A distributed energy storage double-layer optimization configuration method constrained by weak link evaluation

Abstract

The application is a kind of weak link evaluation constraint distributed energy storage double-layer optimization configuration method, belonging to the technical field of smart grid, first, according to the load, weather and historical fault information, the time period fault risk index is constructed, and the target state of charge reserve of each energy storage unit is determined by the fault risk index;Secondly, the weak link comprehensive evaluation index of distribution network is established from three dimensions of structural weakness, operation weakness and risk concentration degree;Thirdly, the access node, rated power and rated capacity of distributed energy storage are determined in the planning layer with the minimum weak link comprehensive evaluation index as the target, and the energy storage charging and discharging plan is determined in the operation layer with the minimum system comprehensive operation cost as the target;Finally, the forward compensation charging and discharging restriction is implemented for the state of charge reserve shortage of high risk period, and the operation checking result is fed back to the planning layer until the energy storage configuration scheme meeting the requirements of economy, safety and power supply capacity is obtained.

Description

Technical Field

[0001] This invention belongs to the field of smart grid technology, specifically relating to a collaborative optimization method for distributed energy storage site selection, fixed power, fixed capacity and operation control for distribution networks with a high proportion of distributed photovoltaic access in mountainous areas. Background Technology

[0002] Mountainous power distribution networks are typically affected by complex terrain, limited line corridors, and low load density, resulting in a radial or quasi-radial network structure. The redundancy of power supply paths is insufficient, and failures in main lines and critical branches can easily cause widespread power outages. Furthermore, residential areas, public service loads, and distributed power sources are highly dispersed in mountainous regions. The location of distributed photovoltaic (PV) installations is limited not only by solar resources but also by terrain, installation, and grid connection conditions, easily leading to spatial mismatches between power sources and loads. As the penetration rate of distributed PV increases, the bidirectional power flow and temporal fluctuations in the distribution network become more pronounced, potentially causing problems such as voltage exceeding limits, line overload, and increased network losses at local nodes.

[0003] Existing distributed energy storage configuration methods mainly target investment costs, main grid power purchase costs, grid losses, or renewable energy absorption rates. While these methods can improve the economic operation of distribution networks to some extent, they still have the following shortcomings: First, existing assessments of weak links are usually used independently for risk analysis, condition monitoring, or resilience assessment. The assessment results lack direct coupling with energy storage site selection, power and capacity determination processes, making it difficult to achieve targeted management of main lines, low-redundancy lines, and voltage-weak areas.

[0004] Second, some energy storage configuration methods describe system safety using only a single voltage index, load rate index, or failure rate index, failing to consider the grid structure, operating status, and risk concentration at the same time. This can easily lead to one-sided evaluation results, resulting in a mismatch between the energy storage configuration location and the actual weak links.

[0005] Third, current energy storage operation and control primarily focus on peak shaving and valley filling, reducing electricity purchase costs, mitigating renewable energy fluctuations, or improving voltage. They fail to translate load levels, weather conditions, and historical fault information into time-specific fault risks, nor do they establish executable state-of-charge (SOC) reserve requirements for high-risk periods. During extreme weather, evening peak hours, or periods of concentrated historical faults, energy storage may already be at a low SOC due to economic dispatch, insufficient to provide adequate energy for critical load power supply and system recovery after a fault.

[0006] Fourth, in the existing planning and operation coordination methods, the planning layer usually only determines the energy storage configuration based on static parameters. The voltage over-limit situation, the degree of improvement of weak links, and the reserve shortage during high-risk periods at the operation layer are difficult to be fed back to the planning layer. This may lead to problems such as insufficient capacity, unreasonable location, or insufficient supply guarantee capacity in the actual operation verification of the configuration scheme.

[0007] Therefore, existing technologies urgently need a new technical solution to address the above problems. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a two-layer optimization configuration method for distributed energy storage with weak link evaluation constraints. This method overcomes the problems of existing energy storage configuration methods that emphasize economic efficiency, are disconnected from weak link evaluation and site selection and capacity determination, have insufficient energy storage reserves during high-risk periods, and lack operational feedback verification of planning results.

