A load group optimal construction method considering distributed power and energy storage

Abstract

This invention discloses an optimal load group construction method considering distributed power sources and energy storage, belonging to the field of distribution network optimization and dispatching technology. The method involves: collecting the topology, line parameters, predicted output curves of distributed power sources, and basic load data for each node of the target distribution network; establishing a lower-level load group model to clarify the technical and economic parameters of the energy storage system; constructing an upper-level optimal power flow model to obtain the optimal dispatching scheme and distribution network node electricity prices; obtaining the optimal capacity configuration and optimal dispatching strategy for the energy storage system; maximizing the utilization of distributed flexible loads; and outputting correct results that meet voltage constraints. This invention, by constructing a unified lower-level load group model, incorporates large-scale transferable loads centrally dispatched by the power grid, distributed loads managed by load aggregators, and distributed energy storage into the same framework, achieving effective integration and coordinated response of multiple types of load resources; and improving the economic efficiency of energy storage investment and the flexibility of operation and dispatching.

Description

Technical Field

[0001] This invention belongs to the field of intelligent distribution network optimization and scheduling technology, specifically relating to an optimal load group construction method that considers distributed power sources and energy storage. Background Technology

[0002] With the rapid development of new power systems, distribution networks are undergoing a profound transformation from the traditional one-way power supply mode to a multi-dimensional interactive mode involving power generation, grid, load, and storage. Against this backdrop, the penetration rate of distributed power sources (such as photovoltaic and wind power) continues to increase, and the intermittency and strong uncertainty of their output pose significant challenges to the power balance, voltage stability, and operational economy of distribution networks.

[0003] Load aggregators integrate adjustable loads on the user side to form large-scale flexible resources, providing a new means of regulation for the power grid. Meanwhile, energy storage systems, due to their flexible charging and discharging characteristics, have become key devices for smoothing fluctuations and improving absorption capacity. Current research on load resource aggregation and regulation has made some progress, but traditional load group construction methods often focus on single types of distributed adjustable loads, or are limited to response optimization in local areas, or even fail to consider network topology. In actual distribution network operation, load resources exhibit multi-level and multi-attribute characteristics: they include both geographically dispersed distributed adjustable loads with varying response characteristics, and large-scale transferable loads that can be directly and centrally controlled by the dispatch center.

[0004] Therefore, designing an optimal load group construction method that can comprehensively consider the output characteristics of distributed power sources, the charging and discharging strategies of distributed energy storage, the response capability of distributed adjustable loads, and the potential of centralized transferable load dispatching in complex distribution network environments has become a key issue that urgently needs to be addressed in the operation and planning of smart distribution networks.

[0005] Considering the topology and physical constraints of distribution networks, traditional centralized optimal power flow methods rely heavily on global information and unified control, making them ill-suited for distribution network environments with numerous distributed resources. In particular, when network constraints (such as voltage limits) are directly incorporated into the optimization model, the locality and asynchronicity of responses from distributed load aggregators and distributed energy storage may lead to infeasible solutions or convergence difficulties in the distributed algorithm. Furthermore, existing methods typically separate energy storage capacity planning from operation scheduling, and voltage safety checks are often performed offline or post-hoc, lacking effective mechanisms for real-time prevention and correction of voltage exceedances during the dynamic response of distributed resources, potentially leading to localized voltage instability risks.

[0006] Therefore, there is an urgent need for a method that can coordinate and optimize distributed power sources, load aggregators, and energy storage systems in a distributed computing environment, so as to achieve a comprehensive improvement in operational economy and resource utilization efficiency while ensuring the voltage safety of the distribution network. Summary of the Invention

[0007] The purpose of this invention is to provide a wireless high-voltage current measurement method with filtering and anti-interference function to solve the problems mentioned in the background art.

[0008] The objective of this invention is achieved as follows: an optimal load group construction method considering distributed power sources and energy storage, characterized by the following steps:

[0009] Step S1: Collect the topology, line parameters, predicted output curves of distributed power sources, and basic load data of each node of the target distribution network.

