Source network user shared energy storage optimal operation scheduling method considering energy storage capacity distribution mechanism

By generating typical daily curves and a dynamic capacity allocation mechanism, and combining the particle swarm optimization algorithm to optimize the scheduling model, the problem of insufficient flexibility in the existing source-grid-user shared energy storage scheduling model is solved. This achieves efficient utilization of the energy storage system and three-way collaborative optimization, thereby improving the scientific nature and flexibility of power grid operation.

CN120879792APending Publication Date: 2025-10-31HEFEI UNIV OF TECH
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202511029067.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-25
Publication Date
2025-10-31

AI Technical Summary

Technical Problem

Existing optimal energy storage operation and scheduling models lack research on energy storage sharing between power sources, grids, and users. They have limited application scenarios, insufficient system scheduling flexibility, and fail to fully consider the differentiated needs of the power source, grid, and user sides, resulting in low energy storage utilization and limited effect in mitigating fluctuations.

Method used

By generating typical daily new energy output, user-side load, and grid-side load curves, a source-grid-user-side energy storage capacity allocation mechanism is established. The particle swarm optimization algorithm is used to solve the optimal operation and scheduling model of shared energy storage. By combining the capacity demand benchmark and the high-dimensional particle swarm optimization solution strategy, dynamic allocation and optimized scheduling of energy storage capacity are achieved.

Benefits of technology

It significantly improves energy storage utilization, effectively reduces system power deficit, smooths out new energy fluctuations and load deviations, enhances the engineering practicality and adaptability of the scheduling model, ensures that charging and discharging strategies meet actual needs, and achieves a win-win situation for the power generation, grid, and users.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120879792A_ABST
    Figure CN120879792A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of energy storage systems, in particular to a source network user shared energy storage optimal operation scheduling method considering energy storage capacity distribution. The method comprises the following steps: acquiring historical power generation and power consumption data and shared energy storage power station configuration data of a source network user side, and generating a typical daily new energy output and load curve; providing an energy storage capacity distribution mechanism of a source network user side, calculating a demand factor of each side for the energy storage capacity, and carrying out quantitative evaluation distribution on the energy storage capacity; based on a capacity allocation mechanism, establishing a source network user side shared energy storage optimal operation scheduling model; constructing a source network user side shared energy storage charging and discharging operation constraint and an energy storage capacity constraint; and solving the model by adopting a particle swarm algorithm. According to the method, an energy storage capacity distribution mechanism is researched, a source network user side shared energy storage optimal operation scheduling model is established, and the operation scheduling model is solved according to historical data, so that an optimal operation scheduling strategy is found, and basis and reference are provided for power grid operation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of energy storage system technology, and more specifically to an optimal operation and scheduling method for shared energy storage by source, grid and user, taking into account the energy storage capacity allocation mechanism. Background Technology

[0002] With the increasing global demand for clean energy and growing environmental awareness, renewable energy sources, such as solar and wind power, are playing an increasingly important role in the power system. While renewable energy boasts advantages like abundant resources and environmental friendliness, it also suffers from intermittency, volatility, and randomness. Energy storage systems can smooth out fluctuations in intermittent energy sources through reasonable charging and discharging strategies, improving the absorption and regulation capabilities of renewable energy generation. On the power generation side, energy storage is needed to assist in stable power generation and participate in frequency and voltage regulation; on the grid side, it is needed to optimize power flow and enhance transmission reliability; and on the user side, it is expected to reduce electricity costs and improve power supply reliability. However, due to the high construction costs of energy storage systems, independent configurations are insufficient to meet the economic and functional needs of diverse stakeholders. Therefore, shared energy storage models are becoming a trend, and energy storage operation and scheduling models provide crucial support for the scientific capacity allocation and optimal operation scheduling of energy storage systems.

[0003] Currently, there is considerable research on the optimized scheduling of energy storage technology. Existing research on energy storage scheduling mainly focuses on establishing shared energy storage scheduling models from both the power supply side and the user side. Regarding the power supply side, Bao Lushan, in his paper "Research on Joint Optimization Scheduling Method of Wind, Solar and Energy Storage in Power Systems [J]", published *Power Plant Auxiliary Machinery*, Vol. 46, No. 2, 2025-6, disclosed a joint optimization scheduling method of wind, solar and energy storage. This method considers factors such as power balance, charging and discharging efficiency, and ramp rate as constraints, and aims to achieve the synergistic goal of minimizing system operating costs and maximizing renewable energy consumption. It establishes a joint scheduling model of wind, solar and energy storage to improve renewable energy consumption and the flexibility of the power supply side of the power system. On the user side, Fu Linbei, Wang Haisheng, Li Zhongzhong, Ma Lihong, Zhou Hang, Zhang Changjun, and Cui Jianzhao proposed an optimized dispatch strategy for mountain microgrids considering user-side shared energy storage [J]. Zhejiang Electric Power, Vol. 44, No. 6, 2025-6, aiming to reduce the total cost of microgrid systems, to further integrate a high proportion of renewable energy into the microgrid and improve the utilization rate of renewable energy. Huo Jinquan, Lin Hong, and Tian Yizhi proposed a two-stage optimized operation strategy for multi-microgrid systems considering shared energy storage leasing [J]. New Energy Storage and Charging Facility Planning, Vol. 53, No. 6, 2025, addressing the issues of power mutual assistance and local consumption of new energy in multi-microgrid grid-connected systems, to improve the energy efficiency and economy of microgrid systems. In the field of source-grid-load-storage coordinated optimization scheduling, Song Mingshu, Su Changsheng, Xu Xinyu, Zhu Qing, Wu Maoqian, and Xiao Zhongjie, "Source-grid-load-storage coordinated optimization scheduling considering power system flexibility," Information Technology, 2025, No. 6, established an objective function by coordinating factors such as fuel cost and start-up / shutdown cost, deep peak-shaving cost, energy storage call-up cost, and renewable energy curtailment penalty. Considering constraints such as unit output, ramp-up, start-up / shutdown constraints, system balance, grid capacity, and energy storage charging and discharging power, a source-grid-load-storage coordinated optimization scheduling model considering power system flexibility was established.