[0009] To address the aforementioned technical problems, this invention provides a two-layer optimal configuration method for distributed energy storage with weak link evaluation constraints, comprising the following steps, which are performed sequentially. Step 1: Obtain the network topology, line parameters, node loads, predicted output of distributed photovoltaic power, conventional power parameters, upper-level grid switching power limits, time-of-use tariffs, weather levels, historical fault records, and energy storage device parameters of the distribution network to be planned; determine the set of candidate nodes that are allowed to install distributed energy storage based on site conditions, access conditions, and electrical constraints. Step 2: Construct a time-period failure risk index, including the load factor index for time period t, the weather index for time period t, the historical failure rate index for time period t, construct a linear intermediate risk quantity, and the failure risk index for time period t; Step 3: Construct a target state of charge reserve model for high-risk periods to obtain a set of high-risk periods that require key protection; Step 4: Implement forward compensation charging and high-risk period discharge restriction. First, compensate charging in the low-risk preceding period, and then restrict discharge in the high-risk period to ensure that the energy storage meets the target state of charge reserve and reduce the disturbance to the original economic dispatch plan. Step 5: Construct structural weakness indicators, operational weakness indicators, and risk concentration indicators to form a comprehensive evaluation index for weak links; Step Six: Construct a two-layer optimization model for planning and operation. The planning layer optimization control model is established with the goal of minimizing the comprehensive evaluation index of the weak links in the distribution network. The operation layer is constructed with the goal of minimizing the overall system cost under the conditions of energy storage location, power and capacity given by the planning layer. It also satisfies the constraints of power flow equations, node voltage constraints, system active and reactive power balance constraints, energy storage operation constraints, curtailed solar power constraints and upper-level grid exchange power constraints. Step 7: Transmit the energy storage access nodes, rated power, and rated capacity output from the planning layer to the operation layer; after completing economic dispatch, risk identification, forward compensation charging, discharge limitation, and power flow verification, the operation layer outputs node voltage, line operational vulnerability, reserve shortage during high-risk periods, and comprehensive cost, and recalculates the comprehensive evaluation index of vulnerability. If a node voltage exceeds the limit, the reserve shortage during high-risk periods is not eliminated, or the change in the comprehensive evaluation index of the weak link between two adjacent iterations exceeds the preset convergence threshold, the operation verification results will be fed back to the planning layer to update the energy storage location, power, and capacity. If the voltage, state of charge, power, and convergence conditions are met, the final configuration scheme and operation plan will be output.

[0010] The load factor index for time period t mentioned in step two for: , In the formula, The total active power load of the entire network during time period t; and These are the minimum and maximum values ​​of the total active power load of the entire network during the assessment period, respectively. Weather indicators for the time period t for: , In the formula, Let t be the weather level for time period t. The worse the weather, the higher the weather level. The larger; The highest preset weather level; The historical failure rate index for the time period t for: , In the formula, The average number of failures within time period t in the historical statistical period; This represents the maximum average number of failures across all time periods on a typical day. The construction of linear intermediate risk quantity : , In the formula, , and These are the weighting coefficients for load factor, weather, and historical failure rate, respectively. For bias terms; The fault risk index during the time period t The linear intermediate risk is mapped using the Sigmoid function as follows: , In the formula, k is the slope adjustment coefficient, which is used to adjust the sensitivity of the risk index to the intermediate risk level.

[0011] The set of high-risk periods requiring special protection as described in step three is as follows: , In the formula, The high-risk threshold The formula for reserve deficit is: , when When, it indicates that there is insufficient reserves during time period t; In the formula: The target state of charge of energy storage unit h in time period t; This refers to the energy storage state of charge, which is updated hourly.

[0012] The formula for the hourly update of the energy storage state of charge is: , In the formula, Let h be the self-discharge coefficient of the energy storage unit; For time step; and These are charging efficiency and discharging efficiency, respectively. and These are charging power and discharging power, respectively. This is the rated capacity.

[0013] Step 5 addresses the weaknesses in the distribution network structure, including improving electrical betweenness and path redundancy. An improved electrical betweenness is constructed for line l considering the source load size and electrical coupling relationship: , In the formula, and These are the sets of power supply nodes and the sets of load nodes, respectively. and These are the number of power supply nodes and the number of load nodes, respectively. and They are nodes Power supply active power and nodes The active power of the load; For power nodes To load node Equivalent admittance; To be in node pair After applying a unit current between the lines The current caused by the above; After removing line l, define the line path redundancy: , In the formula, The set of nodes affected after removing line l; This represents the number of independent paths from power node s to node k in the original network. The number of independent paths from node s to node k after removing line l; Normalizing the improved electrical betweenness and path redundancy yields the distribution network structural weaknesses: , In the formula, Total number of lines; and As the structural indicator weights, and .

[0014] The method for constructing operational vulnerability indicators as described in step five is as follows: The node set is divided into a condition source-side node set A and a load node set B. The node admittance matrix is ​​then divided into sets. The blocks are as follows: , Define the coupling coefficient matrix: , For any load node Define a dimensionless index for voltage stability margin: , In the formula Indicates source-side node For load nodes The coupling coefficient; Map node metrics to lines The line voltage stability sensitivity was obtained as follows: , Normalize the range of each line during the same time period: , The average value over the entire assessment period is used to determine the operational vulnerabilities of the distribution network. , In the formula, E is the set of lines, and T is the total number of running segments.

[0015] The risk concentration index mentioned in step five is: Define the voltage exceedance limit of node i in time period t: , In the formula, This represents the actual voltage at the node. and These are the lower and upper limits of the allowable node voltage, respectively. The Gini coefficient is used to describe the concentration of voltage exceedances across nodes: , The risk concentration is obtained by averaging over the operating cycle: , In the formula: T is the total number of runtime segments.

[0016] The comprehensive evaluation index for the weak links mentioned in step five is as follows: , In the formula, , and For comprehensive weighting, and The comprehensive weights are determined using the analytic hierarchy process (AHP), entropy weighting, or a combination of weighting methods. The larger the value, the higher the overall weakness of the power distribution network.

