A park power distribution network distributed energy storage optimization configuration method, system, device, medium and product

Abstract

The application discloses a kind of park power distribution network distributed energy storage optimization configuration method, system, equipment, medium and product, it is related to distributed energy storage technical field, obtains the operation data and operation information of park power distribution network, based on preset supply adaptation function and preset new energy consumption space function, according to operation data, determine corresponding node energy storage basic state of charge lower limit and node energy storage basic state of charge upper limit, according to language big model, operation data, operation information is corrected to each node energy storage basic state of charge lower limit and each node energy storage basic state of charge upper limit, obtain multiple dynamic node state of charge constraint, each dynamic node state of charge constraint and operation data are used to optimize and solve distributed energy storage optimization configuration model, and obtain distributed energy storage optimization configuration scheme.The technical problem that it is easy to appear emergency power shortage and reduce the reliability of park power distribution network operation is solved in the park distributed energy storage optimization configuration method.

Description

Technical Field

[0001] This invention relates to the field of distributed energy storage technology, and in particular to a method, system, equipment, medium and product for optimizing the configuration of distributed energy storage in a park power distribution network. Background Technology

[0002] With the rapid development of distributed energy, large-scale integration of distributed photovoltaic, wind power, adjustable loads, electric vehicle charging facilities, and critical loads has occurred in industrial parks, commercial parks, data center parks, and other scenarios. The park's distribution network has transformed from a traditional unidirectional power supply network into a complex system involving the coordinated operation of multiple entities including power sources, grids, loads, and energy storage. Against this backdrop, park distribution networks commonly face challenges such as large peak-to-valley load differences, significant fluctuations in renewable energy output, high requirements for power supply assurance for critical loads, localized line overloads, node voltage exceeding limits, and renewable energy curtailment. Distributed energy storage, with its advantages of rapid power regulation, bidirectional energy buffering, and flexible deployment, is widely used in scenarios such as peak shaving and valley filling, local renewable energy consumption, improved power supply reliability, voltage support, and line overload mitigation, becoming a key piece of equipment for the efficient operation of park distribution networks.

[0003] Currently, most methods for optimizing the configuration of distributed energy storage in industrial parks are based on structured numerical data such as load forecasting, renewable energy output forecasting, time-of-use pricing, distribution network topology parameters, and energy storage costs. These methods construct energy storage location and capacity optimization models with the goals of economy, safety, and absorption capacity. Under constraints such as energy storage power capacity, fixed SOC upper and lower limits, node voltage constraints, and line capacity constraints, mathematical programming or intelligent optimization algorithms are used to solve for energy storage access locations, rated power, rated capacity, and operating strategies. However, these methods generally use fixed SOC upper and lower limits or single-scenario safety margins, making it difficult to simultaneously address the demand for raising the lower limit of SOC to ensure the supply of important loads and the demand for lowering the upper limit of SOC to absorb renewable energy. This can easily lead to contradictions such as insufficient emergency power or insufficient absorption space, reducing the reliability of the industrial park's distribution network operation. Summary of the Invention

[0004] This invention provides a method, system, equipment, medium, and product for optimizing the configuration of distributed energy storage in industrial park distribution networks. It solves the technical problem that most methods for optimizing the configuration of distributed energy storage in industrial parks are based on structured numerical data such as load forecasting, new energy output forecasting, time-of-use pricing, topology parameters, and energy storage costs. These methods construct energy storage location and capacity optimization models and solve configuration schemes under fixed SOC constraints, voltage, and line capacity constraints. This often leads to contradictions such as insufficient emergency power supply or insufficient reserved absorption space, which reduces the reliability of industrial park distribution network operation.

[0005] The first aspect of this invention provides a method for optimizing the configuration of distributed energy storage in a park power distribution network, comprising:

[0006] With the goal of minimizing operating costs, energy storage constraints are set, and energy storage configuration data is used as decision variables to construct a distributed energy storage optimization configuration model;

[0007] Obtain the operation data and information of the park's power distribution network, and determine the corresponding node energy storage basic state of charge limit based on the preset power supply adaptation function and the operation data.

[0008] Based on the preset new energy consumption space function, the upper limit of the basic state of charge of the corresponding node energy storage is determined according to the operating data.

[0009] Based on the pre-trained language model, the running data, and the running information, the lower limit and upper limit of the state of charge of the energy storage base of each node are corrected to obtain multiple dynamic node state of charge constraints.

[0010] The distributed energy storage optimization configuration model is optimized and solved using the state of charge constraints of each dynamic node and the operating data to obtain the corresponding distributed energy storage optimization configuration scheme.

[0011] Optionally, the operational data includes multiple important load level coefficients, multiple important load power, multiple continuous supply guarantee times, multiple minimum states of charge, and multiple node energy storage rated capacities. The step of determining the corresponding node energy storage basic state of charge limit based on the operational data according to a preset supply guarantee adaptation function includes:

[0012] The running data is input into a preset supply guarantee adaptation function to obtain multiple supply guarantee adaptation coefficients;

[0013] Based on the preset supply guarantee demand allocation function, the demand is allocated to each of the supply guarantee adaptation coefficients to obtain multiple supply guarantee demand undertaking ratios.

[0014] Each of the aforementioned supply guarantee demand sharing ratios is multiplied by the corresponding important load level coefficient, important load power, and continuous supply guarantee time to obtain multiple first multiplication values;

[0015] The first multiplication values ​​corresponding to each of the lowest states of charge are summed to obtain multiple reserved basic charges.

[0016] Each of the reserved basic power quantities is compared with the corresponding node energy storage rated capacity to obtain multiple first ratios;

[0017] Each of the lowest states of charge is summed with its corresponding first ratio to obtain the lower limit of the basic state of charge of multiple node energy storage.

[0018] Optionally, the operational data further includes multiple predicted outputs of new energy sources, multiple absorbed power from new energy sources, and multiple maximum states of charge. The step of determining the upper limit of the corresponding node energy storage base state of charge based on the operational data according to a preset new energy absorption space function includes:

[0019] The predicted output of each new energy source is compared with the corresponding absorbed new energy power to obtain multiple first differences.

[0020] When the first difference is greater than or equal to the preset benchmark power curtailment, the first difference is determined as the predicted power curtailment.

[0021] When the first difference is less than the benchmark power curtailment, the benchmark power curtailment is determined as the predicted power curtailment.

[0022] Each of the predicted curtailed power is input into a preset new energy consumption space function to obtain multiple reserved charging spaces;

[0023] Each of the reserved rechargeable spaces is compared with the corresponding node energy storage rated capacity to obtain multiple second ratios;

[0024] The difference between each highest state of charge and the corresponding second ratio is processed to obtain the upper limit of the basic state of charge of multiple node energy storage.

[0025] Optionally, the step of correcting the lower limit and upper limit of the basic state of charge of each node's energy storage based on the pre-trained language model, the running data, and the running information to obtain multiple dynamic node state of charge constraints includes:

[0026] The running data and the running information are input into a pre-trained language model to obtain multiple upper limit corrections for the charge state and multiple upper limit corrections for the charge state.

[0027] Each of the above-mentioned charge state lower limit corrections and each of the above-mentioned charge state upper limit corrections are input into a preset boundary verification function to obtain multiple target charge state lower limit corrections and multiple target charge state upper limit corrections.

[0028] Each node energy storage basic state of charge limit is summed with the corresponding target state of charge limit correction amount to obtain multiple corrected state of charge limits.

[0029] The difference between the upper limit of the basic state of charge of each node energy storage and the corresponding target upper limit of the state of charge correction is processed to obtain multiple corrected upper limits of the state of charge.

[0030] When the modified charge state limit is less than or equal to the corresponding modified charge state upper limit, the modified charge state limit and the corresponding modified charge state upper limit are used to construct the corresponding dynamic node charge state constraint.

[0031] When the modified state of charge limit is greater than the corresponding modified state of charge upper limit, the target state of charge limit correction amount and the target state of charge upper limit correction amount corresponding to the modified state of charge limit are multiplied by the preset backoff coefficient to obtain the new target state of charge limit correction amount and the new target state of charge limit correction amount.

[0032] Jump to execute the step of summing the basic state of charge (SBC) limit of each node energy storage with the corresponding target SBC correction amount to obtain multiple corrected SBC limits.

[0033] Optionally, the step of optimizing the distributed energy storage configuration model using the state-of-charge constraints of each dynamic node and the operating data to obtain the corresponding distributed energy storage configuration scheme includes:

[0034] By combining the charge state constraints of each dynamic node with the distributed energy storage optimization configuration model, the corresponding target distributed energy storage optimization configuration model is obtained.

