sequence generated in step S2 and moves to the power restoration process of the next priority zone; when the power restoration completion rate is less than 95% or the load balance evaluation index is less than 0.9, the problem type is first located through monitoring data. If it is caused by a sudden increase in load power, the power allocation of the corresponding node is increased; if it is caused by insufficient energy storage output, the energy storage combination is adjusted, and then the process returns to step S4 to readjust the power allocation coefficient of the energy storage unit. The monitoring process is repeated until both of the above indicators meet the requirements.

[0089] The power restoration calculation formula is as follows:

[0090] ;

[0091] In the formula, The power restoration completion rate is represented by a value ranging from 0 to 1. The closer the value is to 1, the better the power restoration effect. A value of 0.95 or higher is considered to indicate that the power restoration of the current zone has met the standard. It is the set of all load nodes within the power outage area; For the first Each node at time... Real-time load power; For the first Each node at time... The power supply status indicator has a value of 1, which indicates that the node has resumed normal power supply, and a value of 0 indicates that the node has not resumed power supply or the power supply status is abnormal.

[0092] A power distribution network in an old residential area suffered a line fault due to heavy rain, resulting in a complete power outage in an area including three residential areas, one community hospital, and two convenience stores. This energy storage-assisted power restoration method was used for emergency response.

[0093] S1: Deploy 12 distributed sensing nodes in the power outage area to collect real-time data on grid topology (including 3 main lines and 15 branch nodes), load parameters (200kW for the community hospital, 800kW for the residential area, and 30kW for each convenience store), operating parameters of distributed energy storage units (3 200kWh energy storage power stations and 5 energy storage modules for electric vehicle charging piles), as well as environmental humidity, temperature, and other data.

[0094] S2: Based on load importance, the core area (community hospital) is of level 1 priority, the living support area (residential area) is of level 2 priority, and the auxiliary area (convenience store) is of level 3 priority, forming a power supply priority sequence.

[0095] S3: Construct a dual-objective optimization model with the objectives of "fastest power restoration + maximum load coverage" and constrain the energy storage output power to ≤150kW / unit. Determine the optimal combination of "2 energy storage power stations supplying the core area + 1 power station + 3 charging piles supplying energy storage to residential area No. 1".

[0096] S4: Real-time acquisition of node load and energy storage output power during power supply. It was found that the start-up of medical equipment in the core area caused load fluctuations. The energy storage output distribution coefficient was adjusted through a dynamic load balancing model, increasing the output ratio of one power station from 40% to 50% to stabilize the load.

[0097] S5: Monitoring shows that the voltage and frequency in the core area meet the standards, the power supply restoration completion rate is 100%, the load balance is 92%, and power supply has been switched to residential areas. The entire process of restoring power to the core area was completed in just 25 minutes, and the entire power outage area was fully restored within 1 hour, with the load balance maintained above 89%.

Claims

1. A method for energy storage coordinated power distribution network power failure area fast power supply recovery and load balancing control, characterized in that, Includes the following steps: S1: Conduct power outage state perception and information collection, deploy distributed sensing nodes to implement real-time monitoring of the entire distribution network, and comprehensively collect the power grid topology parameters, load parameters of each node, energy storage unit operating parameters and environmental parameters of the power outage area; S2: Based on the topology parameters and load importance collected in the previous step, the power outage area is partitioned, a power supply priority evaluation model is constructed to determine the power supply order of each partition, and a power supply priority sequence is formed; S3: Based on the power supply priority sequence, construct an energy storage and power supply combination optimization model with the dual objectives of the shortest power supply recovery time and the largest power supply coverage load. After introducing constraints, solve the optimal energy storage and power supply combination scheme through the algorithm. S4: Based on the optimal energy storage power supply combination scheme, the load power and energy storage output power of each node are collected in real time during the power supply process, a dynamic load balancing control model is constructed, and load balancing is achieved by adjusting the output power distribution coefficient of the energy storage unit. S5: Real-time monitoring of power supply voltage, current, frequency and load operation status of each node in the power outage area, calculation of power supply restoration completion rate. If both the completion rate and load balance rate meet the standards, the current zone power supply restoration is determined to be complete and the process is transferred to the next priority zone power supply restoration process. If the standards are not met, the process returns to the previous step to readjust the energy storage power allocation coefficient until the requirements are met.

2. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 1, characterized in that, Step S1 specifically includes: uniformly deploying distributed sensing nodes at distribution network line nodes, energy storage unit interfaces, and load ends to synchronously monitor power outage areas of the distribution network in real time and collect multi-dimensional core parameters; among them, the grid topology parameters specifically include the line impedance matrix and the node connection relationship matrix. In the node connection relationship matrix, a value of 1 indicates that the corresponding two nodes are directly connected through the distribution line, and a value of 0 indicates that the two nodes have no direct connection or the line is disconnected due to a fault; the load parameters specifically include the real-time load power, load impedance, and load importance level of each node. The load importance level is divided according to the load type: level 1 is general civil load, level 2 is commercial load, level 3 is ordinary industrial load, level 4 is public service load, and level 5 is medical emergency load; the energy storage unit operating parameters specifically include energy storage capacity, real-time remaining power, upper limit of charging and discharging power, and charging and discharging efficiency; the environmental parameters specifically include real-time wind speed and light intensity. All collected data are preprocessed locally by the sensing nodes and then synchronously transmitted to the control center to ensure data transmission delay consistency.

3. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 1, characterized in that, Step S2 specifically includes: Based on the power grid topology parameters and load importance levels collected in step S1, the power outage area is first partitioned using topology clustering. Specifically, the power outage area is partitioned based on line connectivity, combined with load density and distance to energy storage units. The clustering termination condition is that the number of nodes in the partition is ≤30 or the maximum coverage radius of the partition is ≤3km. After partitioning, a power supply priority evaluation model for the power outage area is constructed, comprehensively considering four core factors: load importance level, distance between the node and the nearest energy storage unit, real-time remaining power of the energy storage unit, and line impedance between nodes. Weighting coefficients are introduced to perform weighted calculations on each factor, with the load importance level having a weight of no less than 0.4 to ensure priority power supply for critical loads; and the distance between the node and energy storage having a weight of no less than 0.2 to ensure power supply efficiency. After calculating the priority values ​​of each partition through the above evaluation model, all partitions in the power outage area are sorted in descending order of the values ​​to form a unique power supply priority sequence. This sequence is synchronously stored in the power supply dispatching system as the core basis for subsequent energy storage dispatching.

4. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 3, characterized in that, The calculation function of the power supply priority evaluation model for power outage areas is: ； In the formula, is the power supply priority value of the first partition, and the higher the value, the earlier the power supply sequence; is the comprehensive weight coefficient of the first partition, is the weighting coefficient of the load power proportion; is the real-time load power of the first node at time ; is the set of all load nodes in the power failure area, represents traversing all load nodes; is the real-time load power of the first node at time ; is the weighting coefficient of the energy storage distance influence factor; is the straight-line distance from the core node in the first partition to the nearest energy storage unit; is the maximum value of the distance from all nodes in the power failure area to the corresponding nearest energy storage unit; is the weighting coefficient of the energy storage remaining capacity proportion; is the real-time remaining capacity of the first energy storage unit at time ; is the maximum remaining capacity of the first energy storage unit; is the weighting coefficient of the line impedance influence; is the total number of nodes included in the first partition; is an element of the node connection relationship matrix, and when the value is 1, it indicates that the first node and the first node are directly connected through the distribution line, and when the value is 0, it indicates that there is no direct communication line between the two nodes or the line is faulted and disconnected; is an element of the line impedance matrix, representing the impedance value of the direct communication line between the first node and the first node; is a correction coefficient, with a value range of 0.05-0.1, used to avoid the denominator being zero or the result being distorted due to the total line impedance being zero or the load power being abnormal.

5. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 1, characterized in that, Step S3 specifically includes: Based on the power supply priority sequence obtained in step S2, constructing an energy storage collaborative power supply combination optimization model for each power-supply zone according to the principle of high priority first, then low priority. This model takes the shortest power supply recovery time and the largest power supply coverage load as dual optimization objectives. The power supply recovery time specifically includes the sum of line deployment time, energy storage unit start-up time, and power adjustment arrival time. The power supply coverage load specifically refers to the total load power that can be stably supplied, including the rated load and 10% of the standby load capacity. The constraints of the model specifically include: firstly, the total output power of energy storage after line loss is not less than the current power supply zone. The optimization model is designed to meet the following criteria: 1) the total load demand of the power distribution zone; 2) the real-time remaining power of the energy storage unit is maintained within a reasonable range of 20%-80%; 3) the charging and discharging power of the energy storage unit does not exceed ±10% of its rated charging and discharging power limit; and 4) the transmission power of the distribution line does not exceed 90% of the line's rated maximum carrying capacity. A non-dominated sorting genetic algorithm is used to solve the above dual-objective optimization model. During the solution process, the energy storage combination scheme population is first initialized, and then candidate schemes are generated through crossover and mutation operations. After sorting by dominance relationship and screening by congestion, the optimal energy storage power supply combination scheme is obtained. The scheme clearly defines the power supply zone, output power, and start-up sequence of each energy storage unit.