[0017] The method for constructing and running the two-level optimization model described in step six is ​​as follows: The planning level aims to minimize the comprehensive evaluation index of weak links in the power distribution network. , The decision variables at the planning level include energy storage access nodes, rated power, and rated capacity, and satisfy the following: , In the formula, and These are the rated power and rated capacity of the energy storage configured for node i, respectively; underlined and overlined variables represent the corresponding lower and upper limits, respectively; For the set of candidate nodes; Given the energy storage location, power, and capacity conditions provided by the planning layer, the operation layer aims to minimize the overall system cost. The converted cost of energy storage investment is as follows: , Network loss cost is: , The cost of purchasing electricity from the main grid is: , In the formula: For energy storage investment costs; For network loss costs; Main grid electricity purchase cost; Rated power of energy storage; This refers to the unit price of investment for energy storage capacity; This refers to the unit price of investment for energy storage capacity. This refers to the rated capacity of the energy storage. This refers to network loss. Price for network loss; Main grid electricity purchase price; The main purpose is to purchase electricity online.

[0018] The energy storage equipment parameters mentioned in Step 1 include at least the upper and lower limits of rated power, the upper and lower limits of rated capacity, charge and discharge efficiency, self-discharge coefficient, upper and lower limits of state of charge, unit price of power investment, and unit price of capacity investment.

[0019] The beneficial effects of this invention are as follows: By constructing a time-period fault risk index, this invention transforms load levels, severe weather, and historical fault frequencies into target state-of-charge reserves for energy storage; by improving electrical betweenness, path redundancy, voltage stability margin, and voltage exceedance concentration, it constructs a comprehensive evaluation index for weak links; and by determining the location, power, capacity, and charging / discharging plan of energy storage through a two-layer iterative "planning-operation" process, it achieves the following objectives: accurately identifying main distribution network lines, low-redundancy lines, voltage-weak areas, and risk-concentrated areas; guiding energy storage to be preferentially deployed at key nodes that can improve system weaknesses; ensuring that energy storage has energy reserves to meet safety and supply requirements during high-risk periods; eliminating voltage exceedances at an acceptable incremental investment cost, improving power supply reliability and system resilience; and achieving closed-loop coordination between energy storage planning decisions and actual operation control.

[0020] Further beneficial effects of the present invention are as follows: 1. Targeted management of weak links. This invention incorporates improvements in electrical betweenness, path redundancy, voltage stability margin, and voltage over-limit concentration into energy storage planning. This can simultaneously reflect line power flow load, post-fault power transfer capacity, the degree to which nodes approach voltage instability boundaries, and the spatial concentration of over-limit risks, avoiding site selection bias caused by relying on a single indicator.

[0021] 2. Establishing executable safety and supply reserve constraints. This invention transforms load, weather, and historical fault information into a continuous fault risk index, and dynamically raises the target state of charge based on the risk index, so that the safety and supply demand during high-risk periods can directly affect the energy storage charging and discharging plan.

[0022] 3. Reduce disturbances to economic dispatch. Forward compensation charging prioritizes the use of remaining charging margins in low-risk preceding periods, and only restricts discharge in high-risk periods after the compensation capacity reaches its physical limit, thereby improving safety reserves while maintaining the original economic dispatch results as much as possible.

[0023] 4. Achieve closed-loop coordination between planning and operation. The results of voltage overruns, reserve shortages, and vulnerability improvements at the operation level can be fed back to the planning level, avoiding problems such as insufficient capacity or unreasonable location in actual operation due to static planning results.

[0024] 5. Balancing safety and economy. In the example system, compared with the configuration method that only considers economy, the energy storage configuration cost of the present invention increases by only about 6.83%, but node voltage overshoot is eliminated, the system voltage is maintained at 0.96 to 1.01 pu, and the average state of charge of energy storage during high-risk periods is improved by 11.3 percentage points, which can significantly enhance the power distribution network's safety and supply capabilities under the condition of acceptable economy. Attached Figure Description

[0025] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0026] Figure 1 This is a diagram illustrating the overall architecture of the distributed energy storage two-layer optimized configuration model of this invention. Figure 2 This is a schematic diagram illustrating the reserve deficit identification between the reference state of charge and the target state of charge in this invention. Figure 3 This is a flowchart illustrating the energy storage configuration and operation optimization implementation of the present invention; Figure 4 This is a topology diagram of the IEEE-33 node power distribution system used in an embodiment of the present invention; Figure 5 A comparison chart showing the energy storage site selection results using the method of this invention, with only an economical configuration scheme as a specific implementation method. Figure 6 This is a comparison chart of the average state of charge of each energy storage unit during high-risk periods under two specific implementation schemes. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] To solve the above-mentioned technical problems, the present invention includes the following steps: Step S1: Acquisition of basic data and determination of candidate node set Obtain the network topology, line parameters, node loads, predicted output of distributed photovoltaic power, conventional power parameters, upper-level grid switching power limits, time-of-use tariffs, weather levels, historical fault records, and energy storage device parameters of the distribution network to be planned; determine the set of candidate nodes that are allowed to install distributed energy storage based on site conditions, access conditions, and electrical constraints.

[0029] The parameters of the energy storage device include at least the upper and lower limits of rated power, the upper and lower limits of rated capacity, charge and discharge efficiency, self-discharge coefficient, upper and lower limits of state of charge, unit price of power investment, and unit price of capacity investment.