[0035] Based on a preset optimization algorithm, the target distributed energy storage optimization configuration model is optimized and solved according to the running data to obtain the corresponding distributed energy storage optimization configuration scheme.

[0036] Optionally, the energy storage constraints include energy storage installation constraints, energy storage charging and discharging power constraints, energy storage state of charge transition constraints, new energy consumption constraints, node voltage constraints, line capacity constraints, and power flow constraints.

[0037] The second aspect of this invention provides a distributed energy storage optimization configuration system for a park power distribution network, comprising:

[0038] The module is used to set energy storage constraints with the goal of minimizing operating costs, and to build a distributed energy storage optimization configuration model with energy storage configuration data as the decision variable.

[0039] The data acquisition module is used to acquire the operation data and information of the park's power distribution network, and based on the preset power supply adaptation function, determine the corresponding node energy storage basic state of charge limit according to the operation data.

[0040] The analysis module is used to determine the upper limit of the basic state of charge of the corresponding node energy storage based on the operating data, according to the preset new energy consumption space function.

[0041] The constraint module is used to correct the lower limit of the state of charge of each node's energy storage base and the upper limit of the state of charge of each node's energy storage base based on the pre-trained language large model, the running data, and the running information, so as to obtain multiple dynamic node state of charge constraints.

[0042] The optimization module is used to optimize and solve the distributed energy storage optimization configuration model by using the charge state constraints of each dynamic node and the operating data, so as to obtain the corresponding distributed energy storage optimization configuration scheme.

[0043] The third aspect of the present invention provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the above-described method for optimizing the configuration of distributed energy storage in a park power distribution network.

[0044] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements the above-described method for optimizing the configuration of distributed energy storage in a park power distribution network.

[0045] The fifth aspect of the present invention provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein when the program instructions are executed by a computer, the computer performs the above-described method for optimizing the configuration of distributed energy storage in a park power distribution network.

[0046] As can be seen from the above technical solutions, the present invention has the following advantages:

[0047] This invention aims to minimize operating costs by setting energy storage constraints and using energy storage configuration data as decision variables to construct a distributed energy storage optimization configuration model. It acquires operational data and information from the park's distribution network, determines the corresponding node energy storage state-of-charge (SOC) limits based on a pre-set supply guarantee adaptation function and the corresponding node SOC upper limits based on the operational data, and further refines these limits using a pre-trained large-scale linguistic model, operational data, and operational information. This results in multiple dynamic node SOC constraints. The distributed energy storage optimization configuration model is then optimized using these constraints and operational data to obtain the corresponding optimized distributed energy storage configuration scheme. This overcomes the technical problems of insufficient emergency power or inadequate consumption space reservation in traditional park-based distributed energy storage optimization configuration methods, which reduce the reliability of the park's distribution network. Compared with traditional distributed energy storage optimization configuration methods, this invention constructs a distributed energy storage optimization configuration model with the goal of minimizing operating costs. Through a supply guarantee adaptation function and a new energy consumption space function, it accurately calculates the lower limit of the node energy storage basic state of charge (BSC) to meet the supply guarantee needs of critical loads and the upper limit of the node energy storage BSC to meet the local consumption needs of new energy, respectively. This overcomes the limitation of traditional fixed BSC constraints, which cannot simultaneously address supply guarantee and consumption. Furthermore, using a pre-trained large-scale linguistic model, combined with operational data and information, the upper and lower limits of the node BSC are dynamically corrected, forming dynamic node BSC constraints that fit actual operating conditions. This effectively resolves the contradiction between insufficient emergency power and insufficient consumption space. Finally, the dynamic constraints are substituted into the distributed energy storage optimization configuration model to obtain an energy storage configuration and operation scheme that balances economy, safety, and reliability. This reduces the operating costs of the park's distribution network while improving the power supply guarantee capacity of critical loads, increasing the level of new energy consumption, and alleviating line overload and node voltage exceedance issues. This makes energy storage operation more adaptable to the complex scenario of source-grid-load-storage coordination in the park, significantly enhancing the reliability of the distribution network operation. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart illustrating the steps of a method for optimizing the configuration of distributed energy storage in a park power distribution network, as provided in Embodiment 1 of the present invention.

[0050] Figure 2This is a flowchart illustrating the steps of a method for optimizing the configuration of distributed energy storage in a park power distribution network, as provided in Embodiment 2 of the present invention.

[0051] Figure 3 This is the IEEE 33 campus topology diagram provided in Embodiment 2 of the present invention;

[0052] Figure 4 This is a schematic diagram of a typical daily park load and photovoltaic curve provided in Embodiment 2 of the present invention;

[0053] Figure 5 This is a schematic diagram of the distribution of critical load supply demand provided in Embodiment 2 of the present invention;

[0054] Figure 6 This is a schematic diagram of the node's supply adaptation coefficient to important loads provided in Embodiment 2 of the present invention;

[0055] Figure 7 This is a schematic diagram of the supply demand allocation matrix provided in Embodiment 2 of the present invention;

[0056] Figure 8 This is a basic SOC lower limit heat map provided in Embodiment 2 of the present invention;

[0057] Figure 9 This is a schematic diagram of the photovoltaic power output and potential curtailment power curves provided in Embodiment 2 of the present invention;

[0058] Figure 10 A thermal diagram showing the reserved space for new energy consumption provided in Embodiment 2 of the present invention;

[0059] Figure 11 This is a heat map of the basic SOC upper limit provided in Embodiment 2 of the present invention;

[0060] Figure 12 This is a schematic diagram of the SOC upper and lower limit correction amount provided in Embodiment 2 of the present invention;

[0061] Figure 13 This is a schematic diagram comparing the SOC corridor before and after the correction provided in Embodiment 2 of the present invention;

[0062] Figure 14 This is a schematic diagram of the rollback correction provided in Embodiment 2 of the present invention;

[0063] Figure 15 This is a cost comparison diagram provided for Embodiment 2 of the present invention;

[0064] Figure 16 This is a schematic diagram of energy storage charging and discharging and SOC trajectory provided in Embodiment 2 of the present invention;

[0065] Figure 17 This is a structural block diagram of a distributed energy storage optimization configuration system for a park power distribution network provided in Embodiment 3 of the present invention;

[0066] Figure 18 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention. Detailed Implementation

[0067] This invention provides a method, system, equipment, medium, and product for optimizing the configuration of distributed energy storage in a park power distribution network. It addresses the technical problem that many methods for optimizing the configuration of distributed energy storage in parks rely on structured numerical data such as load forecasting, renewable energy output forecasting, time-of-use pricing, topology parameters, and energy storage costs to construct energy storage location and capacity optimization models. These models often solve configuration schemes under fixed SOC constraints, voltage constraints, and line capacity constraints, which can easily lead to insufficient emergency power supply or inadequate capacity reservation, thus reducing the reliability of the park power distribution network.

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0069] Please see Figure 1 , Figure 1 The flowchart illustrates the steps of a method for optimizing the configuration of distributed energy storage in a park power distribution network, as provided in Embodiment 1 of the present invention.

[0070] This invention provides a method for optimizing the configuration of distributed energy storage in a park power distribution network, comprising:

[0071] Step 101: With the goal of minimizing operating costs, set energy storage constraints and use energy storage configuration data as decision variables to construct a distributed energy storage optimization configuration model.

[0072] Energy storage configuration data refers to the variables to be solved to determine the layout and operation status of energy storage, including energy storage access status, rated power, rated capacity and charging and discharging strategy.

[0073] In this embodiment of the invention, with the goal of minimizing operating costs, the energy storage installation constraints, energy storage charging and discharging power constraints, energy storage state of charge transition constraints, new energy consumption constraints, node voltage constraints, line capacity constraints, and power flow constraints are used as energy storage constraints. The distributed energy storage optimization configuration model is constructed with energy storage access status, rated power, rated capacity, and charging and discharging strategy as decision variables.