[0010] Step S2: Establish a lower-level load group model to clarify the technical and economic parameters of the energy storage system;

[0011] Step S3: Construct an upper-level optimal power flow model, with the goal of minimizing the total operating cost of the distribution network system, to obtain the optimal scheduling scheme and the electricity price at the distribution network nodes;

[0012] Step S4: Construct the lower-level energy storage configuration and run the joint optimization model to obtain the optimal capacity configuration and optimal scheduling strategy of the energy storage system;

[0013] Step S5: Construct a load dynamic adjustment model based on voltage safety constraints to maximize the utilization of distributed flexible loads;

[0014] Step S6: Establish an iterative solution framework until the system is co-optimal, and output the correct results that meet the voltage constraints.

[0015] Preferably, in step S2, establishing the lower-level load group model to clarify the technical and economic parameters of the energy storage system specifically involves:

[0016] The lower-level load group model includes the transferable load model under centralized grid dispatch and the distributed load aggregator model. The transferable load model is as follows:

[0017] ;

[0018] In the formula, This represents the power regulation of the transferable load i at time t. Represents the set of transferable loads. Let t represent the set of times, i.e., a scheduling cycle; the above formula shows that the total energy of the transferable load model remains unchanged within a complete scheduling cycle, i.e., the load curve in the time dimension satisfies energy conservation.

[0019] ;

[0020] In the formula, and These represent the upper and lower limits of the power regulation of the transferable load, respectively. The above formula shows that, within any scheduling period t, the power regulation value of the transferable load must be within its feasible technical upper and lower limits.

[0021] ;

[0022] In the formula, and These represent the baseline compensation cost and marginal incremental cost coefficient corresponding to the unit power adjustment of the transferable load, respectively.

[0023] Preferably, the distributed load aggregator model is as follows:

[0024] ;

[0025] In the formula, This represents the total load power aggregated by distributed load aggregator i at time t. Represents the set of distributed load aggregators;

[0026] ;

[0027] In the formula, the coefficients and These represent the baseline revenue coefficient and the marginal revenue diminishing coefficient, respectively. The distributed load aggregator model defines the revenue function of the distributed load aggregator to quantify the economic incentives it receives for participating in system scheduling.

[0028] Preferably, the construction of the upper-level optimal power flow model in step S3 specifically includes:

[0029] Based on second-order cone programming, the power flow equations of the distribution network are first treated with phase angle relaxation and second-order cone relaxation, resulting in the following constraints:

[0030] ;

[0031] ;

[0032] ;

[0033] ;

[0034] ;

[0035] ;

[0036] ;

[0037] ;

[0038] in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the resistance and reactance on line l, respectively. This represents the square of the current in line l. and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the predicted amount of new energy power generation. Represents a collection of distributed new energy sources; and These are the indices of the start and end nodes of the power distribution line l; and The set of indices representing the starting and ending nodes of the power distribution line l, respectively, and N is the index of the power distribution network node;

[0039] In a distribution network, the objective function should be to minimize the total operating cost of the distribution network. The total operating cost of the system mainly includes two items: the cost of purchasing electricity from the upper-level grid to make up for the power shortage, and the compensation cost paid for regulating the transferable loads under centralized dispatch. The upper-level optimal power flow model is as follows:

[0040] ;

[0041] in, and These represent the electricity purchase price from the upstream power grid and the amount of electricity purchased from the upstream power grid at time t, respectively.

[0042] Preferably, the optimal scheduling scheme and distribution network node electricity price are obtained in step S3, specifically as follows:

[0043] The convexized optimal power flow model was numerically solved using the Gurobi commercial solver. A branch-and-bound algorithm was employed to correct relaxation errors, ensuring the solution accuracy met engineering requirements. Two core results were output: first, the optimal power of transferable loads under centralized grid dispatch in each time period; and second, the marginal electricity price at each node. The electricity price is represented by the shadow price of node power, reflecting the impact of changes in unit power at that node on the total operating cost of the system.

[0044] Preferably, in step S4, a joint optimization model for lower-level energy storage configuration and operation is constructed to obtain the optimal capacity configuration and optimal scheduling strategy of the energy storage system, specifically as follows:

[0045] The joint optimization model for lower-level energy storage configuration and operation is as follows:

[0046] ;

[0047] ;

[0048] ;

[0049] ;

[0050] In the formula, This represents the total cost of the joint optimization model for energy storage configuration and operation. Represents the daily investment cost of energy storage, etc. Represents the daily operating cost of energy storage. Y represents the daily maintenance cost of energy storage; Y represents the annual planning cycle of energy storage. Represents the annual interest rate. and These represent the rated capacity of the energy storage configuration and the marginal investment cost coefficient per unit capacity of energy storage, respectively. Index representing energy storage devices; and These represent the energy storage charging / discharging power, The maintenance cost coefficient representing the unit power of energy storage.