[0030] Step S2: Construct a time-period failure risk index Define the load factor index for time period t. for: In the formula, The total active power load of the entire network during time period t; and These are the minimum and maximum values ​​of the total active power load of the entire network during the assessment period, respectively.

[0031] Define weather indicators for time period t for: In the formula, Let t be the weather level for time period t. The worse the weather, the higher the weather level. The larger; This is the preset highest weather level.

[0032] Define the historical failure rate index for time period t for: In the formula, The average number of failures within time period t in the historical statistical period; This represents the maximum average number of failures per time period on a typical day.

[0033] Constructing a linear intermediate risk quantity : In the formula, , and These are the weighting coefficients for load factor, weather, and historical failure rate, respectively. This is a bias term.

[0034] The Sigmoid function is used to map the linear intermediate risk quantity to a time-period failure risk index. : In the formula, This is the slope adjustment coefficient, used to adjust the sensitivity of the risk index to intermediate risk levels.

[0035] according to Risk is categorized by time period: when When it is low risk; when At that time, it was considered a medium-risk period; when At that time, it was high-risk; when This is at an extremely high risk.

[0036] Step S3: Construct a target state of charge reserve model for high-risk periods The additional risk reserve for energy storage unit h in time period t is: The target state of charge of energy storage unit h in time period t is: In the formula, Risk response coefficient; This represents the basic reserve state of charge for energy storage unit h. This represents the upper limit of the state of charge of the energy storage unit.

[0037] The energy storage state of charge is updated hourly according to the following formula: In the formula, Let h be the self-discharge coefficient of the energy storage unit; For time step; and These are charging efficiency and discharging efficiency, respectively. and These are charging power and discharging power, respectively. This is the rated capacity.

[0038] The baseline state of charge obtained from economic dispatch is compared with the target state of charge, and the reserve deficit is defined as: when When, it indicates that there is insufficient reserves during the period t.

[0039] Let the high-risk threshold be The high-risk periods that require special attention are: Preferably, .

[0040] Step S4: Implement forward compensation charging and discharge limitation during high-risk periods. For any high-risk period The state-of-charge deficit is converted into the amount of electricity required to compensate the grid side: Before time period t, candidate time periods with risk below the threshold and still having charging margin are selected to form a set of candidate charging time periods: In the formula, The benchmark charging power under economic dispatch, This is the upper limit of the charging power for energy storage.

[0041] Candidate period The margin for additional charging power is: Following the principle of minimum disturbance, compensation power is preferentially allocated to candidate time periods with larger recharge margins, while satisfying the following: as well as: The compensated charging power is After recalculating the state of charge, the remaining reserve deficit is defined as follows: In the formula, The state of charge after forward compensation charging.

[0042] when At that time, the baseline discharge power for high-risk periods is reduced, and the adjusted discharge power is: in: By first compensating for charging during low-risk periods and then restricting discharge during high-risk periods, energy storage can meet the target state of charge reserve as much as possible, while reducing disturbances to the original economic dispatch scheme.

[0043] Step S5: Construct structural weakness indicators An improved electrical betweenness is constructed for line l, taking into account the source load size and electrical coupling relationship: In the formula, G and D are the set of power supply nodes and the set of load nodes, respectively; and These are the number of power supply nodes and the number of load nodes, respectively. and These are the active power of the power supply at node g and the active power of the load at node d, respectively. The equivalent admittance from power node g to load node d; To be in node pair After applying a unit current between the lines The current caused by the above.

[0044] After removing line l, define the line path redundancy: In the formula, The set of nodes affected after removing line l; This represents the number of independent paths from power node s to node k in the original network. This represents the number of independent paths from node s to node k after removing line l.

[0045] Normalizing the improved electrical betweenness and path redundancy yields the distribution network structural weaknesses: In the formula, Total number of lines; and As the structural indicator weights, and .

[0046] Step S6: Construct operational vulnerability indicators The node set is divided into a condition source-side node set A and a load node set B. The node admittance matrix is ​​then divided into sets. The blocks are as follows: Define the coupling coefficient matrix: For any load node Define a dimensionless index for voltage stability margin: Map node metrics to lines The line voltage stability sensitivity was obtained as follows: Normalize the range of each line during the same time period: The average value over the entire assessment period is used to determine the operational vulnerabilities of the distribution network. In the formula, For the collection of routes, This represents the total number of runtime segments.

[0047] Step S7: Construct a comprehensive evaluation index for risk concentration and weak links Define the voltage exceedance limit of node i in time period t: In the formula, This represents the actual voltage at the node. and These represent the lower and upper limits of the node voltage, respectively.

[0048] The Gini coefficient is used to describe the concentration of voltage exceedances across nodes: The risk concentration is obtained by averaging over the operating cycle: A comprehensive evaluation index for weak links in the power distribution network is constructed by considering structural weaknesses, operational weaknesses, and risk concentration. In the formula, , and For comprehensive weighting, and The weights can be determined using the analytic hierarchy process (AHP), entropy weighting, or a combination of weighting methods. The larger the value, the higher the overall weakness of the power distribution network.