[0004] Existing optimal operation and scheduling models for energy storage lack research on the direction of shared energy storage between the source, grid, and user, have limited application scenarios, and insufficient system scheduling flexibility. Therefore, the operation and scheduling problem of shared energy storage between the source, grid, and user requires further research. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes an optimal operation and scheduling model for shared energy storage that considers the energy storage capacity allocation mechanism. By finding the optimal operation and scheduling strategy for the source, grid, and users, it provides a basis and reference for grid operation, thereby achieving a win-win situation for the source, grid, and users and promoting the efficient application and sustainable development of the shared energy storage model in the power system.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] An optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, includes the following steps:

[0008] S1. Obtain historical power generation and consumption data from the source-grid-user side, as well as configuration data of shared energy storage power stations;

[0009] S2. Based on historical power generation and consumption data from the source-grid-user side, generate typical daily curves for renewable energy output, user-side load, and grid-side load.

[0010] S3. Establish a source-grid-user side energy storage capacity allocation mechanism and calculate the energy storage capacity allocated to each side;

[0011] S4. Based on the energy storage capacity allocation mechanism of the source-grid-user side, establish an optimal operation and scheduling model for shared energy storage on the source-grid-user side;

[0012] S5. Use the particle swarm optimization algorithm to solve the optimal operation scheduling model of shared energy storage on the source-grid-user side, and determine the optimal charging and discharging operation scheduling strategy of shared energy storage on the source-grid-user side that optimizes the objective.

[0013] Furthermore, the historical power generation and consumption data on the source-grid-user side includes: historical output data of new energy power plants, historical load data of the power grid, and historical electricity consumption data on the user side; the configuration data of the shared energy storage power station includes the technical parameters of the shared energy storage, which include charge and discharge efficiency, rated capacity, and rated power.

[0014] Furthermore, step S2 specifically includes the following steps:

[0015] S21. Use clustering algorithms to determine the optimal number of typical days;

[0016] S22. Based on the optimal number of typical days, generate the typical day renewable energy output curve, the typical day load curve on the user side, and the typical day load curve on the grid side.

[0017] Furthermore, step S21 specifically includes the following steps:

[0018] S211, Preset upper limit of the number of typical days Different values ​​were selected sequentially as the candidate number of typical days. Determine the number of typical days The range of values ​​for: .

[0019] S212, For each candidate The value is calculated using equation (1). Cost function corresponding to the value :

[0020] (1);

[0021] in, The k-th cluster center is randomly selected, where k represents the cluster center number. Indicates the first Output curve, , Indicates load or new energy source. Indicates the source network user side, Therefore Clusters centered around.

[0022] S213, Obtain the cost function The point where the rate of descent suddenly decreases, and the point corresponding to this sudden decrease. The value is used as the optimal number of typical days.

[0023] Furthermore, in step S22, generating a typical daily renewable energy output curve specifically includes the following steps:

[0024] S221. Based on the historical power output curve of the power supply side, select... Cluster centers .

[0025] S222. For each historical output curve, calculate its correlation with... Cluster centers The Euclidean distance is calculated, and the curve is assigned to the nearest cluster.

[0026] S223. Calculate the mean of the historical power output curves of all new energy sources on the power supply side within each cluster, and use this mean as the new cluster center. .

[0027] S224. Repeat steps S221-S223 until cluster centers are reached. When the change is less than a preset threshold or the maximum number of iterations is reached, the iteration stops and outputs the result. Cluster centers This forms a typical daily new energy output curve.

[0028] Furthermore, step S3 specifically includes the following steps:

[0029] S31. Based on the historical power generation and consumption data of the source-grid-household side, determine the capacity demand benchmark for each side.

[0030] S32. Based on the capacity demand benchmarks of each side of the source, grid, and user, calculate the demand factor of each side for energy storage capacity using equation (2). :

[0031] (2);

[0032] in, This is a demand factor for energy storage capacity on each side, used to quantitatively assess the allocation of energy storage capacity; For source network user-side identification; This serves as the baseline for capacity requirements on each side.

[0033] S33. Calculate the energy storage capacity allocated to each side of the source grid and user using formula (3). :

[0034] (3);

[0035] in, This indicates the rated capacity of the shared energy storage.

[0036] Furthermore, the optimization objective of the source-grid-user side shared energy storage optimal operation scheduling model is to minimize the system power deficit, and the decision variables are the hourly charging and discharging power of the power source side energy storage, the hourly charging and discharging power of the grid side energy storage, and the hourly charging and discharging power of the user side energy storage.

[0037] Furthermore, step S4 specifically includes the following steps:

[0038] S41. Calculate the minimum system power deficit using equation (4):

[0039] (4);

[0040] in, for The power deficit of the source-grid-user shared energy storage system at any given time. Let t be the load fluctuation power of the shared energy storage system between the source, grid, and users. The fluctuating power of new energy generation during time period t. express Side energy storage Charging power at any time Indicates energy storage on the power supply side Charging power at any time Indicates grid-side energy storage Charging power at any time Indicates user-side energy storage Charging power at any time express Side energy storage Discharge power at any given time , and These represent energy storage on the power supply side, grid side, and user side, respectively. Indicates energy storage on the power supply side Discharge power at any given time Indicates grid-side energy storage Discharge power at any given time Indicates user-side energy storage Discharge power at any given moment.

[0041] S42. Using equation (5), establish shared energy storage charging and discharging operation constraints for the power source, power grid, and user side:

[0042] (5);

[0043] in, This is for energy storage charging and discharging identification. Indicates charging. Indicates discharge; express Side energy storage The charging and discharging state at any given moment; for Maximum charging power of side energy storage; for Maximum discharge power of side-stored energy.