[0074] It should be noted that the specific distributed energy storage optimization configuration model is as follows:

[0075] ;

[0076] ;

[0077] ;

[0078] ;

[0079] ;

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] ;

[0085] ;

[0086] ;

[0087] ;

[0088] ;

[0089] ;

[0090] ;

[0091] in, The objective function value, Annualized cost of energy storage investment For energy storage operation and maintenance costs, For distribution network loss costs, The cost of penalizing the peak-valley difference. To incur the penalty cost for abandoning renewable energy, Penalty costs for critical load failures Penalty cost for exceeding voltage limits Penalty costs for overloading lines or equipment. To address the penalty costs for defaulting on dynamic SOC corridors, This indicates whether node i has energy storage installed; a value of 1 indicates installation, and a value of 0 indicates otherwise. The unit power investment cost, Rated energy storage power for node i The unit capacity investment cost The rated energy storage capacity of node i. This is the discount rate (usually ranging from 3% to 8%). This is a set of candidate energy storage access nodes. Where i represents the energy storage lifespan (typically 8–15 years), and i is the energy storage node index. Operating and maintenance costs per unit of charge / discharge capacity. The charging power for energy storage at node i during time period t. The discharge power of the energy stored at node i during time period t. t represents the time interval, and t is the time period index. To optimize the time period set, Let t be the electricity price for time period t. Let t be the power loss of the distribution network during time period t. For the collection of routes, Let the resistance of the l-th line be... Let be the current of the l-th line during time period t. The power purchased by the park from the upper-level power grid during time period t. Let i be the load power of node i in time period t. Let i be the actual renewable energy consumption power of node i in time period t. For a set of nodes, This is the peak-valley difference penalty coefficient. To optimize the maximum power purchase capacity of the park within the cycle, To optimize the minimum power purchase capacity of the park within the cycle, The penalty coefficient for curtailment of renewable energy. Let i be the actual power curtailment at node i during time period t. To contribute to the prediction of new energy sources for node i. Let i be the actual renewable energy consumption power at node i. This is the penalty factor for critical load failure. For the m-th important load level coefficient, Let m be the set of critical loads, and m be the index of the critical load. Let m be the power loss of the critical load during time period t. Let m be the power demand of the m-th critical load in time period t. This is the voltage over-limit penalty coefficient. Let be the voltage relaxation variable of node i over time period t. Let be the over-limit relaxation variable of node i under the voltage in time period t. For node voltage, This is the upper limit of the node voltage. This is the lower limit of the node voltage. This is the overload penalty factor for the line or equipment. Let l be the overload relaxation variable of the l-th line or equipment during time period t. The apparent power of the l-th line or equipment during time period t. For the allowable capacity of the l-th line or equipment, The penalty coefficient for defaulting on a SOC corridor. For node i, store energy below the dynamic SOC lower limit, which is a slack variable. For node i, store energy exceeding the dynamic SOC upper limit, a slack variable. Let SOC be the decision variable for energy storage at time t. This is the lower limit of dynamic SOC. This is the upper limit of dynamic SOC. The maximum energy storage capacity allowed to be connected to node i. Let i be the maximum energy storage capacity that node i is allowed to access. For charging state variables, For discharge state variables, For optimal energy storage and discharge efficiency, a value of 0.90 to 0.98 is preferred. Let SOC be the decision variable for energy storage at time t+1. Let i be the actual power curtailment at node i during time period t. Let represent the actual renewable energy power consumed by node i during time period t. To allow for a curtailment rate of renewable energy, a range of 0% to 5% is preferred. For the line The active power transmitted in time period t, For the line The reactive power transmitted during time period t, Let j be the active power load of node j in time period t. Let j be the reactive load of node j in time period t. For node j, the active power contribution of new energy sources during time period t. For node j, the reactive power output of new energy sources during time period t. For the line The resistance, For the line Reactance, For the line The current, Let J be the voltage at node j. Let be the voltage at node i, and j be the first index of node i.

[0092] Step 102: Obtain the operation data and information of the park's power distribution network, and determine the corresponding node energy storage basic charge limit based on the preset power supply adaptation function and the operation data.

[0093] Operational data refers to structured numerical information related to power generation, grid, load, and energy storage within the park. This includes the set of distribution network nodes, line connection relationships, and candidate energy storage access node locations; rated power of multiple important loads, multiple important load level coefficients, multiple supply guarantee levels and continuous supply guarantee times; predicted output of multiple new energy sources, power of multiple absorbed new energy sources, predicted output of distributed new energy sources for each time period, and upper limit of absorbable power; charging and discharging efficiency of energy storage devices, rated charging and discharging power; rated capacity of energy storage at multiple nodes, multiple minimum states of charge, and multiple maximum states of charge; allowable upper and lower limits of voltage at each node of the distribution network; capacity limits of each line and transformer; line impedance and reliable operation probability; and also includes time-of-use electricity price data, distribution network loss parameters, adjustable load regulation capacity range, and structured parameters that can be directly used for mathematical calculations, such as energy storage unit power investment cost, unit capacity investment cost, and annual operation and maintenance cost.

[0094] Operational information refers to textual or semi-structured task information that affects the operational boundary of energy storage. This includes, but is not limited to, unstructured or semi-structured task information that can directly affect the operational boundary of energy storage SOC, such as park production plans, lists of important loads to be supplied, equipment maintenance plans, dispatching and operation instructions, notices of new energy generation restrictions, operation and maintenance alarm information, and power supply guarantee requirements during special periods. Textual information involving raising the lower limit of SOC includes content such as raising the level of important loads, extending the duration of supply guarantee, maintenance of power supply paths, dispatching requirements to reserve emergency power, and increased local power supply risks. Textual information involving lowering the upper limit of SOC includes content such as limited new energy consumption, risk warnings of large-scale photovoltaic power generation, dispatching requirements to reserve charging space, and increased risks of line backfeeding or voltage exceeding the upper limit. All textual operational information is used for semantic parsing of the language big model and converted into calculable SOC correction quantities.

[0095] The minimum allowable state of charge (SOC) of a node energy storage unit is determined based on the supply requirements of critical loads, to prevent excessive discharge of energy storage from failing to guarantee power supply to critical loads.

[0096] In this embodiment of the invention, operational data and information of the park's power distribution network are acquired. The operational data is input into a preset supply guarantee adaptation function to obtain the supply guarantee adaptation coefficient of each energy storage node for important loads. Supply guarantee demand allocation is completed based on the supply guarantee adaptation coefficients using a preset supply guarantee demand allocation function, resulting in the supply guarantee demand undertaking ratio corresponding to each energy storage node. The supply guarantee demand undertaking ratio, important load level coefficient, important load power, and continuous supply guarantee time corresponding to each energy storage node are input into a preset reserved basic energy function to obtain multiple reserved basic energy quantities. The reserved basic energy quantity, node energy storage rated capacity, and minimum state of charge corresponding to each energy storage node are input into a preset basic state of charge limit function to obtain multiple node energy storage basic state of charge limits.

[0097] It should be noted that the specific supply guarantee adaptation function is as follows:

[0098] ;

[0099] in, Let be the supply adaptation coefficient of energy storage node i to critical load m. For the reliability of the power supply path from energy storage node i to critical load m, Let i be the electrical distance from energy storage node i to critical load m. To prevent small positive numbers with a denominator of zero, it is preferable to take... , Let l be the impedance of the l-th line. Let be the reliable operating probability of the l-th line, with a value ranging from 0 to 1. <rl≤1, This is the set of lines included in the power supply path from energy storage node i to critical load m. This is a route index.

[0100] It is worth mentioning that if reliable operational data is lacking, then Take 1.

[0101] The specific function for reserving basic power is as follows:

[0102] ;

[0103] in, Reserved basic power capacity for energy storage node i For energy storage nodes For the first The proportion of supply demand borne by each important load, For the m-th important load level coefficient, For the m-th critical load power, This represents the m-th consecutive supply guarantee period.

[0104] The specific bound function of the fundamental state of charge is as follows:

[0105] ;

[0106] in, This represents the state-of-charge limit of the energy storage base at the i-th node. For the i-th lowest state of charge, Let i be the rated energy storage capacity of the i-th node.

[0107] Step 103: Based on the preset new energy consumption space function, determine the upper limit of the basic state of charge of the corresponding node energy storage according to the operating data.

[0108] The upper limit of the base state of charge of nodal energy storage refers to the maximum allowable SOC value of energy storage determined based on the demand for new energy consumption, so as to avoid the inability to absorb surplus new energy after it is fully charged.

[0109] In this embodiment of the invention, the predicted output of each new energy source and the corresponding absorbed new energy power are input into a preset predicted curtailment power function to obtain multiple predicted curtailment powers. Each predicted curtailment power is input into a preset new energy absorption space function to obtain multiple reserved charging spaces. Each reserved charging space, along with the corresponding node energy storage rated capacity and maximum state of charge, is input into a preset baseline state of charge upper limit function to obtain multiple node energy storage baseline state of charge upper limits.