[0049] Step S8: Establish a two-tier optimization configuration model of "planning-operation". The planning level aims to minimize the comprehensive evaluation index of weak links in the power distribution network. The decision variables at the planning level include energy storage access nodes, rated power, and rated capacity, and satisfy the following: In the formula, and These are the rated power and rated capacity of the energy storage configured for node i, respectively; underlined and overlined variables represent the corresponding lower and upper limits, respectively; This is the set of candidate nodes.

[0050] Given the energy storage location, power, and capacity conditions provided by the planning layer, the operation layer aims to minimize the overall system cost. The converted cost of energy storage investment is as follows: Network loss cost is: The cost of purchasing electricity from the main grid is: In the formula: For energy storage investment costs; For network loss costs; Main grid electricity purchase cost; Rated power of energy storage; This refers to the unit price of investment for energy storage capacity; This refers to the unit price of investment for energy storage capacity. This refers to the rated capacity of the energy storage. This refers to network loss. Price for network loss; Main grid electricity purchase price; The main purpose is to purchase electricity online.

[0051] The runtime layer must satisfy at least the following constraints.

[0052] 1. Power flow equation constraints: In the formula, , and These represent the injected active power, injected reactive power, and voltage amplitude at node i, respectively. , and These represent the line conductance, susceptance, and voltage phase angle difference between nodes i and j, respectively.

[0053] 2. Node voltage constraints: In the formula, Let be the voltage at node i at time t; These are the upper and lower limits of the node voltage, respectively.

[0054] 3. System active and reactive power balance constraints: In the formula, The active power exchanged between the distribution network and the upstream power grid; The active power output of the conventional power source at node i during time period t; Let be the active power output of the distributed photovoltaic system at node i during time period t; Distributed energy storage discharge power at node i; Active power demand of the load at node i; Distributed energy storage charging power at node i; The active power loss of the distribution network at time t. The reactive power exchanged between the distribution network at node i and the upstream power grid; The reactive power output of the conventional power supply at node i; The reactive power output of the photovoltaic inverter at node i; The reactive power output of the bidirectional inverter at node i; The reactive power provided by the parallel static compensation device at node i; Reactive load at node i; The reactive power loss of the entire network during time period t; N is the number of nodes.

[0055] 4. Energy storage operation constraints: In the formula, These are the upper and lower limits of the energy storage state of charge. These represent the upper limits of the charging and discharging power of the energy storage unit.

[0056] Furthermore, binary state variables or mutual exclusion logic constraints are used to prevent the same energy storage unit from charging and discharging simultaneously during the same period.

[0057] 5. Curtailment power constraint: In the formula, Let represent the curtailed power of energy storage unit h during time period t.

[0058] 6. Upstream power grid switching power constraints: In the formula, These are the upper and lower limits of active and reactive power injected into the distribution network from the upper-level power grid, respectively.

[0059] Step S9: Planning and Operational Iterative Feedback The energy storage access nodes, rated power, and rated capacity output from the planning layer are transmitted to the operation layer. After completing economic dispatch, risk identification, forward compensation charging, discharge limitation, and power flow verification, the operation layer outputs node voltage, line operational vulnerability, reserve shortage during high-risk periods, and comprehensive cost, and recalculates the comprehensive evaluation index of the vulnerability. .

[0060] If a node voltage exceeds the limit, the reserve shortage during high-risk periods is not eliminated, or the change in the comprehensive evaluation index of the weak link between two adjacent iterations exceeds the preset convergence threshold, the operation verification results will be fed back to the planning layer to update the energy storage location, power, and capacity. If the voltage, state of charge, power, and convergence conditions are met, the final configuration scheme and operation plan will be output.

[0061] In one implementation, the planning layer uses a particle swarm optimization algorithm to search for energy storage access nodes, rated power, and rated capacity, while the operation layer uses YALMIP for modeling and calls the CPLEX solver for solving. Alternatively, genetic algorithms, differential evolution algorithms, second-order cone optimization, or other solution methods that can achieve the same technical functions can be used.

[0062] One specific implementation is verified using the IEEE-33 node power distribution system, with the system topology as follows: Figure 4 As shown. The system's base capacity is 10 MVA, the base voltage is 12.66 kV, and the allowable node voltage range is 0.95–1.05 pu. The typical daily total load is 69122.02 kWh, the typical daily total power generation of distributed photovoltaic is 7870.17 kWh, and the photovoltaic access nodes are node 6, node 16, node 25, and node 30.

[0063] Node load, predicted photovoltaic output, electricity price, weather level, and historical fault count are input into the model according to 24 time periods. Weather level and historical fault count are normalized, and the fault risk index for each time period is calculated using the method described in step S2. Based on the risk classification results, H16–H20 are designated as the key high-risk periods in this embodiment, and a high-risk threshold is set. .

[0064] The upper and lower limits of the energy storage state of charge are set to 0.2 and 0.9, respectively. The charge and discharge efficiency is given according to the parameters of the energy storage equipment, and the initial state of charge of the energy storage is determined according to the operating requirements of the first period of a typical day. Candidate nodes are selected by combining site installation conditions, voltage level, and grid-connected capacity.

[0065] To verify the effectiveness of this invention, the following three schemes are proposed: Option 1: Without energy storage, the system relies solely on conventional power sources and electricity purchased from the upstream grid to meet load demands, and performs operational calculations under power flow and voltage constraints.