[0051] Preferably, the constraints of the lower-level energy storage configuration-operation joint optimization model are:

[0052] ;

[0053] in, Represents the state of charge of the energy storage device. and These represent the energy storage charging / discharging efficiency, Represents a unit of time step; Constrained state of charge of energy storage;

[0054] Limiting energy storage while maintaining charge / discharge states:

[0055] ;

[0056] The real-time charging and discharging power of energy storage batteries should not exceed their rated power.

[0057] ;

[0058] ;

[0059] in, and These represent the proportional coefficients for charging / discharging power and energy storage capacity, respectively.

[0060] Preferably, in step S5, a load dynamic adjustment model based on voltage safety constraints is constructed to maximize the utilization of distributed flexible loads, specifically as follows:

[0061] The load dynamic adjustment model is as follows:

[0062] ;

[0063] in, This represents the originally planned grid-connected power of the distributed load aggregator. This represents the final grid-connected power of the load aggregator after dynamic adjustment;

[0064] The constraints are:

[0065] ;

[0066] ;

[0067] ;

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the square of the current in line l; and These represent the resistance and reactance on line l, respectively; This represents the voltage at the starting node of power distribution line l. This represents the voltage at the termination point of the power distribution line l; and These represent the lower and upper limits of reactive load on node n, respectively; This represents the voltage at node n.

[0073] Preferably, the load dynamic adjustment model adds voltage safety constraints, which are as follows:

[0074] ;

[0075] ;

[0076] ;

[0077] in, and These represent the upper and lower limits of the voltage at node n, respectively;

[0078] The design of voltage safety constraints is to relax the voltage constraints and use the preliminary solution of distributed computing to allow the voltage to temporarily deviate from the rated range, while retaining the deviation identifier of the voltage exceeding the limit. After the distributed preliminary solution is generated, based on the voltage deviation identifier and the load dynamic adjustment model based on the voltage safety constraints, the load power of the nodes that exceed the voltage limit is optimized and reduced in a centralized coordination manner. For nodes with excessive power injection, the load injection power is reduced, and for nodes with excessive power outflow, the load outflow power is reduced, so as to correct the node voltage to the safe and stable range.

[0079] Compared with the prior art, the present invention has the following improvements and advantages:

[0080] 1. By constructing a unified lower-level load group model, large-scale transferable loads that can be centrally dispatched by the power grid, distributed loads managed by load aggregators, and distributed energy storage are incorporated into the same framework, thereby achieving effective integration and coordinated response of multiple types of load resources.

[0081] 2. By using the energy storage configuration-operation joint optimization method, the traditional paradigm of separating planning and operation is broken through, which significantly improves the economic efficiency of energy storage investment and the flexibility of operation scheduling, so that its regulation capacity can be fully utilized.

[0082] 3. By using a dynamic adjustment model based on voltage safety constraints and embedding a dynamic prevention-correction mechanism, the risk of voltage exceeding limits can be detected and automatically corrected in real time during the load aggregator response process. This transforms safety verification from "post-event verification" to "process control," effectively improving voltage stability and power supply reliability during distributed resource interaction. Attached Figure Description

[0083] Figure 1 This is a flowchart illustrating the method of the present invention.

[0084] Figure 2 This is a coupled topology diagram of the distributed resources within the power distribution network of this invention.

[0085] Figure 3This is a schematic diagram showing the predicted output results of photovoltaic and wind power.

[0086] Figure 4 This is a graph showing the changes in power and state of charge of the energy storage facility.

[0087] Figure 5 The result is a graph showing the voltage change curve and the identification of the over-limit region.

[0088] Figure 6 The graph shows the verification results of the adjusted voltage curve.

[0089] Figure 7 The diagram shows the optimization results of distributed load aggregator scheduling. Detailed Implementation

[0090] The invention will be further summarized below with reference to the accompanying drawings.

[0091] like Figure 1 As shown, step S1: Collect the topology, line parameters, predicted output curves of distributed power sources, and basic load data of each node of the target distribution network.