[0044] S43. Use equation (6) to establish constraints on the energy storage capacity of the power source, power grid, and user side:

[0045] (6);

[0046] in, for Time period The capacity of side energy storage for Time period The capacity of side energy storage for The charging efficiency of side-side energy storage for The discharge efficiency of side-stored energy for Minimum capacity limit for side-side energy storage for Maximum capacity limit for side-side energy storage This indicates the total number of time periods in a time cycle.

[0047] S44. Establish system power balance constraints using equation (7):

[0048] (7).

[0049] Furthermore, step S5 specifically includes the following steps:

[0050] S51. Based on a set of three-sided energy storage charging and discharging power sequences, the Zth particle is obtained using equation (8). The position of the control strategy is determined by equation (9), which is the current value of the control strategy. Particles The speed, i.e., the amount of change that the control strategy is about to produce. :

[0051] (8);

[0052] (9);

[0053] in, , The particle number is used to indicate the particle's number. Represents the total number of particles. The initial values ​​of each particle are randomly generated according to the rules of formulas (10) and (11); subscript Desirable , and These represent energy storage on the power supply side, grid side, and user side, respectively. This is for energy storage charging and discharging identification. Indicates energy storage charging. It indicates energy storage discharge; the superscript ESS indicates energy storage, meaning that the control strategy controls energy storage. Indicates time, ; This indicates the amount of change in each component; The data in the t-th row and i-th column This represents the charging / discharging power of the energy stored on side i at time t.

[0054] S52. Use equation (10) to generate the initial position of each particle, and use equation (11) to generate the initial velocity of each particle:

[0055] (10);

[0056] (11);

[0057] in, express Particle velocity at the next iteration Indicates inertia weight, express Particle velocity at the next iteration For particles The individual optimal solution. express The particle's spatial position at the next iteration. This is the globally optimal solution. express The particle's spatial position at the next iteration, subscript , Both represent the number of iterations. and Represents the learning factor. and A random number between 0 and 1.

[0058] S53. Calculate the particle fitness function value using equation (12) as a criterion for judging whether a particle is good or bad, and penalize particles that do not meet the constraints:

[0059] (12);

[0060] in, For particles The fitness function value, This is a simplified version of equation (4); p is a fixed penalty term and is a constant.

[0061] S54. Calculate the position, velocity, and fitness function of each particle's next generation, perform particle iteration, and update the individual optimal solution and the global optimal solution based on the comparison results of the fitness function value after iteration with the individual optimal solution and the global optimal solution.

[0062] S55. Repeat step S54 until the maximum number of iterations is reached. Stop iteration and output the global optimal solution. As the optimal set of control strategies, the shared energy storage charging and discharging operation scheduling strategy for the source, grid, and user sides is obtained.

[0063] Furthermore, step S54 specifically includes the following steps:

[0064] S541, For m particles Calculate according to formula (11) respectively The position of the particle at that time is calculated according to equation (10). The velocity of the particle at that time is calculated according to equation (12). The fitness function of each particle is used for particle iteration;

[0065] S542, For m particles If the fitness function value after iteration Greater than the individual optimal solution of this particle fitness function value Then the current Updated to the individual optimal solution of the particle. If the fitness function value of an individual optimal solution with particles among all particles is greater than that of the global optimal solution. fitness function value Then the particle's Updated to the global optimal solution for the population. ; This represents the z-th particle, i.e., the z-th control strategy.

[0066] Beneficial effects

[0067] Existing shared energy storage technologies typically employ static capacity allocation strategies, and their scheduling models do not fully consider the differentiated needs of the power supply side, grid side, and user side, resulting in low energy storage utilization and limited effectiveness in mitigating fluctuations. This invention, through a dynamic capacity allocation mechanism, proposes for the first time a demand factor calculation model based on capacity demand benchmarks. This model can accurately quantify the differences in demand across the three sides, enabling dynamic allocation of energy storage capacity on demand. This not only significantly improves energy storage utilization but also ensures that charging and discharging strategies align with actual needs by embedding the allocation results as hard constraints into the scheduling model. It collaboratively optimizes the charging and discharging behaviors of the three sides, thereby effectively reducing system power deficits and mitigating new energy fluctuations and load deviations. Furthermore, this invention combines a typical daily curve adaptive generation method with a high-dimensional particle swarm optimization strategy, improving the representativeness of input data and reducing computational complexity. The scheduling model constructed in this invention possesses a complete constraint system and can be directly embedded into the grid dispatch system. Its parameter adaptive characteristics can flexibly adapt to energy storage power stations of different scales, enhancing engineering practicality while ensuring optimization accuracy. Attached Figure Description

[0068] Figure 1 This is a flowchart of the optimal operation and scheduling method for shared energy storage on the source-grid-user side, which takes into account the allocation of energy storage capacity in this invention. Detailed Implementation

[0069] To provide a better understanding of the structural features and effects achieved by the present invention, a detailed description is provided below, accompanied by preferred embodiments and accompanying drawings:

[0070] This invention relates to an optimal operation and scheduling method for shared energy storage on the grid-source and user sides, considering energy storage capacity allocation. The method includes: acquiring historical power generation and consumption data and shared energy storage power station configuration data from the grid and user sides to generate typical daily renewable energy output and load curves; proposing an energy storage capacity allocation mechanism on the grid-source and user sides; calculating the demand factor for energy storage capacity on each side; and quantitatively evaluating and allocating energy storage capacity. Based on the capacity allocation mechanism, an optimal operation and scheduling model for shared energy storage on the grid-source and user sides is established. Operational constraints for charging and discharging of shared energy storage and energy storage capacity constraints are constructed. The model is solved using a particle swarm optimization algorithm to find the optimal operation and scheduling strategy, providing a basis and reference for grid operation.