[0110] It should be noted that the function for predicting the power curtailment is as follows:

[0111] ;

[0112] in, Let i be the predicted power curtailment at time t. For the predicted output of the i-th new energy source at time t, Let t be the i-th source of renewable energy power absorbed at time t.

[0113] The specific space function for renewable energy consumption is as follows:

[0114] ;

[0115] in, Reserved charging space for the i-th time t. For optimal energy storage charging efficiency, a value of 0.90 to 0.98 is preferred. for The predicted power curtailment at time i, The time index is used for prediction, and H is the length of the prediction window for renewable energy consumption. In the day-ahead optimization, 4 to 24 hours is preferred, while in the intraday rolling optimization, 1 to 4 hours is preferred.

[0116] The upper bound function for the fundamental state of charge is as follows:

[0117] ;

[0118] in, This represents the upper limit of the state of charge of the energy storage base at the i-th node at time t. This represents the i-th highest state of charge at time t. Let be the rated energy storage capacity of the i-th node at time t.

[0119] Step 104: Based on the pre-trained language model, running data, and running information, correct the lower limit of the basic state of charge of each node's energy storage and the upper limit of the basic state of charge of each node's energy storage to obtain multiple dynamic node state of charge constraints.

[0120] In this embodiment of the invention, operational data and information are input into a pre-trained language large model to obtain multiple state-of-charge (SOC) limit corrections and multiple SOC upper limit corrections. Each SOC limit correction and each SOC upper limit correction is then input into a preset boundary verification function to obtain multiple target SOC limit corrections and multiple target SOC upper limit corrections. Each node's basic SOC limit is summed with its corresponding target SOC limit correction to obtain multiple corrected SOC limits. Each node's basic SOC upper limit is then subtracted from its corresponding target SOC upper limit correction to obtain multiple corrected SOC upper limits. When a corrected SOC limit is less than or equal to its corresponding corrected SOC upper limit, the corrected SOC limit and the corresponding corrected SOC upper limit are used to construct the corresponding dynamic node SOC constraint. When the corrected state of charge lower limit is greater than the corresponding corrected state of charge upper limit, the corrected state of charge lower limit and the corrected state of charge upper limit are reduced by a preset backoff coefficient and then the corrected state of charge lower limit and the corrected state of charge upper limit are recalculated until the corrected state of charge lower limit is less than or equal to the corresponding corrected state of charge upper limit.

[0121] Step 105: Use the state-of-charge constraints and operating data of each dynamic node to optimize and solve the distributed energy storage configuration model, and obtain the corresponding distributed energy storage configuration scheme.

[0122] In this embodiment of the invention, the state of charge constraints of each dynamic node and the operating data are substituted into the constructed distributed energy storage optimization configuration model. Combined with the operating data, the distributed energy storage optimization configuration model is solved by a preset optimization algorithm to obtain a distributed energy storage optimization configuration scheme that includes the optimal access location of energy storage, rated power, rated capacity, charging and discharging strategies for each time period and the SOC operating trajectory.

[0123] In this embodiment of the invention, by setting energy storage constraints with the goal of minimizing operating costs, and using energy storage configuration data as decision variables, a distributed energy storage optimization configuration model is constructed. The model acquires the operating data and information of the park's distribution network. Based on a preset supply guarantee adaptation function, the corresponding node energy storage basic state-of-charge (BSOC) limits are determined according to the operating data. Based on a preset new energy consumption space function, the corresponding node energy storage BSOC upper limit is determined according to the operating data. The BSOC limits and upper limits of each node energy storage are corrected based on a pre-trained large-scale linguistic model, operating data, and operating information, resulting in multiple dynamic node BSOC constraints. These dynamic node BSOC constraints and operating data are then used to optimize and solve the distributed energy storage optimization configuration model, yielding the corresponding distributed energy storage optimization configuration scheme. This overcomes the technical problem of insufficient emergency power or inadequate consumption space reservation in traditional park distributed energy storage optimization configuration methods, which reduces the reliability of the park's distribution network operation. Compared with traditional distributed energy storage optimization configuration methods, this invention constructs a distributed energy storage optimization configuration model with the goal of minimizing operating costs. Through a supply guarantee adaptation function and a new energy consumption space function, it accurately calculates the lower limit of the node energy storage basic state of charge (BSC) to meet the supply guarantee needs of critical loads and the upper limit of the node energy storage BSC to meet the local consumption needs of new energy, respectively. This overcomes the limitation of traditional fixed BSC constraints, which cannot simultaneously address supply guarantee and consumption. Furthermore, using a pre-trained large-scale linguistic model, combined with operational data and information, the upper and lower limits of the node BSC are dynamically corrected, forming dynamic node BSC constraints that fit actual operating conditions. This effectively resolves the contradiction between insufficient emergency power and insufficient consumption space. Finally, the dynamic constraints are substituted into the distributed energy storage optimization configuration model to obtain an energy storage configuration and operation scheme that balances economy, safety, and reliability. This reduces the operating costs of the park's distribution network while improving the power supply guarantee capacity of critical loads, increasing the level of new energy consumption, and alleviating line overload and node voltage exceedance issues. This makes energy storage operation more adaptable to the complex scenario of source-grid-load-storage coordination in the park, significantly enhancing the reliability of the distribution network operation.

[0124] Please see Figure 2 , Figure 2 The flowchart illustrates the steps of a method for optimizing the configuration of distributed energy storage in a park power distribution network, as provided in Embodiment 2 of the present invention.

[0125] This invention provides a method for optimizing the configuration of distributed energy storage in a park power distribution network, comprising:

[0126] Step 201: With the goal of minimizing operating costs, set energy storage constraints and use energy storage configuration data as decision variables to construct a distributed energy storage optimization configuration model.

[0127] In this embodiment of the invention, with the minimum operating cost as the optimization objective, an objective function is constructed by comprehensively considering the annualized cost of energy storage investment, operation and maintenance cost, distribution network loss cost, peak-valley difference penalty cost, renewable energy curtailment penalty cost, critical load failure penalty cost, voltage over-limit penalty cost, and line overload penalty cost. At the same time, energy storage constraints are set, and energy storage configuration data is used as decision variables to construct a distributed energy storage optimization configuration model.

[0128] It should be noted that energy storage constraints include energy storage installation constraints, energy storage charging and discharging power constraints, energy storage state of charge transition constraints, new energy consumption constraints, node voltage constraints, line capacity constraints, and power flow constraints.

[0129] The specific constraints on energy storage installation are as follows:

[0130] ;

[0131] The specific constraints on energy storage charging and discharging power are as follows:

[0132] ;

[0133] The specific constraints for energy storage charge state transition are:

[0134] ;

[0135] The specific constraints on the absorption of new energy sources are as follows:

[0136] ;

[0137] The node voltage constraints are as follows:

[0138] ;

[0139] The specific line capacity constraints are as follows:

[0140] ;

[0141] The specific constraints of current flow are:

[0142] ;

[0143] Step 202: Obtain the operation data and information of the park's power distribution network, and determine the corresponding node energy storage basic charge state limit based on the preset power supply adaptation function and the operation data.

[0144] Furthermore, the operational data includes multiple critical load level coefficients, multiple critical load power, multiple continuous supply guarantee times, multiple minimum states of charge, and multiple node energy storage rated capacities. Step 202 includes the following sub-steps:

[0145] S11. Input the running data into the preset supply guarantee adaptation function to obtain multiple supply guarantee adaptation coefficients.

[0146] The power supply adaptation coefficient refers to the dimensionless value calculated by the power supply adaptation function, which is used to characterize the strength of the power supply support of energy storage nodes for important loads.

[0147] In this embodiment of the invention, the operating data is input into a preset supply adaptation function to obtain the supply adaptation coefficients of each energy storage node corresponding to different important loads.

[0148] S12. Based on the preset supply guarantee demand allocation function, the demand is allocated to each supply guarantee adaptation coefficient to obtain multiple supply guarantee demand undertaking ratios.

[0149] The supply guarantee demand sharing ratio refers to the share of responsibility that each energy storage node should bear for guaranteeing the supply of important loads, which is used to allocate the amount of power to be guaranteed.

[0150] In this embodiment of the invention, the supply matching coefficients of each energy storage node corresponding to different important loads are input into a preset supply demand allocation function to obtain multiple supply demand sharing ratios.

[0151] It should be noted that the specific function for allocating supply demand is as follows:

[0152] ;

[0153] in, is the supply adaptation coefficient of energy storage node j to important load m.

[0154] S13. Multiply each supply guarantee demand share by the corresponding important load level coefficient, important load power and continuous supply guarantee time to obtain multiple first multiplication values.