[0092] Step S2: Establish a lower-level load group model to clarify the technical and economic parameters of the energy storage system, specifically:

[0093] The lower-level load group model includes the transferable load model under centralized grid dispatch and the distributed load aggregator model. The transferable load model is as follows:

[0094] ;

[0095] In the formula, This represents the power regulation of the transferable load i at time t. Represents the set of transferable loads. Let t represent the set of times, i.e., a scheduling cycle; the above formula shows that the total energy of the transferable load model remains unchanged within a complete scheduling cycle, i.e., the load curve in the time dimension satisfies energy conservation.

[0096] ;

[0097] In the formula, and These represent the upper and lower limits of the power regulation of the transferable load, respectively. The above formula shows that, within any scheduling period t, the power regulation value of the transferable load must be within its feasible technical upper and lower limits.

[0098] ;

[0099] In the formula, and These represent the baseline compensation cost and marginal incremental cost coefficient corresponding to the unit power adjustment of the transferable load, respectively.

[0100] The distributed load aggregator model is as follows:

[0101] ;

[0102] In the formula, This represents the total load power aggregated by distributed load aggregator i at time t. Represents the set of distributed load aggregators;

[0103] ;

[0104] In the formula, the coefficients and These represent the baseline revenue coefficient and the marginal revenue diminishing coefficient, respectively. The distributed load aggregator model defines the revenue function of the distributed load aggregator to quantify the economic incentives it receives for participating in system scheduling.

[0105] Establish a mapping relationship between various loads and energy storage systems at the access nodes in the distribution network, which serves as the physical basis for subsequent power regulation and voltage control.

[0106] In step S3, an upper-level optimal power flow model is constructed with the objective of minimizing the total operating cost of the distribution network system. This yields the optimal scheduling scheme and the electricity price at each node of the distribution network, which are:

[0107] Based on second-order cone programming, the power flow equations of the distribution network are first treated with phase angle relaxation and second-order cone relaxation, resulting in the following constraints:

[0108] ;

[0109] ;

[0110] ;

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] ;

[0116] in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the resistance and reactance on line l, respectively. This represents the square of the current in line l. and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the predicted amount of new energy power generation. Represents a collection of distributed new energy sources; and These are the indices of the start and end nodes of the power distribution line l; and The set of indices representing the starting and ending nodes of the power distribution line l, respectively, and N is the index of the power distribution network node;

[0117] Voltage constraints are not considered in the optimal power flow model because the integration of distributed generation sources and loads can alter power flow and potentially cause local voltage exceedances. In centralized optimization, this problem can be addressed through global coordination; however, in distributed computing, each node optimizes independently based on its own information, making global voltage coordination difficult and potentially leading to unsolvable problems. To ensure the model is solvable in a distributed architecture, voltage constraints are not considered at this stage.

[0118] In a distribution network, the objective function should be to minimize the total operating cost of the distribution network. The total operating cost of the system mainly includes two items: the cost of purchasing electricity from the upper-level grid to make up for the power shortage, and the compensation cost paid for regulating the transferable loads under centralized dispatch. The upper-level optimal power flow model is as follows:

[0119] ;

[0120] in, and These represent the electricity purchase price from the upstream power grid and the amount of electricity purchased from the upstream power grid at time t, respectively.

[0121] The convexized optimal power flow model was numerically solved using the Gurobi commercial solver. A branch-and-bound algorithm was employed to correct relaxation errors, ensuring the solution accuracy met engineering requirements. Two core results were output: first, the optimal power of transferable loads under centralized grid dispatch in each time period; and second, the marginal electricity price at each node. The electricity price is represented by the shadow price of node power, reflecting the impact of changes in unit power at that node on the total operating cost of the system.

[0122] The nodal electricity price in the distribution network is the nodal marginal electricity price. By solving the optimized distribution network power flow model, and using the Lagrange multiplier method or duality theory in optimization theory, the nodal marginal electricity price of each node at each time step is obtained. Its value is the dual variable of the active power balance constraint of the distribution network. The marginal electricity price at a node is essentially a reflection of the spatiotemporal value of electrical energy. Its value equals the marginal system cost incurred to meet the increased unit electricity demand at a given moment at that node, while satisfying grid operation constraints. It can be decomposed into the external grid purchase price, congestion price, and reactive power support price.