[0071] The invention will now be described in detail with reference to the accompanying drawings. Figure 1As shown, the optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, includes the following steps:

[0072] S1. Obtain historical power generation and consumption data from the source-grid-user side, as well as configuration data of shared energy storage power stations.

[0073] The historical power generation and consumption data on the source-grid-user side include: historical output data of new energy power plants, historical load data of the power grid, and historical electricity consumption data on the user side; the configuration data of the shared energy storage power station includes the technical parameters of the shared energy storage, which include charging and discharging efficiency, rated capacity, rated power, etc.

[0074] Step S1 aims to collect basic data to provide data support for subsequent typical daily curve generation, capacity allocation, and scheduling model construction. The historical power generation and consumption data from the source-grid-user side includes data from the power source side, grid side, and user side. The power source side mainly refers to the historical output data of new energy power plants, including records of the power generation changes of renewable energy sources such as solar and wind power over time, used to analyze the intermittent and fluctuating characteristics of new energy output. The grid side focuses on historical grid load data, reflecting the overall electricity demand and power flow distribution of the grid at different times, and is the basis for assessing the grid's demand for energy storage capacity. The user side covers historical electricity consumption data, reflecting the electricity consumption patterns of users (such as industrial users, commercial users, and residential users) (such as peak and valley periods, load fluctuation amplitude, etc.), providing a basis for user-side energy storage demand analysis. The configuration data of the shared energy storage power station mainly consists of the technical parameters of the shared energy storage system, including: charge and discharge efficiency, rated capacity, and rated power. Charge and discharge efficiency refers to the energy conversion efficiency during the charging and discharging process of the energy storage system, affecting the actual available energy storage capacity and the design of charging and discharging strategies. Rated capacity refers to the maximum energy storage capacity of an energy storage power station, serving as the overall benchmark for capacity allocation. Rated power refers to the maximum charging and discharging power of the energy storage system per unit time, limiting the charging and discharging rate and being a key parameter constraining operation. The accuracy and completeness of these data directly affect the representativeness of subsequent typical daily curves, the rationality of capacity allocation, and the optimization effect of the scheduling model, and are prerequisites for the implementation of the entire method.

[0075] S2. Based on historical power generation and consumption data from the source, grid, and user sides, generate typical daily curves for renewable energy output, user-side load, and grid-side load.

[0076] Step S2 specifically includes the following steps:

[0077] S21. Determine the optimal number of typical days using a clustering algorithm; the core of step S21 is to determine the optimal number of typical days using a clustering algorithm, which will contribute to the subsequent generation of representative typical days for renewable energy. User-side load and grid-side load The curve lays the foundation.

[0078] Step S21 specifically includes the following steps:

[0079] S211, Preset upper limit of the number of typical days Different values ​​were selected sequentially as the candidate number of typical days. Determine the number of typical days The range of values ;

[0080] S212, For each candidate The value is calculated using equation (1). Cost function corresponding to the value :

[0081] (1);

[0082] in, The k-th cluster center is randomly selected, where k represents the cluster center number. Indicates the first Output curve, , Indicates load or new energy source. Indicates the source network user side, Therefore Clusters centered on;

[0083] S213, Obtain the cost function The point where the rate of descent suddenly decreases, and the point corresponding to this sudden decrease. The value is used as the optimal number of typical days.

[0084] along with The value increases, The number of clusters shows a downward trend, and the point where the rate of decrease suddenly drops can represent the true number of clusters. Cost function The physical meaning of is the sum of squared distances from all curves to the center of their respective clusters, reflecting the overall error of the clustering results. This varies with the number of typical days. The increase in cluster size is due to the fact that more clusters can fit the data features more precisely. It shows a downward trend, but the rate of decline will gradually slow down. When When the rate of descent decreases sharply, the corresponding The value is considered the true cluster number, i.e., the optimal number of typical days. This ensures the representativeness of the typical day curves while avoiding data redundancy caused by an excessive number of typical days. Step S21 quantitatively analyzes the relationship between clustering error and the number of typical days to ensure that the subsequently generated typical day curves accurately reflect the actual changes in the output and load of new energy sources on the source, grid, and user sides, providing a reliable input basis for energy storage capacity allocation and scheduling models.

[0085] S22. Based on the optimal number of typical days, generate the typical day renewable energy output curve, the typical day load curve on the user side, and the typical day load curve on the grid side.

[0086] Step S22 generates typical daily renewable energy output curves, typical daily load curves on the user side, and typical daily load curves on the grid side through clustering iteration. Its core is clustering optimization based on Euclidean distance. This process extracts the core features of historical data through clustering algorithms, ensuring that the generated typical daily curves not only reflect the actual power generation and consumption patterns but also simplify the data dimensions, providing an efficient and reliable input basis for subsequent optimized scheduling.

[0087] The methods for generating typical daily load curves on the user side and the grid side are completely consistent with those for generating typical daily renewable energy output curves. The only difference is that the curve used on the power generation side is the output curve, while the curves used on the grid side and the user side are load curves. The curves on the power generation, grid, and user sides are generated separately, but the generation methods are exactly the same.

[0088] The following section uses the power supply side as an example to introduce the method for generating curves.

[0089] In step S22, generating a typical daily renewable energy output curve specifically includes the following steps:

[0090] S221. Based on the historical power output curve of the power supply side, select... Cluster centers .

[0091] The Cluster centers The typical power output curves are selected from the historical power output curves on the power source side, which are the 24-hour power output curves of new energy sources. When generating typical power output curves on the power source side, K cluster centers are selected based on all historical power output curves of new energy sources on the power source side (one curve per day). When generating typical load curves on the grid side, K cluster centers are selected based on all historical load curves of the grid side. When generating typical load curves on the user side, K cluster centers are selected based on all historical load curves of the user side.