[0155] The critical load level coefficient refers to a quantitative coefficient set according to the priority of power supply protection for critical loads. The higher the level, the larger the value. For example, the critical load level is 1-1.2, the critical load level is 0.7-1, and the critical load level is 0.4-0.7.

[0156] Important load power refers to the rated active power of critical loads within the park.

[0157] Continuous power supply time refers to the duration for which critical loads require energy storage to continuously supply power.

[0158] In this embodiment of the invention, the multiplication values ​​between each supply guarantee demand sharing ratio and the corresponding important load level coefficient, important load power and continuous supply guarantee time are calculated to obtain multiple first multiplication values ​​used to characterize the size of the supply guarantee demand shared by each node.

[0159] S14. Sum the first multiplication values ​​corresponding to each lowest state of charge to obtain multiple reserved basic charges.

[0160] Reserved basic power capacity refers to the minimum amount of power that an energy storage node must reserve in advance to ensure power supply to critical loads.

[0161] In this embodiment of the invention, the sum of the first multipliers corresponding to each lowest state of charge is calculated to obtain multiple reserved basic charge values.

[0162] S15. Ratio the reserved basic power capacity with the corresponding node energy storage rated capacity to obtain multiple first ratios.

[0163] The rated capacity of a node energy storage refers to the nominal rated capacity of the energy storage device configured in the energy storage node, which is a fixed design parameter.

[0164] In this embodiment of the invention, the ratio between each reserved basic power supply and the corresponding node energy storage rated capacity is calculated to obtain multiple first ratios used to characterize the proportion of power supply guaranteed.

[0165] S16. Sum the minimum state of charge of each node with the corresponding first ratio to obtain the lower limit of the basic state of charge of the energy storage at multiple nodes.

[0166] Minimum state of charge (SOC) refers to the lowest SOC baseline value that an energy storage device is allowed to operate at, used to prevent damage from over-discharge of the battery.

[0167] In this embodiment of the invention, the sum of each minimum state of charge and the corresponding first ratio is calculated to obtain the minimum state of charge of multiple node energy storage.

[0168] Step 203: Based on the preset new energy consumption space function, determine the upper limit of the basic state of charge of the corresponding node energy storage according to the operating data.

[0169] Furthermore, the operational data also includes multiple predicted outputs from new energy sources, multiple absorbed power from new energy sources, and multiple maximum states of charge. Step 203 includes the following sub-steps:

[0170] S21. The predicted output of each new energy source is processed by the difference between the corresponding absorbed new energy power to obtain multiple first differences.

[0171] New energy forecast output refers to the predicted power generation of distributed energy sources such as photovoltaics and wind power in the park at different times.

[0172] Absorbing new energy power refers to the maximum power limit of new energy that the distribution network and local loads can safely absorb.

[0173] In this embodiment of the invention, the difference between the predicted output of each new energy source and the corresponding absorbed new energy power is calculated to obtain a number of first differences that reflect the size of the new energy absorption gap.

[0174] S22. When the first difference is greater than or equal to the preset benchmark power curtailment, the first difference is determined as the predicted power curtailment.

[0175] The baseline curtailment power refers to a pre-set threshold for judging curtailment power, used to distinguish whether effective curtailment has occurred, and is usually set to 0.

[0176] Predicted curtailment power refers to the surplus power generated by new energy sources that exceeds the capacity of the power grid and load to absorb it.

[0177] In this embodiment of the invention, when the first difference is greater than or equal to the preset benchmark power curtailment, the first difference is determined as the predicted power curtailment of the node corresponding to the current time period.

[0178] S23. When the first difference is less than the baseline power curtailment, the baseline power curtailment is determined as the predicted power curtailment.

[0179] In this embodiment of the invention, when the first difference is less than the reference curtailment power, the reference curtailment power is determined as the predicted curtailment power of the node corresponding to the current time period.

[0180] S24. Input each predicted power curtailment into the preset new energy consumption space function to obtain multiple reserved charging spaces.

[0181] Reserved charging space refers to the spare charging capacity that energy storage needs to maintain in advance to avoid power waste.

[0182] In this embodiment of the invention, each predicted power curtailment is input into a preset new energy consumption space function, and the calculation is completed by combining the energy storage charging efficiency and the preset prediction window duration to obtain multiple reserved charging spaces that energy storage needs to reserve in advance for new energy consumption.

[0183] It should be noted that the forecast window refers to the length of the future period used to predict the demand for renewable energy consumption in advance.

[0184] S25. Ratio each reserved rechargeable space with the corresponding node energy storage rated capacity to obtain multiple second ratios.

[0185] In this embodiment of the invention, the ratio between each reserved rechargeable space and the corresponding node energy storage rated capacity is calculated to obtain a number of second ratios used to characterize the proportion of reserved capacity to be consumed.

[0186] S26. The difference between each highest state of charge and the corresponding second ratio is processed to obtain the upper limit of the basic state of charge of multiple node energy storage.

[0187] Maximum state of charge (SOC) refers to the highest SOC reference value that an energy storage device is allowed to operate at, used to prevent battery damage from overcharging.

[0188] In this embodiment of the invention, the difference between each highest state of charge and the corresponding second ratio is calculated to obtain the upper limit of the basic state of charge of multiple node energy storage.

[0189] Step 204: Based on the pre-trained language big data model, running data, and running information, correct the lower limit of the basic state of charge of each node's energy storage and the upper limit of the basic state of charge of each node's energy storage to obtain multiple dynamic node state of charge constraints.

[0190] Further, step 204 includes the following sub-steps:

[0191] S31. Input the running data and running information into the pre-trained language large model to obtain multiple upper limit corrections for charge states and multiple upper limit corrections for charge states.

[0192] The state-of-charge lower limit correction refers to the correction value used to raise the lower limit of the high-temperature baseline, adapting to scenarios such as power supply upgrades and increased power supply risks.

[0193] The upper limit correction of the state of charge refers to the correction value used to reduce the upper limit of the base SOC, which is adapted to scenarios such as limited consumption of new energy and large-scale photovoltaic power generation.

[0194] In this embodiment of the invention, the running data and running information are input into a pre-trained language model for semantic parsing and feature extraction, and the corresponding state of charge lower limit correction amount and state of charge upper limit correction amount are output for each energy storage node.

[0195] It should be noted that the large language model can be a lightweight encoder-decoder architecture, constructed using a Transformer-based simplified encoding layer, semantic feature extraction layer, constraint classification layer, and numerical output layer. The encoding layer is used to embed features from textual operational information and numerical operational data. The semantic feature extraction layer is used to identify key information such as supply guarantee level, maintenance period, absorption risk, and scheduling requirements. The constraint classification layer is used to distinguish between the state-of-charge (SOC) limit correction type and the SOC upper limit correction type. The numerical output layer is used to output quantized correction values. This structure completes the mapping and transformation from textual information to constraint correction values, resulting in multiple SOC limit correction values ​​and multiple SOC upper limit correction values. The large language model can also use Qwen2.5 / Qwen30.5B, 1.5B, 3B, 4B, 7B, etc.

[0196] S32. Input the lower limit correction amount and the upper limit correction amount of each charge state into the preset boundary verification function to obtain multiple target charge state lower limit correction amounts and multiple target charge state upper limit correction amounts.

[0197] The target state of charge limit correction refers to the state of charge limit correction after boundary verification.

[0198] The target upper limit correction amount refers to the upper limit correction amount of the upper limit of the charge state after boundary verification.

[0199] In this embodiment of the invention, each state of charge lower limit correction amount and each state of charge upper limit correction amount are input into a preset boundary verification function. The boundary verification function performs upper and lower limit verification and reasonableness constraints on the correction amount. The correction amount that exceeds the preset allowable range is truncated to obtain multiple target state of charge lower limit correction amounts and multiple target state of charge upper limit correction amounts.

[0200] It should be noted that the boundary check function is as follows:

[0201] ;

[0202] in, For the limit correction of the i-th target charge state, This is the upper limit correction for the i-th target state of charge. This is the limit correction amount for the i-th state of charge. This is the upper limit correction for the i-th state of charge.

[0203] S33. Sum the basic state of charge (SUC) limit of each node energy storage with the corresponding target SUC correction amount to obtain multiple corrected SUC limits.

[0204] In this embodiment of the invention, the sum between the basic state of charge limit of each node energy storage and the corresponding target state of charge limit correction amount is calculated to obtain multiple corrected state of charge limits.

[0205] S34. The difference between the upper limit of the basic state of charge of each node energy storage and the corresponding target upper limit of the state of charge correction is processed to obtain multiple corrected upper limits of the state of charge.