[0123] In step S4, a joint optimization model for lower-level energy storage configuration and operation is constructed to obtain the optimal capacity configuration and optimal scheduling strategy of the energy storage system, specifically as follows:

[0124] The joint optimization model for lower-level energy storage configuration and operation is as follows:

[0125] ;

[0126] ;

[0127] ;

[0128] ;

[0129] In the formula, This represents the total cost of the joint optimization model for energy storage configuration and operation. Represents the daily investment cost of energy storage, etc. Represents the daily operating cost of energy storage. Y represents the daily maintenance cost of energy storage; Y represents the annual planning cycle of energy storage. Represents the annual interest rate. and These represent the rated capacity of the energy storage configuration and the marginal investment cost coefficient per unit capacity of energy storage, respectively. Index representing energy storage devices; and These represent the energy storage charging / discharging power, The maintenance cost coefficient representing the unit power of energy storage.

[0130] The constraints of the joint optimization model for layered energy storage configuration and operation are as follows:

[0131] ;

[0132] in, Represents the state of charge of the energy storage device. and These represent the energy storage charging / discharging efficiency, Represents a unit of time step; Constrained state of charge of energy storage;

[0133] Limiting energy storage while maintaining charge / discharge states:

[0134] ;

[0135] The real-time charging and discharging power of energy storage batteries should not exceed their rated power.

[0136] ;

[0137] ;

[0138] in, and These represent the proportional coefficients for charging / discharging power and energy storage capacity, respectively.

[0139] Step S5 involves constructing a load dynamic adjustment model based on voltage safety constraints to maximize the utilization of distributed flexible loads. Specifically:

[0140] The load dynamic adjustment model is as follows:

[0141] ;

[0142] in, This represents the originally planned grid-connected power of the distributed load aggregator. This represents the final grid-connected power of the load aggregator after dynamic adjustment;

[0143] The constraints are:

[0144] ;

[0145] ;

[0146] ;

[0147] ;

[0148] ;

[0149] ;

[0150] ;

[0151] in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the square of the current in line l; and These represent the resistance and reactance on line l, respectively; This represents the voltage at the starting node of power distribution line l. This represents the voltage at the termination point of the power distribution line l; and These represent the lower and upper limits of reactive load on node n, respectively; This represents the voltage at node n.

[0152] To verify the feasibility and effectiveness of this invention, the following experiments were conducted:

[0153] This invention selects an improved IEEE 33-node distribution network for simulation. The coupling positions of distributed resources, distributed energy storage facilities, centrally dispatched transferable loads, and distributed load aggregators with the IEEE 33-node distribution network are as follows: Figure 2 As shown.

[0154] Basic parameter settings: and Defined as 0.002 yuan / kWh 2 and 0.1 yuan / kWh; and Defined as 0.0005 yuan / kWh 2 And 1.2 yuan / kWh; the annual planning period Y for energy storage is set to 10 years, and the marginal investment price cost coefficient per unit capacity of energy storage is... Set to 0.0008 yuan / kWh 2 , Set to 0.002 yuan / kWh 2 . and All were set to 0.98. and All are set to 1 / 3. The power baseline for the distribution network is 10MWh, and the upper and lower limits of the voltage amplitude at each node are 1.05pu and 0.9pu, respectively; the predicted values ​​for distributed energy resources are as follows: Figure 3 As shown.

[0155] Figure 4 illustrates the optimal construction method for transferable loads described in this invention. The transferable loads achieve peak shaving and valley filling through positive and negative transfers. The bar chart below quantifies the percentage change in load relative to each time period. The transferable loads increase electricity consumption during periods of excess renewable energy output (such as at night) and decrease electricity consumption during periods of insufficient output (such as midday peak), creating space for renewable energy consumption.

[0156] Figure 5 illustrates the coordinated operation characteristics of energy storage facilities under the optimal load group framework of this invention, including the optimized energy storage capacity configuration and 24-hour energy storage power scheduling results. The state-of-charge curve in the figure clearly and intuitively demonstrates the dynamic changes in the storage and release of stored electrical energy.