[0066] Option 2: Configure distributed energy storage without considering vulnerability assessment and safe supply operation strategies. Both planning and operation are primarily guided by economic efficiency. On the planning side, the location, power, and capacity of energy storage are determined to minimize investment and operating costs. On the operation side, charging and discharging scheduling is carried out based on minimizing overall costs.

[0067] Option 3: Employing the method of this invention. The planning layer uses comprehensive evaluation indicators based on weak links. With the goal of minimizing overall operating costs, the operation layer incorporates strategies based on fault risk, such as target state of charge reserves, forward compensation charging, and discharge limitation during high-risk periods.

[0068] The evaluation process for weak links is as follows: First, the improved electrical betweenness of each line is calculated based on the network topology, source-load distribution, and line parameters. Lines carrying a large downstream load, located on the main line, and close to the power source have a larger improved electrical betweenness. Then, each line is removed one by one, and the number of independent power supply paths for each node after a fault is calculated to obtain the path redundancy. For a single power supply channel, the path redundancy is close to zero, indicating a lack of alternative power supply paths after a line fault.

[0069] Secondly, voltage stability margin indices are calculated based on node voltages and node admittance matrices for each time period, and these indices are mapped to adjacent lines. The end with the poorer voltage stability margin among the nodes at both ends of a line determines the operational risk level of that line. The operational vulnerability is obtained by averaging the results over all time periods.

[0070] Next, the voltage exceedance rate of each node is calculated, and the Gini coefficient is used to describe the spatial concentration of exceedance risk. When a few nodes experience significant exceedances while the voltages of other nodes remain normal, the risk concentration is high; when none of the nodes exceed their limits, the risk concentration is zero.

[0071] Finally, a comprehensive evaluation index was obtained by weighting structural weaknesses, operational weaknesses, and risk concentration using preset weights. In this embodiment, the average vulnerability of trunk lines 1 to 7 is approximately 0.62, which is higher than the average level of 0.35 for the entire network. Line 2 carries a large amount of downstream load, with an improved electrical dielectric constant of approximately 0.95. Its 24-hour average load rate is 78.6%, and its load rate during the evening peak period is close to 92%, thus it is identified as a critical vulnerable line.

[0072] The energy storage configuration at the planning layer is as follows: The planning layer encodes the energy storage access location, rated capacity, and rated power as particle positions, and constructs the objective function and constraints according to step S8. For each candidate configuration scheme, the operation layer is invoked to complete the economic scheduling and safety supply verification, and the comprehensive weak link evaluation index is returned. If there are voltage over-limit issues, insufficient reserves during high-risk periods, or failure to converge the index, the particle positions are updated and the search continues.

[0073] The energy storage configuration results obtained using the method of this invention are shown in the table below.

[0074]

[0075] In comparison, Scheme 2 has a total energy storage capacity of 4737.86 kWh and a total power of 2395.06 kW; Scheme 3 has a total capacity that is about 4.5% higher than Scheme 2, and a total power that is about 15.79% higher. The energy storage locations in Scheme 3 are changed from nodes 5, 8, 16, 20, 24, and 25 to nodes 2, 5, 8, 14, 19, and 24, making the energy storage closer to the weak lines, critical branches, and voltage risk areas of the main trunk.

[0076] Given a specific energy storage configuration, the operation layer first aims to minimize the sum of the energy storage investment cost, grid loss cost, main grid power purchase cost, and curtailment cost to solve for the baseline charge-discharge plan. Subsequently, based on load factor, weather level, and historical failure rate, the failure risk index for each time period is calculated to obtain the target state-of-charge curve.

[0077] Taking the energy storage at node 5 as an example, Scheme 3 increases the energy storage capacity of this node from approximately 773 kWh in Scheme 2 to approximately 973 kWh, giving it a longer energy support duration. The operation layer compares the baseline state of charge (SOC) and the target SOC hourly. When a reserve deficit is detected during a high-risk period, charging power is first increased during previously low-risk periods with sufficient charging margin. If forward compensation charging is still insufficient to eliminate the deficit, the planned discharge power during the high-risk period is reduced.

[0078] In this embodiment, Scheme 3 initiates compensatory charging at H4, increasing the state of charge (SBC) at node 5 from approximately 37.6% to approximately 58.8%. At H9, both schemes reach their upper limit of SBC, but due to the larger storage capacity of Scheme 3, its actual stored energy is higher than that of Scheme 2. After discharging at H12, Scheme 3 maintains a SBC of approximately 59.5%, while Scheme 2 drops to approximately 29.4%. From H13 to H16, Scheme 3 maintains a SBC of approximately 40.2%, while Scheme 2 approaches its upper limit. Before entering the high-risk period at H18, Scheme 3 has increased the SBC to approximately 93%, providing more sufficient emergency energy reserves for the evening peak and potential faults.

[0079] After completing the charge and discharge correction, the node voltage, line power flow, network loss, reserve deficit during high-risk periods, and comprehensive evaluation indicators of weak links are recalculated. If the voltage of any node exceeds the allowable range, or if there is still a reserve deficit during any high-risk period that cannot be eliminated through charging compensation and discharging restrictions, the corresponding node and period information is fed back to the planning level to increase the energy storage capacity of key nodes, increase the rated power, or adjust the access location.