[0206] In this embodiment of the invention, the difference between the upper limit of the basic state of charge of each node's energy storage and the corresponding correction amount of the target upper limit of the state of charge is calculated to obtain multiple corrected upper limits of the state of charge.

[0207] S35. When the modified charge state lower limit is less than or equal to the corresponding modified charge state upper limit, the corresponding dynamic node charge state constraint is constructed using the modified charge state lower limit and the corresponding modified charge state upper limit.

[0208] In this embodiment of the invention, when the modified charge state limit is less than or equal to the corresponding modified charge state upper limit, it is determined that the modified charge state limit and the corresponding modified charge state upper limit meet the operating logic requirements, and the corresponding dynamic node charge state constraint is constructed using the modified charge state limit and the corresponding modified charge state upper limit.

[0209] S36. When the corrected state of charge limit is greater than the corresponding corrected state of charge upper limit, the target state of charge limit correction amount and the target state of charge upper limit correction amount corresponding to the corrected state of charge limit are multiplied by the preset backoff coefficient to obtain the new target state of charge limit correction amount and the new target state of charge limit correction amount.

[0210] The rollback factor is a pre-set proportional coefficient less than 1, used to adjust the correction amount that exceeds the limit, thereby reducing the correction range.

[0211] In this embodiment of the invention, when the corrected state of charge limit is greater than the corresponding corrected state of charge upper limit, it is determined that the corrected state of charge limit and the corresponding corrected state of charge upper limit do not meet the operating logic requirements. The target state of charge limit correction amount and the target state of charge upper limit correction amount corresponding to the corrected state of charge limit are multiplied by the preset backoff coefficient to obtain the new target state of charge limit correction amount and the new target state of charge limit correction amount.

[0212] S37. Jump to execute the step of summing the basic state of charge (SBC) limit of each node's energy storage with the corresponding target SBC limit correction amount to obtain multiple corrected SBC limits.

[0213] In this embodiment of the invention, the step of summing the basic state of charge (SBC) limit of each node's energy storage with the corresponding target SBC correction amount is executed to obtain multiple corrected SBC limits. The corrected SBC limits and corrected SBC upper limits are then iteratively calculated using the scaled target SBC correction amounts.

[0214] It is worth mentioning that the methods S36-S37 are used to perform backtracking corrections on the target state of charge upper limit correction and the target state of charge upper limit correction. (See [reference needed]). Figure 14 As shown, the gap was positive before the rollback correction, indicating that the lower limit was higher than the upper limit during certain periods. After the rollback correction, the gap fell back to a non-positive value, indicating that the lower and upper limits of the corrected charge state were restored to the executable range. The bars in the figure show the number of rollbacks, revealing the periods when conflicts occurred in concentrated areas.

[0215] Step 205: Combine the state-of-charge constraints of each dynamic node with the distributed energy storage optimization configuration model to obtain the corresponding target distributed energy storage optimization configuration model.

[0216] In this embodiment of the invention, the state-of-charge constraints of each dynamic node are combined with the distributed energy storage optimization configuration model to obtain a target distributed energy storage optimization configuration model that adapts to actual operating conditions.

[0217] Step 206: Based on the preset optimization algorithm, optimize and solve the target distributed energy storage optimization configuration model according to the running data to obtain the corresponding distributed energy storage optimization configuration scheme.

[0218] Optimization algorithms refer to numerical algorithms used to solve planning models and search for optimal configuration results, possessing efficient optimization capabilities. Examples include, but are not limited to, mixed-integer linear programming methods, mixed-integer second-order cone programming methods, particle swarm optimization, genetic algorithms, or hierarchical optimization algorithms.

[0219] In this embodiment of the invention, according to a preset optimization algorithm, the target distributed energy storage optimization configuration model is iteratively optimized and solved using running data. Under the premise of satisfying all constraints, the optimal solution is obtained, and a distributed energy storage optimization configuration scheme adapted to the current operating conditions is obtained.

[0220] It is worth mentioning that the IEEE 33-node public topology is used as the underlying network, and energy storage nodes, critical load nodes, and photovoltaic nodes are specified in this network, such as... Figure 3 As shown in the diagram. The lines in the diagram represent branches, the node numbers correspond to the IEEE 33 node numbers, the ring-highlighted nodes represent candidate energy storage locations, the square markers represent important loads, and the triangular markers represent photovoltaic access points.

[0221] See Figure 4 As shown, the curves of total load and total photovoltaic power output of the park over 24 hours are presented, showing the temporal contradiction between the peak photovoltaic power generation at noon and the peak load in the evening. This intuitively reflects the reverse constraint of supply guarantee and new energy consumption on SOC, and provides data support for the temporal construction of the upper and lower limits of basic SOC.

[0222] See Figure 5 As shown, the power supply requirements for four types of important loads are statistically analyzed, and it is determined that the power supply requirement for data center loads is the largest. The required energy storage lock-in power for each important load is calculated according to load level, power, and supply time, providing the input for supply demand calculation of the basic SOC lower limit.

[0223] See Figure 6As shown, a heat map is used to display the electrical distance and path reliability of each candidate energy storage node to the important load. The darker the color, the stronger the adaptability, reflecting the principle of "local energy storage giving priority to supporting local important loads" and providing a weighting basis for the allocation of supply demand.

[0224] See Figure 7 As shown, based on the supply matching coefficient, the supply demand of each important load is proportionally allocated to each candidate energy storage node, clarifying the supply responsibility to be borne by each node, and completing the detailed mapping from "network-wide supply guarantee" to "node-level supply guarantee".

[0225] See Figure 8 As shown, the power supply locked after being allocated to each node is converted into the basic SOC lower limit for each time period of 24 hours. The darker the color, the higher the minimum power supply needs to be reserved for that time period to meet the emergency power supply requirements of important loads, thus forming the dynamic SOC boundary on the power supply side.

[0226] See Figure 9 As shown, by comparing the predicted output of photovoltaic power with the local absorption capacity in different time periods, the peak value of potential curtailed power during the noon period is extracted, the gap in the absorption of new energy is identified, and curtailed power data is provided for calculating the reserved space for energy storage charging.

[0227] See Figure 10 As shown, Figure 9 The potential risk of power curtailment becomes the charging space that each node needs to reserve. Based on the potential power curtailment in the next 4 hours, the charging space that each node needs to reserve for each time period is calculated. The darker the color, the more power needs to be released in advance, providing the input of consumption demand for the calculation of the basic SOC upper limit.

[0228] See Figure 11 As shown, the calculated peak of the node's reserved charging space is approximately 2.800 MWh, occurring around node 25 at time 11:00. This indicates that the pressure on renewable energy consumption is strongest during the middle of the day.

[0229] See Figure 12 As shown, after the language big data model analyzes the text operation information, it can be seen that the upper limit correction of node 18 is the strongest, with a peak value of 0.1242. This indicates that the photovoltaic-related nodes in the R&D area need to reserve space the most during the noon period.

[0230] See Figure 13 As shown, taking node 14 as an example, the shapes of the corridor (i.e., dynamic node charge state constraints) before and after the correction are compared. The lower limit curve is raised during the supply guarantee event, while the upper limit curve is lowered during the peak photovoltaic power generation period, indicating that the text task has been transformed into an executable running boundary.

[0231] See Figure 15 and Figure 16As shown, Case A is without energy storage baseline; Case B is the result obtained by optimization only under the minimum and maximum state of charge (SOC) of node energy storage; Case C is the result obtained by sampling the optimization method of this invention. Case B-under-AI is the result obtained by optimization without backoff correction in this application. After optimization configuration using the method of this invention, the charging and discharging power and SOC trajectory curves of distributed energy storage in the park on a typical day are shown. The horizontal axis represents the 24-hour operation period of a typical day. The upper part represents the changes in charging and discharging power of energy storage in each period, with positive values ​​for charging and negative values ​​for discharging. The lower part simultaneously gives the real-time SOC of energy storage, the final dynamic SOC lower limit and SOC upper limit after language large model correction and boundary verification and backoff correction. As can be seen from the curve changes, during the nighttime off-peak hours when there is no photovoltaic output, the energy storage charges with a small power, and the SOC rises steadily and remains within a reasonable range. This utilizes low-cost electricity to complete basic power replenishment while reserving sufficient space for daytime photovoltaic consumption and supply to important loads. During the midday peak photovoltaic power generation period, due to the demand for new energy consumption and the correction effect of the big data model and operational information, the upper limit of SOC is moderately reduced. The energy storage actively increases its charging power to absorb excess photovoltaic output, and the real-time SOC operates close to the corrected upper limit, effectively reducing the curtailment rate. During the evening peak electricity consumption period in the park and the supply period for important loads, the big data model raises the lower limit of SOC based on operational information. The energy storage discharges with a large power to support the peak load, and the real-time SOC is always maintained above the corrected lower limit, ensuring the reliability of power supply to important loads and emergency power reserves. After the load decreases at night, the energy storage completes the SOC adjustment through small charging and discharging, returning to the safe operating range. Throughout the process, the real-time SOC of the energy storage remained strictly within the dynamic SOC corridor after the language large model correction, without any exceeding of the limit. The charging and discharging behavior satisfied the power constraints and SOC state transition constraints, which directly verified the proposed method for optimizing the configuration of distributed energy storage in the park based on the language large model-corrected SOC corridor. This method can transform textual operation information into executable operation constraints, enabling energy storage operation to simultaneously meet the requirements of peak shaving and valley filling, new energy consumption, supply guarantee for important loads, and safe operation of the distribution network. The configuration scheme and operation strategy obtained by solving the problem have engineering feasibility and constraint compliance, and can stably adapt to the complex and ever-changing actual operation scenarios in the park.