[0157] Figure 6 compares the voltage operation status of distribution network nodes without and with the strategy of this invention. Without the optimal load group construction method described in this invention, due to the lack of overall control over transferable loads and energy storage within the load group, the voltage curves of each node show significant deviations from the safe operating limit at 1-9, 13-17, and 24 hours, falling below the safe operating lower limit of 0.9 pu. However, after optimization using the optimal load group construction method of this invention, which considers distributed power sources and energy storage, through the coordinated control of each unit within the load group, the voltage curves of all nodes stabilize within the safe amplitude range above 0.9 pu, effectively improving the overall voltage stability of the distribution network.

[0158] Figure 7 shows the comparison results before and after optimizing the load aggregator load using the load group optimal construction method considering distributed power sources and energy storage described in this invention, with voltage safety constraints as boundary conditions.

[0159] From the simulation and quantitative indicators, the proposed method shows significant effects in renewable energy consumption and system voltage stabilization: without rational load group construction, the total daily curtailment of wind and solar power in the system was 2903.2 kW; after constructing load groups, the total daily curtailment of wind and solar power in the system decreased to 598.81 kW. Without rational load group construction, the total daily operating cost of the distribution network system was 20387.2 yuan, indicating that a large amount of curtailed wind and solar power did not play a role in renewable energy; with rational load group construction, the total daily operating cost of the distribution network system was 9876.3 yuan.

[0160] The above description is merely an embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of the claims of the present invention.

Claims

1. A method for optimal construction of a load group considering distributed power and energy storage, characterized in that: The method includes the following steps: Step S1: Collect the topology, line parameters, predicted output curves of distributed power sources, and basic load data of each node of the target distribution network. Step S2: Establish a lower-level load group model to clarify the technical and economic parameters of the energy storage system; Step S3: Construct an upper-level optimal power flow model, with the goal of minimizing the total operating cost of the distribution network system, to obtain the optimal scheduling scheme and the electricity price at the distribution network nodes; Step S4: Construct the lower-level energy storage configuration and run the joint optimization model to obtain the optimal capacity configuration and optimal scheduling strategy of the energy storage system; Step S5: Construct a load dynamic adjustment model based on voltage safety constraints to maximize the utilization of distributed flexible loads; Step S6: Establish an iterative solution framework until the system is co-optimal, and output the correct results that meet the voltage constraints.

2. The method of claim 1, wherein the method further comprises: In step S2, establishing the lower-level load group model clarifies the technical and economic parameters of the energy storage system, specifically as follows: The lower-level load group model includes the transferable load model under centralized grid dispatch and the distributed load aggregator model. The transferable load model is as follows: ； wherein, denotes the power adjustment amount of the transferable load i at time t, denotes the set of transferable loads, denotes the set of times, i.e. one scheduling period; the above equation shows that the total energy of the transferable load model remains constant over a complete scheduling period, i.e. the flatness of its load curve in time dimension satisfies the energy conservation; ； In the formula, and These represent the upper and lower limits of the power regulation of the transferable load, respectively. The above formula shows that, within any scheduling period t, the power regulation value of the transferable load must be within its feasible technical upper and lower limits; ； In the formula, and These represent the baseline compensation cost and marginal incremental cost coefficient corresponding to the unit power adjustment of the transferable load, respectively.

3. The optimal load group construction method considering distributed power sources and energy storage according to claim 2, characterized in that: The distributed load aggregator model is as follows: ； In the formula, This represents the total load power aggregated by distributed load aggregator i at time t. Represents the set of distributed load aggregators; ； In the formula, the coefficients and These represent the baseline revenue coefficient and the marginal revenue diminishing coefficient, respectively. The distributed load aggregator model defines the revenue function of the distributed load aggregator to quantify the economic incentives it receives for participating in system scheduling.

4. The optimal load group construction method considering distributed power sources and energy storage according to claim 1, characterized in that: The construction of the upper-level optimal power flow model in step S3 is specifically as follows: Based on second-order cone programming, the power flow equations of the distribution network are first treated with phase angle relaxation and second-order cone relaxation, resulting in the following constraints: ； ； ； ； ； ； ； ； in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the resistance and reactance on line l, respectively. This represents the square of the current in line l. and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the predicted amount of new energy power generation. Represents a collection of distributed new energy sources; and These are the indices of the start and end nodes of the power distribution line l; and The set of indices representing the starting and ending nodes of the power distribution line l, respectively, and N is the index of the power distribution network node; In a distribution network, the objective function should be to minimize the total operating cost of the distribution network. The total operating cost of the system mainly includes two items: the cost of purchasing electricity from the upper-level grid to make up for the power shortage, and the compensation cost paid for regulating the transferable loads under centralized dispatch. The upper-level optimal power flow model is as follows: ； in, and These represent the electricity purchase price from the upstream power grid and the amount of electricity purchased from the upstream power grid at time t, respectively.