[0080] The iteration terminates when the following condition is met: 1. All node voltages are within the allowable range; 2. The energy storage charging and discharging power, state of charge, and power exchanged with the upstream power grid all meet the constraints; 3. The target state of charge reserve meets the preset tolerance during high-risk periods; 4. The two adjacent iterations The change is less than a preset threshold; 5. Overall operating costs are within acceptable limits.

[0081] When only economic considerations are taken into account in Scheme 2, nodes 12 and 13 experience voltage drops below the allowable lower limit during certain periods. After adopting the method of this invention, the system node voltage is maintained between 0.96 and 1.01 pu, meeting the voltage quality requirement of 0.95 to 1.05 pu, and the node voltage exceeding the limit is eliminated.

[0082] The energy storage configuration cost of this invention increases by approximately 6.83% compared to Scheme 2, but the average state of charge (SOC) of the entire system during high-risk periods increases from approximately 55.7% to approximately 67.0%, an increase of 11.3 percentage points. The standard deviation of the SOC decreases from approximately 25.2% to approximately 22.1%, indicating that the fluctuation of energy storage levels is reduced and reserves are more stable during critical periods. The average SOC of all six energy storage units during high-risk periods is higher than that of Scheme 2, with some energy storage units showing an increase of more than 12 percentage points. This indicates that the increase in safety and supply reserves is concentrated during the truly needed high-risk periods, rather than an indiscriminate increase in the overall daily operating plan.

[0083] The above results show that the present invention can transform energy storage configuration from simply pursuing economic optimization to safety-economic synergistic optimization oriented towards improving weak links and ensuring safe supply, with relatively small incremental investment costs, and has good engineering applicability.

[0084] This invention is not limited to the IEEE-33 node system, but can also be applied to actual mountain power distribution networks, rural power distribution networks, island power distribution networks, active power distribution networks in industrial parks, and power distribution systems containing wind power, photovoltaic power, or multiple distributed power sources.

[0085] Weather indicators can be composed of one or more of the following: meteorological warning level, wind speed, rainfall, icing thickness, lightning density, or temperature anomaly level; historical failure rate can be statistically analyzed separately for lines, equipment, or regions; target state of charge can be set differently for important users, critical load levels, or fault recovery time.

[0086] Energy storage types can include lithium-ion batteries, sodium-ion batteries, flow batteries, supercapacitors, or other energy storage devices with bidirectional power regulation capabilities. For multiple types of energy storage, constraints on lifetime loss, cycle life, and power response speed can be further increased.

[0087] Without changing the core technical idea of ​​this invention, the planning layer and the operation layer can be solved using other equivalent algorithms, and the weight of the comprehensive evaluation index of weak links can also be determined based on expert experience, historical data or data-driven methods.

[0088] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0089] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A two-layer optimal configuration method for distributed energy storage with weak link evaluation constraints, characterized in that: Includes the following steps, And the following steps are performed in sequence. Step 1: Obtain the network topology, line parameters, node loads, predicted output of distributed photovoltaic power, conventional power parameters, upper-level grid switching power limits, time-of-use tariffs, weather levels, historical fault records, and energy storage device parameters of the distribution network to be planned; determine the set of candidate nodes that are allowed to install distributed energy storage based on site conditions, access conditions, and electrical constraints. Step 2: Construct a time-period failure risk index, including the load factor index for time period t, the weather index for time period t, the historical failure rate index for time period t, construct a linear intermediate risk quantity, and the failure risk index for time period t; Step 3: Construct a target state of charge reserve model for high-risk periods to obtain a set of high-risk periods that require key protection; Step 4: Implement forward compensation charging and high-risk period discharge restriction. First, compensate charging in the low-risk preceding period, and then restrict discharge in the high-risk period to ensure that the energy storage meets the target state of charge reserve and reduce the disturbance to the original economic dispatch plan. Step 5: Construct structural weakness indicators, operational weakness indicators, and risk concentration indicators to form a comprehensive evaluation index for weak links; Step Six: Construct a two-layer optimization model for planning and operation. The planning layer optimization control model is established with the goal of minimizing the comprehensive evaluation index of the weak links in the distribution network. The operation layer is constructed with the goal of minimizing the overall system cost under the conditions of energy storage location, power and capacity given by the planning layer. It also satisfies the constraints of power flow equations, node voltage constraints, system active and reactive power balance constraints, energy storage operation constraints, curtailed solar power constraints and upper-level grid exchange power constraints. Step 7: Transmit the energy storage access nodes, rated power, and rated capacity output from the planning layer to the operation layer; after completing economic dispatch, risk identification, forward compensation charging, discharge limitation, and power flow verification, the operation layer outputs node voltage, line operational vulnerability, reserve shortage during high-risk periods, and comprehensive cost, and recalculates the comprehensive evaluation index of vulnerability. ; If a node voltage exceeds the limit, the reserve shortage during high-risk periods is not eliminated, or the change in the comprehensive evaluation index of the weak link between two adjacent iterations exceeds the preset convergence threshold, the operation verification results will be fed back to the planning layer to update the energy storage location, power, and capacity. If the voltage, state of charge, power, and convergence conditions are met, the final configuration scheme and operation plan will be output.

2. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The load factor index for time period t mentioned in step two for: , In the formula, The total network active load for time period t; and These are the minimum and maximum values ​​of the total active power load of the entire network during the assessment period, respectively. Weather indicators for the time period t for: , In the formula, Let t be the weather level for time period t. The worse the weather, the higher the weather level. The larger; The highest preset weather level; The historical failure rate index for the time period t for: , In the formula, The average number of failures within time period t in the historical statistical period; This represents the maximum average number of failures across all time periods on a typical day. The construction of linear intermediate risk quantity : , In the formula, , and These are the weighting coefficients for load factor, weather, and historical failure rate, respectively. For bias terms; The fault risk index during the time period t The linear intermediate risk is mapped using the Sigmoid function as follows: , In the formula, k is the slope adjustment coefficient, which is used to adjust the sensitivity of the risk index to the intermediate risk level.

3. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The set of high-risk periods requiring special protection as described in step three is as follows: , In the formula, The high-risk threshold The formula for reserve deficit is: , when When, it indicates that there is insufficient reserves during the period t; In the formula: The target state of charge of energy storage unit h in time period t; This refers to the energy storage state of charge, which is updated hourly.

4. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 3, characterized in that: The formula for the hourly update of the energy storage state of charge is: , In the formula, Let h be the self-discharge coefficient of the energy storage unit; For time step; and These are charging efficiency and discharging efficiency, respectively. and These are charging power and discharging power, respectively. This is the rated capacity.

5. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: Step 5 addresses the weaknesses in the distribution network structure, including improving electrical betweenness and path redundancy. An improved electrical betweenness is constructed for line l considering the source load size and electrical coupling relationship: , In the formula, and These are the sets of power supply nodes and the sets of load nodes, respectively. and These are the number of power supply nodes and the number of load nodes, respectively. and They are nodes Power supply active power and nodes The active power of the load; For power nodes To load node Equivalent admittance; To be in node pair After applying a unit current between the lines The current caused by the above; After removing line l, define the line path redundancy: , In the formula, The set of nodes affected after removing line l; This represents the number of independent paths from power node s to node k in the original network. The number of independent paths from node s to node k after removing line l; Normalizing the improved electrical betweenness and path redundancy yields the distribution network structural weaknesses: , In the formula, Total number of lines; and As the structural indicator weights, and .

6. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The method for constructing operational vulnerability indicators as described in step five is as follows: The node set is divided into a condition source-side node set A and a load node set B. The node admittance matrix is ​​then divided into sets. The blocks are as follows: , Define the coupling coefficient matrix: , For any load node Define a dimensionless index for voltage stability margin: , In the formula Indicates source-side node For load nodes The coupling coefficient; Map node metrics to lines The line voltage stability sensitivity was obtained as follows: , Normalize the range of each line during the same time period: , The average value over the entire assessment period is used to determine the operational vulnerabilities of the distribution network. , In the formula, E is the set of lines, and T is the total number of running segments.

7. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The risk concentration index mentioned in step five is: Define the voltage limit of node i in time period t: , In the formula, This represents the actual voltage at the node. and These are the lower and upper limits of the allowable node voltage, respectively. The Gini coefficient is used to describe the concentration of voltage exceedances across nodes: , The risk concentration is obtained by averaging over the operating cycle: , In the formula: T is the total number of runtime segments.

8. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The comprehensive evaluation index for the weak links mentioned in step five is as follows: , In the formula, , and For comprehensive weighting, and The comprehensive weights are determined using the analytic hierarchy process (AHP), entropy weighting, or a combination of weighting methods. The larger the value, the higher the overall weakness of the power distribution network.

9. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that: The method for constructing and running the two-level optimization model described in step six is ​​as follows: The planning level aims to minimize the comprehensive evaluation index of weak links in the power distribution network. , The decision variables at the planning level include energy storage access nodes, rated power, and rated capacity, and satisfy the following: , In the formula, and These are the rated power and rated capacity of the energy storage configured for node i, respectively; underlined and overlined variables represent the corresponding lower and upper limits, respectively; For the set of candidate nodes; Given the energy storage location, power, and capacity conditions provided by the planning layer, the operation layer aims to minimize the overall system cost. The converted cost of energy storage investment is as follows: , Network loss cost is: , The cost of purchasing electricity from the main grid is: , In the formula: For energy storage investment costs; For network loss costs; Main grid electricity purchase cost; Rated power of energy storage; This refers to the unit price of investment for energy storage capacity; This refers to the unit price of investment for energy storage capacity. This refers to the rated capacity of the energy storage. This refers to network loss. Price for network loss; Main grid electricity purchase price; The main purpose is to purchase electricity online.

10. The distributed energy storage two-layer optimization configuration method with weak link evaluation constraints according to claim 1, characterized in that, The energy storage equipment parameters mentioned in Step 1 include at least the upper and lower limits of rated power, the upper and lower limits of rated capacity, charge and discharge efficiency, self-discharge coefficient, upper and lower limits of state of charge, unit price of power investment, and unit price of capacity investment.

Images