[0232] In this embodiment of the invention, by setting energy storage constraints with the goal of minimizing operating costs, and using energy storage configuration data as decision variables, a distributed energy storage optimization configuration model is constructed. The model acquires the operating data and information of the park's distribution network. Based on a preset supply guarantee adaptation function, the corresponding node energy storage basic state-of-charge (BSOC) limits are determined according to the operating data. Based on a preset new energy consumption space function, the corresponding node energy storage BSOC upper limit is determined according to the operating data. The BSOC limits and upper limits of each node energy storage are corrected based on a pre-trained large-scale linguistic model, operating data, and operating information, resulting in multiple dynamic node BSOC constraints. These dynamic node BSOC constraints and operating data are then used to optimize and solve the distributed energy storage optimization configuration model, yielding the corresponding distributed energy storage optimization configuration scheme. This overcomes the technical problem of insufficient emergency power or inadequate consumption space reservation in traditional park distributed energy storage optimization configuration methods, which reduces the reliability of the park's distribution network operation. Compared with traditional distributed energy storage optimization configuration methods, this invention constructs a distributed energy storage optimization configuration model with the goal of minimizing operating costs. Through a supply guarantee adaptation function and a new energy consumption space function, it accurately calculates the lower limit of the node energy storage basic state of charge (BSC) to meet the supply guarantee needs of critical loads and the upper limit of the node energy storage BSC to meet the local consumption needs of new energy, respectively. This overcomes the limitation of traditional fixed BSC constraints, which cannot simultaneously address supply guarantee and consumption. Furthermore, using a pre-trained large-scale linguistic model, combined with operational data and information, the upper and lower limits of the node BSC are dynamically corrected, forming dynamic node BSC constraints that fit actual operating conditions. This effectively resolves the contradiction between insufficient emergency power and insufficient consumption space. Finally, the dynamic constraints are substituted into the distributed energy storage optimization configuration model to obtain an energy storage configuration and operation scheme that balances economy, safety, and reliability. This reduces the operating costs of the park's distribution network while improving the power supply guarantee capacity of critical loads, increasing the level of new energy consumption, and alleviating line overload and node voltage exceedance issues. This makes energy storage operation more adaptable to the complex scenario of source-grid-load-storage coordination in the park, significantly enhancing the reliability of the distribution network operation.

[0233] Please see Figure 17 , Figure 17 This is a structural block diagram of a distributed energy storage optimization configuration system for a park power distribution network provided in Embodiment 3 of the present invention.

[0234] This invention provides a distributed energy storage optimization configuration system for a park power distribution network, comprising:

[0235] Module 301 is used to construct a distributed energy storage optimization configuration model with the goal of minimizing operating costs, setting energy storage constraints, and using energy storage configuration data as decision variables.

[0236] The data acquisition module 302 is used to acquire the operation data and information of the park's power distribution network, and based on the preset power supply adaptation function, determine the corresponding node energy storage basic charge state limit according to the operation data.

[0237] Analysis module 303 is used to determine the upper limit of the basic state of charge of the corresponding node energy storage based on the preset new energy consumption space function and the operating data.

[0238] The constraint module 304 is used to correct the lower limit of the basic charge state of each node and the upper limit of the basic charge state of each node based on the pre-trained language large model, running data and running information, so as to obtain multiple dynamic node charge state constraints.

[0239] The optimization module 305 is used to optimize and solve the distributed energy storage optimization configuration model by using the charge state constraints and operating data of each dynamic node, so as to obtain the corresponding distributed energy storage optimization configuration scheme.

[0240] Furthermore, the operational data includes multiple critical load level coefficients, multiple critical load power, multiple continuous supply guarantee times, multiple minimum states of charge, and multiple node energy storage rated capacities. The data acquisition module 302 includes:

[0241] The supply guarantee adaptation submodule is used to input the running data into the preset supply guarantee adaptation function to obtain multiple supply guarantee adaptation coefficients;

[0242] Based on the preset supply demand allocation function, demand is allocated to each supply matching coefficient to obtain multiple supply demand undertaking ratios.

[0243] The allocation submodule is used to multiply the proportion of each supply guarantee demand with the corresponding important load level coefficient, important load power and continuous supply guarantee time to obtain multiple first multiplication values;

[0244] The first multiplier corresponding to each lowest state of charge is summed to obtain multiple reserved base charges;

[0245] Each reserved basic power capacity is compared with the corresponding node energy storage rated capacity to obtain multiple first ratios;

[0246] Each minimum state of charge is summed with its corresponding first ratio to obtain the minimum state of charge of multiple node energy storage systems.

[0247] Furthermore, the operational data also includes multiple predicted outputs from new energy sources, multiple absorbed power from new energy sources, and multiple maximum states of charge. The analysis module 303 includes:

[0248] The predicted curtailment power submodule is used to perform difference processing between the predicted output of each new energy source and the corresponding absorbed new energy power to obtain multiple first differences.

[0249] When the first difference is greater than or equal to the preset baseline power curtailment, the first difference is determined as the predicted power curtailment.

[0250] When the first difference is less than the baseline curtailment power, the baseline curtailment power is determined as the predicted curtailment power.

[0251] The analysis submodule is used to input each predicted power curtailment into a preset new energy consumption space function to obtain multiple reserved charging spaces;

[0252] Each reserved rechargeable space is compared with the corresponding node energy storage rated capacity to obtain multiple second ratios;

[0253] The difference between each highest state of charge and the corresponding second ratio is processed to obtain the upper limit of the basic state of charge of multiple node energy storage.

[0254] Furthermore, constraint module 304 includes:

[0255] The parsing submodule is used to input the running data and running information into the pre-trained language model to obtain multiple charge state lower limit corrections and multiple charge state upper limit corrections.

[0256] Each charge state lower limit correction and each charge state upper limit correction are input into a preset boundary verification function to obtain multiple target charge state lower limit corrections and multiple target charge state upper limit corrections.

[0257] The correction submodule is used to sum the basic state of charge (SBC) limit of each node's energy storage with the corresponding target SBC correction amount to obtain multiple corrected SBC limits.

[0258] The difference between the upper limit of the basic state of charge of each node's energy storage and the corresponding correction amount of the target upper limit of the state of charge is processed to obtain multiple corrected upper limits of the state of charge.

[0259] When the modified charge state lower limit is less than or equal to the corresponding modified charge state upper limit, the corresponding dynamic node charge state constraint is constructed using the modified charge state lower limit and the corresponding modified charge state upper limit.

[0260] When the corrected state of charge limit is greater than the corresponding corrected state of charge upper limit, the target state of charge limit correction amount and the target state of charge upper limit correction amount corresponding to the corrected state of charge limit are multiplied by the preset backoff coefficient to obtain the new target state of charge limit correction amount and the new target state of charge limit correction amount.

[0261] The process involves jumping to execute steps that sum the basic state of charge (SBC) limit of each node's energy storage with the corresponding target SBC correction amount to obtain multiple corrected SBC limits.

[0262] Furthermore, module 305 is optimized, including:

[0263] The coupling submodule is used to combine the charge state constraints of each dynamic node with the distributed energy storage optimization configuration model to obtain the corresponding target distributed energy storage optimization configuration model.

[0264] The optimization submodule is used to optimize and solve the target distributed energy storage optimization configuration model based on the preset optimization algorithm and the running data, so as to obtain the corresponding distributed energy storage optimization configuration scheme.