5. The optimal load group construction method considering distributed power sources and energy storage according to claim 4, characterized in that: The optimal scheduling scheme and distribution network node electricity price are obtained in step S3 as follows: The convexized optimal power flow model was numerically solved using the Gurobi commercial solver. A branch-and-bound algorithm was employed to correct relaxation errors, ensuring the solution accuracy met engineering requirements. Two core results were output: first, the optimal power of transferable loads under centralized grid dispatch in each time period; and second, the marginal electricity price at each node. The electricity price is represented by the shadow price of node power, reflecting the impact of changes in unit power at that node on the total operating cost of the system.

6. The optimal load group construction method considering distributed power sources and energy storage according to claim 1, characterized in that: In step S4, a joint optimization model for lower-level energy storage configuration and operation is constructed to obtain the optimal capacity configuration and optimal scheduling strategy of the energy storage system. Specifically: The joint optimization model for lower-level energy storage configuration and operation is as follows: ； ； ； ； In the formula, This represents the total cost of the joint optimization model for energy storage configuration and operation. Represents the daily investment cost of energy storage, etc. Represents the daily operating cost of energy storage. Represents the daily maintenance cost of energy storage; Y represents the annual planning cycle for energy storage. Represents the annual interest rate. and These represent the rated capacity of the energy storage configuration and the marginal investment cost coefficient per unit capacity of energy storage, respectively. Index representing energy storage devices; and These represent the energy storage charging / discharging power, The maintenance cost coefficient representing the unit power of energy storage.

7. The optimal load group construction method considering distributed power sources and energy storage according to claim 1, characterized in that: The constraints of the lower-level energy storage configuration-operation joint optimization model are as follows: ； in, Represents the state of charge of the energy storage device. and These represent the energy storage charging / discharging efficiency, Represents a unit of time step; Constrained state of charge of energy storage; Limiting energy storage while maintaining charge / discharge states: ； The real-time charging and discharging power of energy storage batteries should not exceed their rated power. ； ； in, and These represent the proportional coefficients for charging / discharging power and energy storage capacity, respectively.

8. The optimal load group construction method considering distributed power sources and energy storage according to claim 1, characterized in that: In step S5, a load dynamic adjustment model based on voltage safety constraints is constructed to maximize the utilization of distributed flexible loads. Specifically: The load dynamic adjustment model is as follows: ； in, This represents the originally planned grid-connected power of the distributed load aggregator. This represents the final grid-connected power of the load aggregator after dynamic adjustment; The constraints are: ； ； ； ； ； ； ； in, and These represent the active and reactive power flows flowing through line l at time t, respectively. This represents the net load of each distributed resource on node n; and These represent the conductance and susceptance of node n, respectively. Represents complex power. This represents the square of the current in line l; and These represent the resistance and reactance on line l, respectively; This represents the voltage at the starting node of power distribution line l. This represents the voltage at the termination point of the power distribution line l; and These represent the lower and upper limits of reactive load on node n, respectively; This represents the voltage at node n.

9. The optimal load group construction method considering distributed power sources and energy storage according to claim 8, characterized in that: The load dynamic adjustment model adds voltage safety constraints, which are as follows: ； ； ； in, and These represent the upper and lower limits of the voltage at node n, respectively; The design of voltage safety constraints is to relax the voltage constraints and use the preliminary solution of distributed computing to allow the voltage to temporarily deviate from the rated range, while retaining the deviation identifier of the voltage exceeding the limit. After the distributed preliminary solution is generated, based on the voltage deviation identifier and the load dynamic adjustment model based on the voltage safety constraints, the load power of the nodes that exceed the voltage limit is optimized and reduced in a centralized coordination manner. For nodes with excessive power injection, the load injection power is reduced, and for nodes with excessive power outflow, the load outflow power is reduced, so as to correct the node voltage to the safe and stable range.

Images