[0092] S222. For each historical output curve, calculate its correlation with... Cluster centers The Euclidean distance is calculated, and the curve is assigned to the nearest cluster. For the , New energy power output curve ,calculate arrive The Euclidean distance, and Assign it to the nearest cluster.

[0093] S223. Calculate the mean of the historical power output curves of all new energy sources on the power supply side within each cluster, and use this mean as the new cluster center. .

[0094] S224. Repeat steps S221-S223 until cluster centers are reached. When the change is less than a preset threshold or the maximum number of iterations is reached, the iteration stops and outputs the result. Cluster centers This forms a typical daily new energy output curve. After the iteration stops in step S224, the final result is... Each cluster center represents a typical daily new energy output curve.

[0095] For both the grid side and the user side, the output curves in steps S222-S224 are replaced with load curves. The typical daily renewable energy output curve, the typical daily user-side load curve, and the typical daily grid-side load curve represent the renewable energy output, user-side load, and grid-side load, respectively. The typical changes within a typical day are the key foundational data for subsequent steps S3 and S4.

[0096] S3. Establish a source-grid-user side energy storage capacity allocation mechanism and calculate the energy storage capacity allocated to each side.

[0097] The purpose of step S3 is to establish a dynamic energy storage capacity allocation mechanism, which quantitatively calculates the energy storage capacity to be allocated to each side based on the differentiated needs of the source, grid and user sides, and provides a capacity constraint basis for the subsequent operation and scheduling model.

[0098] S31. Based on the historical power generation and consumption data of the source-grid-household side, determine the capacity demand benchmark for each side.

[0099] The capacity requirement benchmarks for each side include the capacity requirement benchmarks for the power supply side. Grid-side capacity demand benchmark and user-side capacity demand benchmark The capacity requirement baseline on the power supply side. The maximum scheduling deviation based on historical statistical data; the capacity demand benchmark on the grid side. The load fluctuation deviation on the grid side; the capacity demand benchmark on the user side. This refers to the peak-valley difference in user-side load.

[0100] S32. Based on the capacity demand benchmarks of each side of the source, grid, and user, calculate the demand factor of each side for energy storage capacity using equation (2). :

[0101] (2);

[0102] in, This is a demand factor for energy storage capacity on each side, used to quantitatively assess the allocation of energy storage capacity; For source network user-side identification; This serves as the baseline for capacity requirements on each side.

[0103] The demand factor is a key indicator for measuring the intensity of demand for shared energy storage capacity from each side (source, grid, and user). Its value directly determines the subsequent allocation ratio of energy storage capacity and serves as a bridge connecting the capacity demand benchmark and the actual capacity allocation. The capacity demand benchmark is the benchmark value determined in step S31 for each side: maximum dispatch deviation for the source side, load fluctuation deviation for the grid side, and load peak-valley difference for the user side. It is used to reflect the basic demand characteristics of each side for energy storage. Demand fluctuation refers to the dynamic change in demand shown by each side in historical data, used to supplement short-term fluctuations in demand not covered by the benchmark value. The weighting coefficients of the capacity demand benchmark and the demand fluctuation are used to adjust the influence weights of the capacity demand benchmark and demand fluctuation in the demand factor, respectively, and can be flexibly set according to actual scenarios (such as grid operation priority, user-side demand intensity, etc.).

[0104] S33. Calculate the energy storage capacity allocated to each side of the source grid and user using formula (3). :

[0105] (3);

[0106] in, This indicates the rated capacity of the shared energy storage.

[0107] S4. Based on the energy storage capacity allocation mechanism of the source-grid-user side, establish an optimal operation and scheduling model for shared energy storage on the source-grid-user side.

[0108] Step S4, based on the energy storage capacity allocation mechanism determined in Step S3, constructs an optimal operation scheduling model for shared energy storage on the source, grid, and user sides. The core of this step is to achieve coordinated optimization of the charging and discharging behavior of energy storage on all three sides by setting optimization objectives and constraints. This step constructs a scheduling model that balances economy and safety by clarifying the optimization objective (minimizing power deficit) and four types of constraints (charging / discharging, capacity, and power balance), laying the framework for solving the optimal charging and discharging strategy in the subsequent step S5.

[0109] Furthermore, the optimization objective of the source-grid-user side shared energy storage optimal operation scheduling model is to minimize the system power deficit, and the decision variables are the hourly charging and discharging power of the power source side energy storage, the hourly charging and discharging power of the grid side energy storage, and the hourly charging and discharging power of the user side energy storage.

[0110] Step S4 specifically includes the following steps:

[0111] S41. Calculate the minimum system power deficit using equation (4):

[0112] (4);

[0113] in, for The power deficit of the source-grid-user shared energy storage system at any given time. Let t be the load fluctuation power of the shared energy storage system between the source, grid, and users. The fluctuating power of new energy generation during time period t. express Side energy storage Charging power at any time Indicates energy storage on the power supply side Charging power at any time Indicates grid-side energy storage Charging power at any time Indicates user-side energy storage Charging power at any time express Side energy storage Discharge power at any given time , and These represent energy storage on the power supply side, grid side, and user side, respectively. Indicates energy storage on the power supply side Discharge power at any given time Indicates grid-side energy storage Discharge power at any given time Indicates user-side energy storage Discharge power at any given moment.

[0114] Formula (4) measures the degree of system power balance by quantifying the difference between the above factors, thereby minimizing the system power deficit. By adjusting the charging and discharging power of energy storage on the source, grid, and user sides, the fluctuations in new energy power generation and load fluctuations are balanced, reducing the imbalance between system power supply and demand.

[0115] S42. Using equation (5), establish shared energy storage charging and discharging operation constraints for the power source, power grid, and user side:

[0116] (5);

[0117] in, This is for energy storage charging and discharging identification. Indicates charging. Indicates discharge; express Side energy storage The charging and discharging state at any given moment; for Maximum charging power of side energy storage; for Maximum discharge power of side-stored energy.