[0265] Furthermore, energy storage constraints include energy storage installation constraints, energy storage charging and discharging power constraints, energy storage state of charge transition constraints, new energy consumption constraints, node voltage constraints, line capacity constraints, and power flow constraints.

[0266] Please see Figure 18 , Figure 18 This is a structural block diagram of an electronic device provided in Embodiment 4 of the present invention.

[0267] An electronic device according to an embodiment of the present invention includes: a memory 401 and a processor 402. The memory 401 stores a computer program. When the computer program is executed by the processor 402, the processor 402 executes the distributed energy storage optimization configuration method for the park distribution network as described in any of the above embodiments.

[0268] Memory 401 may be an electronic memory such as flash memory, EEPROM (Electrically Erasable Programmable Read-Only Memory), EPROM, hard disk, or ROM. Memory 401 has storage space 403 for program code 413 for performing any of the method steps described above. For example, storage space 403 for program code may include individual program codes 413 for implementing the various steps in the methods described above. This program code may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When run by a computing processing device, this code causes the computing processing device to perform the various steps in the methods described above. This program code may be read from or written to one or more computer program products. These computer program products include program code carriers such as hard disks, CDs, memory cards, or floppy disks. The program code may be compressed, for example, in a suitable form. When this code is run by a computing device, it causes the device to execute the various steps in the distributed energy storage optimization configuration method for the campus distribution network described above.

[0269] Embodiment 5 of the present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the method for optimizing the configuration of distributed energy storage in a park power distribution network as described in any of the above embodiments.

[0270] Embodiment 6 of the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer executes the distributed energy storage optimization configuration method for the park distribution network as described in any of the above embodiments.

[0271] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0272] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0273] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0274] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0275] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0276] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the configuration of distributed energy storage in a park power distribution network, characterized in that, include: With the goal of minimizing operating costs, energy storage constraints are set, and energy storage configuration data is used as decision variables to construct a distributed energy storage optimization configuration model; Obtain the operation data and information of the park's power distribution network, and determine the corresponding node energy storage basic state of charge limit based on the preset power supply adaptation function and the operation data. Based on the preset new energy consumption space function, the upper limit of the basic state of charge of the corresponding node energy storage is determined according to the operating data. Based on the pre-trained language model, the running data, and the running information, the lower limit and upper limit of the basic state of charge of each node energy storage are corrected to obtain multiple dynamic node state of charge constraints. The distributed energy storage optimization configuration model is optimized and solved using the state of charge constraints of each dynamic node and the operating data to obtain the corresponding distributed energy storage optimization configuration scheme.

2. The method for optimizing the configuration of distributed energy storage in a park power distribution network according to claim 1, characterized in that, The operational data includes multiple important load level coefficients, multiple important load power, multiple continuous supply guarantee times, multiple minimum states of charge, and multiple node energy storage rated capacities. The step of determining the corresponding node energy storage basic state of charge limit based on the operational data according to the preset supply guarantee adaptation function includes: The running data is input into a preset supply guarantee adaptation function to obtain multiple supply guarantee adaptation coefficients; Based on the preset supply guarantee demand allocation function, the demand is allocated to each of the supply guarantee adaptation coefficients to obtain multiple supply guarantee demand undertaking ratios. Each of the aforementioned supply guarantee demand sharing ratios is multiplied by the corresponding important load level coefficient, important load power, and continuous supply guarantee time to obtain multiple first multiplication values; The first multiplication values ​​corresponding to each of the lowest states of charge are summed to obtain multiple reserved basic charges. Each of the reserved basic power quantities is compared with the corresponding node energy storage rated capacity to obtain multiple first ratios; Each of the lowest states of charge is summed with its corresponding first ratio to obtain the lower limit of the basic state of charge of multiple node energy storage.

3. The method for optimizing the configuration of distributed energy storage in a park power distribution network according to claim 2, characterized in that, The operational data also includes multiple predicted outputs of new energy sources, multiple absorbed new energy power, and multiple maximum states of charge. The step of determining the upper limit of the corresponding node energy storage base state of charge based on the operational data, according to a preset new energy absorption space function, includes: The predicted output of each new energy source is compared with the corresponding absorbed new energy power to obtain multiple first differences. When the first difference is greater than or equal to the preset benchmark power curtailment, the first difference is determined as the predicted power curtailment. When the first difference is less than the benchmark power curtailment, the benchmark power curtailment is determined as the predicted power curtailment. Each of the predicted curtailed power is input into a preset new energy consumption space function to obtain multiple reserved charging spaces; Each of the reserved rechargeable spaces is compared with the corresponding node energy storage rated capacity to obtain multiple second ratios; The difference between each highest state of charge and the corresponding second ratio is processed to obtain the upper limit of the basic state of charge of multiple node energy storage.

4. The method for optimizing the configuration of distributed energy storage in a park power distribution network according to claim 1, characterized in that, The step of correcting the lower limit and upper limit of the basic state of charge of each node's energy storage based on the pre-trained language model, the running data, and the running information to obtain multiple dynamic node state of charge constraints includes: The running data and the running information are input into a pre-trained language model to obtain multiple upper limit corrections for the charge state and multiple upper limit corrections for the charge state. Each of the above-mentioned charge state lower limit corrections and each of the above-mentioned charge state upper limit corrections are input into a preset boundary verification function to obtain multiple target charge state lower limit corrections and multiple target charge state upper limit corrections. Each node energy storage basic state of charge limit is summed with the corresponding target state of charge limit correction amount to obtain multiple corrected state of charge limits. The difference between the upper limit of the basic state of charge of each node energy storage and the corresponding target upper limit of the state of charge correction is processed to obtain multiple corrected upper limits of the state of charge. When the modified charge state limit is less than or equal to the corresponding modified charge state upper limit, the modified charge state limit and the corresponding modified charge state upper limit are used to construct the corresponding dynamic node charge state constraint. When the modified state of charge limit is greater than the corresponding modified state of charge upper limit, the target state of charge limit correction amount and the target state of charge upper limit correction amount corresponding to the modified state of charge limit are multiplied by the preset backoff coefficient to obtain the new target state of charge limit correction amount and the new target state of charge limit correction amount. Jump to execute the step of summing the basic state of charge (SBC) limit of each node energy storage with the corresponding target SBC correction amount to obtain multiple corrected SBC limits.

5. The method for optimizing the configuration of distributed energy storage in a park power distribution network according to claim 1, characterized in that, The step of optimizing the distributed energy storage configuration model using the state-of-charge constraints of each dynamic node and the operating data to obtain the corresponding distributed energy storage configuration scheme includes: By combining the charge state constraints of each dynamic node with the distributed energy storage optimization configuration model, the corresponding target distributed energy storage optimization configuration model is obtained. Based on a preset optimization algorithm, the target distributed energy storage optimization configuration model is optimized and solved according to the running data to obtain the corresponding distributed energy storage optimization configuration scheme.

6. The method for optimizing the configuration of distributed energy storage in a park power distribution network according to claim 1, characterized in that, The energy storage constraints include energy storage installation constraints, energy storage charging and discharging power constraints, energy storage state of charge transition constraints, new energy consumption constraints, node voltage constraints, line capacity constraints, and power flow constraints.

7. A distributed energy storage optimization configuration system for a park power distribution network, characterized in that, include: The module is used to set energy storage constraints with the goal of minimizing operating costs, and to build a distributed energy storage optimization configuration model with energy storage configuration data as the decision variable. The data acquisition module is used to acquire the operation data and information of the park's power distribution network, and based on the preset power supply adaptation function, determine the corresponding node energy storage basic state of charge limit according to the operation data. The analysis module is used to determine the upper limit of the basic state of charge of the corresponding node energy storage based on the operating data, according to the preset new energy consumption space function. The constraint module is used to correct the lower limit of the state of charge of each node's energy storage base and the upper limit of the state of charge of each node's energy storage base based on the pre-trained language large model, the running data, and the running information, so as to obtain multiple dynamic node state of charge constraints. The optimization module is used to optimize and solve the distributed energy storage optimization configuration model by using the charge state constraints of each dynamic node and the operating data, so as to obtain the corresponding distributed energy storage optimization configuration scheme.

8. An electronic device, characterized in that, The system includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the distributed energy storage optimization configuration method for a park power distribution network as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the method for optimizing the configuration of distributed energy storage in a park power distribution network as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, wherein when the program instructions are executed by a computer, the computer performs the distributed energy storage optimization configuration method for the park distribution network as described in any one of claims 1-6.

Images