6. The energy storage coordinated power distribution network power loss area fast service restoration and load balancing control method of claim 1, wherein, Step S4 specifically includes: Based on the optimal energy storage power supply combination scheme determined in step S3, the load power of each node and the output power of the energy storage unit are collected in real time at a frequency of 500ms / time through distributed sensing nodes, and a dynamic load balancing control model is constructed. Load balancing is achieved by dynamically adjusting the output power distribution coefficient of the energy storage unit. At the same time, a load balancing evaluation index is defined. This index is obtained by comparing the sum of the relative deviations between the actual load power of each node and the average load power of the entire partition before and after load balancing control. The index value ranges from 0 to 1. The closer the value is to 1, the more uniform the load power distribution of each node and the better the load balancing effect. The power distribution adjustment logic of the dynamic load balancing control model is as follows: Based on the proportion of the real-time load power of each node to the total load power of the partition, the load importance level, and the deviation value of the load balancing evaluation index, the power distribution amount of each energy storage unit to the corresponding node is dynamically adjusted. For nodes with sudden increases in load power or high importance levels, power supply is prioritized. Through real-time correction of the adjustment coefficient, the load balancing evaluation index is ensured to remain stable above 0.

8.

7. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 6, characterized in that, The power allocation adjustment function of the dynamic load balancing control model is: ； In the formula, For the first Each energy storage unit at time To the Real-time output power allocation of each node; For the first Each energy storage unit at time Total output power; For the first Each node at time... Real-time load power; For the first The load importance weighting coefficient for each node, with values ​​corresponding to the load importance level; For the first Each energy storage unit is responsible for supplying power to all nodes. This is the load balancing adjustment factor; This is a load balancing evaluation index, with a value range of 0 to 1. For the first The power supply coverage area of ​​each energy storage unit is calculated using the following formula: , For set The total number of nodes within.

8. The method for rapid power restoration and load balancing control in power outage areas of a distribution network with energy storage coordination as described in claim 6, characterized in that, The definition of the load balancing degree evaluation index The calculation formula is as follows: ； In the formula, is a load balancing evaluation index, the value range is 0 to 1, the value is closer to 1, the more uniform the load distribution is; is the total number of nodes in the current evaluation partition; is the node number; is the real-time load power of the th node at time ; is the average load power of all nodes in the current partition at time ; is the initial load power of the th node at time before the load balancing control is not carried out.

9. The energy storage coordinated power distribution network power loss area fast service restoration and load balancing control method of claim 1, wherein, Step S5 specifically includes: setting up monitoring points at the beginning, end, and key load nodes of the distribution network line to monitor the power supply voltage, current, frequency, and load operation status of each node in the power outage area in real time; uploading the deviation between the monitoring data and standard parameters to the control center in real time; calculating the power supply restoration completion rate, which is obtained by statistically analyzing the total load power of the nodes that are currently powered normally and comparing it with the total load power of all zones in the power outage area. The calculation uses a 1-minute time window for sliding average to avoid misjudgment caused by instantaneous fluctuations; when the power supply restoration completion rate is not less than 95% and the load is balanced... When the load balance evaluation index is not lower than 0.9, the power supply to the current partition is determined to be restored. The control center automatically calls the power supply priority sequence generated in step S2 and enters the power supply restoration process of the next priority partition. When the power supply restoration completion rate is lower than 95% or the load balance evaluation index is lower than 0.9, the problem type is first located by monitoring data. If it is caused by a sudden increase in load power, the power allocation of the corresponding node is increased. If it is caused by insufficient energy storage output, the energy storage combination is adjusted. Then, the process returns to step S4 to readjust the power allocation coefficient of the energy storage unit and repeats the adjustment monitoring process until both of the above indicators meet the requirements.

10. The energy storage coordinated power distribution network power loss area fast power supply recovery and load balancing control method according to claim 9, characterized in that, The power restoration calculation formula is as follows: ； In the formula, is the power supply recovery completion degree, the value range is 0 to 1, the closer to 1 indicates that the power supply recovery effect is better, and when reaching 0.95 and above, it is determined that the current partition power supply recovery meets the standard; is the set of all load nodes in the power failure area; is the real-time load power of the th node at time ; is the power supply state identifier of the th node at time , the value is 1 when the node has recovered normal power supply, and the value is 0 when the node has not recovered power supply or the power supply state is abnormal.

Images