[0118] Step S42 is used to limit the charging and discharging power of energy storage on each side within a safe range to avoid overload or abnormal operation, thereby ensuring that the energy storage system operates at its rated capacity and protecting equipment safety.

[0119] S43. Use equation (6) to establish constraints on the energy storage capacity of the power source, power grid, and user side:

[0120] (6);

[0121] in, for Time period The capacity of side energy storage for Time period The capacity of side energy storage for The charging efficiency of side-side energy storage for The discharge efficiency of side-stored energy for Minimum capacity limit for side-side energy storage for Maximum capacity limit for side-side energy storage This indicates the total number of time periods in a time cycle.

[0122] Step S43 constrains the capacity between the minimum capacity limit and the maximum capacity limit. The maximum capacity limit is the energy storage capacity allocated on each side in step S33. Step S43 can ensure that the energy storage capacity on each side is within a reasonable range, avoiding overcharging (damaging the equipment) or over-discharging (affecting the lifespan).

[0123] S44. Establish system power balance constraints using equation (7):

[0124] (7).

[0125] The purpose of the system power balance constraint is to ensure that the total power generation on the power source, grid, and user sides is balanced with the total load demand, which is a core condition for the stable operation of the power grid. By comprehensively considering the power generation on the power source side, the transmission power on the grid side, the power consumption on the user side, and the charging and discharging power of energy storage, the algebraic sum of each power component is constrained to be 0, thereby ensuring real-time power supply and demand balance.

[0126] S5. Use the particle swarm optimization algorithm to solve the optimal operation scheduling model of shared energy storage on the source-grid-user side, and determine the optimal charging and discharging operation scheduling strategy of shared energy storage on the source-grid-user side that optimizes the objective.

[0127] S51. Establish the definition of a particle. This represents the z-th particle, which is also the z-th control strategy. It sets the population size. Maximum number of iterations The particle encoding dimension is 3*24, where 3 represents the source, network, and user sides. This represents a set of three-sided energy storage control strategies. The first column of data represents the control strategy on the power source side, the second column represents the control strategy on the grid side, and the third column represents the control strategy on the user side. 24 represents 24 hours, and one column of data represents one day.

[0128] Based on a set of three-sided energy storage charge and discharge power sequences, the Zth particle is obtained using equation (8). The position of the control strategy is determined by equation (9), which is the current value of the control strategy. Particles The speed, i.e., the amount of change that the control strategy is about to produce. :

[0129] (8);

[0130] (9);

[0131] in, , The particle number is used to indicate the particle's number. Represents the total number of particles. The initial values ​​of each particle are randomly generated according to the rules of formulas (10) and (11); subscript Desirable , and These represent energy storage on the power supply side, grid side, and user side, respectively. This is for energy storage charging and discharging identification. Indicates energy storage charging. It indicates energy storage discharge; the superscript ESS indicates energy storage, meaning that the control strategy controls energy storage. Indicates time, ; This indicates the amount of change in each component; for example: The data in row t and column i This represents the charging / discharging power of the energy stored on side i at time t.

[0132] S52. Use equation (10) to generate the initial position of each particle, and use equation (11) to generate the initial velocity of each particle:

[0133] Set inertia weight and learning factors and Once set, it becomes a constant. The larger it is, the harder it is to change its speed; The larger the value, the stronger the particle's ability to learn from the individual optimal solution. The larger the value, the stronger the particle's ability to learn from the swarm's optimal solution. Each particle in the set generates its initial position according to equation (10) and its initial velocity according to equation (11). Since this is the first iteration, ... .

[0134] (10);

[0135] (11);

[0136] in, express Particle velocity at the next iteration Indicates inertia weight, express Particle velocity at the next iteration For particles The individual optimal solution. express The particle's spatial position at the next iteration. This is the globally optimal solution. express The particle's spatial position at the next iteration, subscript , Both represent the number of iterations. and Represents the learning factor. and A random number between 0 and 1 is generated for each iteration to prevent the algorithm from getting trapped in local optima. Global optimal solution. and the individual optimal solution of each particle The initial value is taken as a random value. There is only one global optimal solution, and each particle has one individual optimal solution.

[0137] S53. Calculate the particle fitness function value using equation (12) as a criterion for judging whether a particle is good or bad, and penalize particles that do not meet the constraints:

[0138] (12);

[0139] in, For particles The fitness function value, This is a simplified version of equation (4); p is a fixed penalty term and is a constant.

[0140] S54. Calculate the position, velocity, and fitness function of each particle's next generation, perform particle iteration, and update the individual optimal solution and the global optimal solution based on the comparison results of the fitness function value after iteration with the individual optimal solution and the global optimal solution.

[0141] Step S54 specifically includes the following steps:

[0142] S541, For m particles Calculate according to formula (11) respectively The position of the particle at that time is calculated according to equation (10). The velocity of the particle at that time is calculated according to equation (12). The fitness function of each particle is used for particle iteration;

[0143] S542, For m particles If the fitness function value after iteration Greater than the individual optimal solution of this particle fitness function value Then the current Updated to the individual optimal solution of the particle. If the fitness function value of an individual optimal solution with particles among all particles is greater than that of the global optimal solution. fitness function value Then the particle's Updated to the global optimal solution for the population. ; This represents the z-th particle, which is also the z-th control strategy.

[0144] S55. Repeat step S54 until the maximum number of iterations is reached. Stop iteration and output the global optimal solution. As the optimal set of control strategies, the shared energy storage charging and discharging operation scheduling strategy for the source, grid, and user sides is obtained.

[0145] The innovative point of this invention is:

[0146] (1) Existing technologies lack research on energy storage sharing between the source, grid, and user sides, resulting in limited application scenarios and insufficient system scheduling flexibility. This invention employs a typical daily curve generation method, specifically by improving the clustering algorithm to generate typical curves of new energy output and load. This improved clustering algorithm makes the input data more representative, solving the problem of insufficient representativeness in the input data of existing technologies. This invention accurately reflects the actual changing patterns of new energy output and load on the source, grid, and user sides, providing a reliable input basis for energy storage capacity allocation and scheduling models, and improving the accuracy and reliability of subsequent optimized scheduling. Furthermore, existing technologies do not fully consider the differentiated needs and data characteristics of the source, grid, and user sides when generating typical daily curves. This invention, through improved clustering algorithms, can better adapt to the complex scenarios of shared energy storage between the source, grid, and user sides, improving the generation quality and efficiency of typical daily curves.

[0147] (2) Existing shared energy storage technologies typically employ static capacity allocation strategies, and their scheduling models do not adequately consider the differentiated needs of the power supply side, grid side, and user side, resulting in low energy storage utilization and limited effectiveness in mitigating fluctuations. This invention, however, adopts an energy storage capacity allocation mechanism based on a demand factor calculation model using differentiated capacity demand benchmarks to achieve quantitative allocation of shared energy storage capacity. Based on the capacity demand benchmarks, demand fluctuations, weighting coefficients of the capacity demand benchmarks, and weighting coefficients of the demand fluctuations for each side, the demand factor for energy storage capacity is calculated, and the quantitative allocation of shared energy storage capacity is achieved using this demand factor. Through this method, this invention can accurately quantify the demand differences among each side, achieve dynamic allocation of energy storage capacity on demand, significantly improve energy storage utilization, ensure that the charging and discharging strategy aligns with actual needs, and collaboratively optimize the charging and discharging behavior of the three sides, thereby effectively reducing system power deficits and mitigating new energy fluctuations and load deviations.

[0148] (3) Existing optimal energy storage operation and scheduling models lack research on the direction of shared energy storage between the source, grid, and users, making it difficult to meet the collaborative and win-win needs of the three parties. This invention adopts a shared energy storage scheduling model with coupled capacity constraints, establishing an optimization model with the goal of minimizing the system power deficit, using the capacity allocation result as the constraint condition, and the decision variable as the 24-hour charging and discharging power sequence of the three sides. This model can comprehensively consider the differentiated needs and constraints of the source, grid, and user sides, realize the collaborative optimization of the charging and discharging behavior of energy storage on the three sides, improve the overall performance and operating efficiency of the system, and provide a more scientific basis and reference for grid operation. In addition, existing scheduling models usually only focus on the optimization of one aspect, while this invention couples the shared energy storage scheduling of the source, grid, and user sides, fully considering the capacity constraints of each side, making the scheduling model more complete and scientific, and better able to adapt to complex power system scenarios.

[0149] (4) Existing technologies suffer from drawbacks in solving shared energy storage scheduling models, such as high difficulty in solving high-dimensional optimization problems and high computational complexity, making it difficult to efficiently solve complex shared energy storage scheduling models. This invention employs a high-dimensional solution strategy using particle swarm optimization (PSO), employing 3×24-dimensional particle encoding to represent the charging and discharging power on all three sides, and combining a fitness function with a penalty term to handle constraints. This high-dimensional PSO strategy improves solution efficiency and accuracy, enabling the rapid identification of the optimal source-grid-user-side shared energy storage charging and discharging operation scheduling strategy, reducing computational complexity, and enhancing the model's engineering applicability. This invention combines a typical daily curve adaptive generation method with the PSO strategy, making the solution process more intelligent and efficient, better adapting to the scheduling needs of large-scale energy storage power stations, and improving the model's applicability and flexibility.

[0150] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for optimal operation and scheduling of shared energy storage on the source-grid-user side, considering energy storage capacity allocation, characterized in that, The method includes the following steps: S1. Obtain historical power generation and consumption data from the source-grid-user side, as well as configuration data of shared energy storage power stations; S2. Based on historical power generation and consumption data from the source-grid-user side, generate typical daily curves for renewable energy output, user-side load, and grid-side load. S3. Establish a source-grid-user side energy storage capacity allocation mechanism and calculate the energy storage capacity allocated to each side; S4. Based on the energy storage capacity allocation mechanism of the source-grid-user side, establish an optimal operation and scheduling model for shared energy storage on the source-grid-user side; S5. Use the particle swarm optimization algorithm to solve the optimal operation scheduling model of shared energy storage on the source-grid-user side, and determine the optimal charging and discharging operation scheduling strategy of shared energy storage on the source-grid-user side that optimizes the objective.

2. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... The historical power generation and consumption data on the source-grid-user side include: historical power output data of new energy power plants, historical load data of the power grid, and historical electricity consumption data on the user side; the configuration data of the shared energy storage power station includes the technical parameters of the shared energy storage, which include charge and discharge efficiency, rated capacity, and rated power.

3. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... Step S2 specifically includes the following steps: S21. Use clustering algorithms to determine the optimal number of typical days; S22. Based on the optimal number of typical days, generate the typical day renewable energy output curve, the typical day load curve on the user side, and the typical day load curve on the grid side.

4. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 3, is characterized in that... Step S21 specifically includes the following steps: S211, Preset upper limit of the number of typical days Different values ​​were selected sequentially as the candidate number of typical days. Determine the number of typical days The range of values ​​for: ; S212, For each candidate The value is calculated using equation (1). Cost function corresponding to the value : (1); in, The k-th cluster center is randomly selected, where k represents the cluster center number. Indicates the first Output curve, , Indicates load or new energy source. Indicates the source network user side, Therefore Clusters centered on; S213, Obtain the cost function The point where the rate of descent suddenly decreases, and the point corresponding to this sudden decrease. The value is used as the optimal number of typical days.

5. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 3, is characterized in that... In step S22, generating a typical daily renewable energy output curve specifically includes the following steps: S221. Based on the historical power output curve of the power supply side, select... Cluster centers ; S222. For each historical output curve, calculate its correlation with... Cluster centers The Euclidean distance is calculated, and the curve is assigned to the nearest cluster. S223. Calculate the mean of the historical power output curves of all new energy sources on the power supply side within each cluster, and use this mean as the new cluster center. ; S224. Repeat steps S221-S223 until cluster centers are reached. When the change is less than a preset threshold or the maximum number of iterations is reached, the iteration stops and outputs the result. Cluster centers This forms a typical daily new energy output curve.

6. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... Step S3 specifically includes the following steps: S31. Based on the historical power generation and consumption data of the source-grid-household side, determine the capacity demand benchmark for each side of the source, grid and household. S32. Based on the capacity demand benchmarks of each side of the source, grid, and user, calculate the demand factor of each side for energy storage capacity using equation (2). : (2); in, This is a demand factor for energy storage capacity on each side, used to quantitatively assess the allocation of energy storage capacity; For source network user-side identification; This serves as a baseline for capacity requirements on each side. S33. Calculate the energy storage capacity allocated to each side of the source grid and user using formula (3). : (3); in, This indicates the rated capacity of the shared energy storage.

7. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... The optimization objective of the source-grid-user-side shared energy storage optimal operation and scheduling model is to minimize the system power deficit. The decision variables are the hourly charging and discharging power of the power source-side energy storage, the hourly charging and discharging power of the grid-side energy storage, and the hourly charging and discharging power of the user-side energy storage.

8. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... Step S4 specifically includes the following steps: S41. Calculate the minimum system power deficit using equation (4): (4); in, for The power deficit of the source-grid-user shared energy storage system at any given time. Let t be the load fluctuation power of the shared energy storage system between the source, grid, and users. The fluctuating power of new energy generation during time period t. express Side energy storage Charging power at any time Indicates energy storage on the power supply side Charging power at any time Indicates grid-side energy storage Charging power at any time Indicates user-side energy storage Charging power at any time express Side energy storage Discharge power at any given time , and These represent energy storage on the power supply side, grid side, and user side, respectively. Indicates energy storage on the power supply side Discharge power at any given time Indicates grid-side energy storage Discharge power at any given time Indicates user-side energy storage Discharge power at any given moment; S42. Using equation (5), establish shared energy storage charging and discharging operation constraints for the power source, power grid, and user side: (5); in, This is for energy storage charging and discharging identification. Indicates charging. Indicates discharge; express Side energy storage The charging and discharging state at any given moment; for Maximum charging power of side energy storage; for Maximum discharge power of side-stored energy; S43. Use equation (6) to establish constraints on the energy storage capacity of the power source, power grid, and user side: (6); in, for Time period The capacity of side energy storage for Time period The capacity of side energy storage for The charging efficiency of side-side energy storage for The discharge efficiency of side-stored energy for Minimum capacity limit for side-side energy storage for Maximum capacity limit for side energy storage Indicates the total number of time periods in a time cycle; S44. Establish system power balance constraints using equation (7): (7)。 9. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 1, is characterized in that... Step S5 specifically includes the following steps: S51. Based on a set of three-sided energy storage charging and discharging power sequences, the Zth particle is obtained using equation (8). The position of the control strategy is determined by equation (9), which is the current value of the control strategy. Particles The speed, i.e., the amount of change that the control strategy is about to produce. : (8); (9); in, , The particle number is used to indicate the particle's number. Represents the total number of particles. The initial values ​​of each particle are randomly generated according to the rules of formulas (10) and (11); subscript Pick , and These represent energy storage on the power supply side, grid side, and user side, respectively. This is for energy storage charging and discharging identification. Indicates energy storage charging. It indicates energy storage discharge; the superscript ESS indicates energy storage, meaning that the control strategy controls energy storage. Indicates time, ; This indicates the amount of change in each component; The data in the t-th row and i-th column This represents the charging / discharging power of the energy stored on side i at time t; S52. Use equation (10) to generate the initial position of each particle, and use equation (11) to generate the initial velocity of each particle: (10); (11); in, express Particle velocity at the next iteration Indicates inertia weight, express Particle velocity at the next iteration For particles The individual optimal solution. express The particle's spatial position at the next iteration. This is the globally optimal solution. express The particle's spatial position at the next iteration, subscript , Both represent the number of iterations. and Represents the learning factor. and A random number between 0 and 1; S53. Calculate the particle fitness function value using equation (12) as a criterion for judging whether a particle is good or bad, and penalize particles that do not meet the constraints: (12); in, For particles The fitness function value, This is a simplified version of equation (4); p is a fixed penalty term and is a constant; S54. Calculate the position, velocity, and fitness function of each particle's next generation, perform particle iteration, and update the individual optimal solution and the global optimal solution based on the comparison results of the fitness function value after iteration with the individual optimal solution and the global optimal solution. S55. Repeat step S54 until the maximum number of iterations is reached. Stop iteration and output the global optimal solution. As the optimal set of control strategies, the shared energy storage charging and discharging operation scheduling strategy for the source, grid, and user sides is obtained.

10. The optimal operation and scheduling method for shared energy storage on the source-grid-user side, considering energy storage capacity allocation, as described in claim 9, is characterized in that... Step S54 specifically includes the following steps: S541, For m particles Calculate according to formula (11) respectively The position of the particle at that time is calculated according to equation (10). The velocity of the particle at that time is calculated according to equation (12). The fitness function of each particle is used for particle iteration; S542, For m particles If the fitness function value after iteration Greater than the individual optimal solution of this particle fitness function value Then the current Updated to the individual optimal solution of the particle. If the fitness function value of an individual optimal solution with particles among all particles is greater than that of the global optimal solution. fitness function value Then the particle's Updated to the global optimal solution for the population. ; This represents the z-th particle, i.e., the z-th control strategy.

Citation Information

Cited By

  • Data center centralized standby power energy storage optimal configuration method based on cluster sharing architecture

    CN121529946A