Power distribution area shared energy storage optimization scheduling method based on cooperation of multiple elastic resources

Abstract

The invention discloses a power distribution area shared energy storage optimization scheduling method based on cooperation of multiple elastic resources, and the method comprises the steps: bringing four types of core resources, namely distributed photovoltaic resources, electric vehicle charging station loads, temperature control loads and shared energy storage batteries, into a unified scheduling system, building a system decision model, setting a multi-dimensional constraint condition, and carrying out the optimization scheduling of the shared energy storage of the power distribution area. A'source-storage-load 'coordinated scheduling closed loop is formed, photovoltaic output fluctuation is effectively stabilized, 'unit carbon emission cost' is introduced, carbon emission is indirectly converted into carbon cost to be incorporated into a comprehensive cost objective function, fossil energy complementing carbon emission and energy storage charging and discharging indirect carbon emission caused by insufficient photovoltaic consumption in a transformer area are converted into cost items, and the cost items are optimized. A hierarchical iterative optimization mechanism is constructed, a'user side-transformer area side 'hierarchical optimization framework is innovatively adopted, and a'peak clipping cost-peak load' Pareto optimal curve is generated by combining a user scheme to dynamically adjust energy storage price and regulate and control air conditioning load, so that benefit balance of users and a power grid is realized. A full-process mechanism of real-time data acquisition, cooperative computing platform analysis, abnormity reporting and regulation and control instruction issuing is established, the problem that traditional static scheduling cannot deal with sudden working conditions is effectively avoided through real-time response, it is ensured that the system can still operate stably in photovoltaic output fluctuation and load sudden change scenes, and the power supply reliability and the resource utilization efficiency of a transformer area are improved.

Description

Technical Field

[0001] This invention relates to the field of information technology, specifically to a method for optimized scheduling of shared energy storage in distribution substations that coordinates multiple flexible resources. Background Technology

[0002] As renewable energy sources, represented by photovoltaics, gradually cover distribution areas, in scenarios where distributed photovoltaics and shared energy storage are optimized in a coordinated manner, the characteristics of photovoltaics—small in size, numerous in number, widely distributed, and irregular—make them prone to problems such as severe power flow back and node voltage exceeding limits after being connected to the distribution network. At the same time, the strong fluctuations in photovoltaic output make it difficult for the grid to achieve real-time supply and demand balance, which places higher demands on the flexibility and resource coordination of dispatching strategies.

[0003] How to accurately construct and dynamically adjust the objective functions on the user side and the distribution area side under multiple objective constraints to achieve the optimal solution that balances photovoltaic (PV) consumption, carbon emission control, and economic costs? Distributed PV output within a distribution area exhibits significant diurnal fluctuations, with output almost zero at night. Meanwhile, load demand often aligns with peak PV output periods, leading to incomplete PV power consumption. Furthermore, the charging and discharging strategies for shared energy storage need to balance user electricity purchase costs and peak load reduction requirements while maintaining a safe battery state of charge range. Additionally, the highly dynamic temperature regulation thresholds for temperature-controlled loads and the time-of-day distribution characteristics of electric vehicle charging stations increase the complexity of control period segmentation and command generation. Implicit carbon emission constraints require limiting the proportion of fossil fuel supplementation during peak hours, but existing data acquisition and collaborative computing platforms struggle to quickly identify abnormal data and accurately assess low-carbon performance and optimization when processing real-time data. This can result in generated optimization commands deviating from actual needs, impacting PV consumption rate and system stability. More importantly, the initial parameters of the shared energy storage charging and discharging benchmark price and the time-of-use electricity price range of the distribution area lack an adaptive adjustment mechanism, making it difficult to cope with the dynamic balance requirements under the prediction deviation of the photovoltaic power output increment curve and the scenario of sudden load changes.

[0004] Current distribution network optimization dispatch strategies primarily focus on peak-shaving control of adjustable loads such as electric vehicle loads and temperature-controlled loads (e.g., air conditioning). Their core objectives are to reduce user economic costs, smooth peak-valley load differences, or improve grid stability. However, these strategies generally suffer from two major technical limitations: First, existing strategies often fail to adequately consider energy storage battery resources connected to the distribution network. They fail to mitigate photovoltaic volatility from both the source and storage sides through multi-resource synergy involving distributed photovoltaics, adjustable loads, and energy storage batteries, resulting in limited dispatch effectiveness. Second, under the policy background of "carbon peaking and carbon neutrality," existing strategies do not incorporate carbon emissions into the optimization system. They fail to quantify "carbon emission costs" (e.g., cost accounting based on carbon taxes and carbon quota trading) and integrate them with electricity purchase costs and energy storage usage costs as optimization targets. Furthermore, they do not design suitable solutions for scenarios requiring carbon emission control, such as low-carbon demonstration parks and zero-carbon communities. This results in dispatch results that fail to meet regional carbon emission control needs and are out of sync with current green energy development trends. This problem involves a deep coupling of multi-dimensional data fusion, real-time anomaly identification, and multi-objective iterative optimization in terms of technical implementation. It is urgent to break through the limitations of existing models and algorithms in order to adapt to the refined management requirements of the complex operating environment of the transformer substation.

[0005] To address the above issues, Chinese Patent Publication No. CN113629705B discloses a user-side energy storage optimization configuration method, which includes: establishing a user-side energy storage optimization configuration model that considers development factors, with the goal of maximizing the monthly comprehensive benefits of user energy storage; selecting an energy storage type; solving the user-side energy storage optimization configuration model to obtain the initial rated capacity and rated power of the user-side energy storage optimization configuration; and correcting the initial rated power and rated capacity of the user-side energy storage optimization configuration to give the final rated capacity and rated power of the user-side energy storage optimization configuration.

[0006] For example, Chinese patent CN119051091B discloses a method for optimizing the scheduling of distributed photovoltaic energy storage. It determines the objective function and constraints for optimizing the scheduling of distributed photovoltaic energy storage, constructs a model for optimizing the scheduling of distributed photovoltaic energy storage based on the objective function and constraints, and uses an improved multi-objective optimization algorithm to solve the model for multi-objective optimization, thereby obtaining a distributed photovoltaic energy storage scheduling scheme.

[0007] Currently, existing energy storage optimization technologies still have shortcomings: both patents improve the reliability of power grid operation to some extent, but they do not comprehensively assess line fault conditions by combining specific application scenarios, especially user-side status information, and cannot accurately obtain energy storage optimization data in the distribution network. Furthermore, the power dispatch control system and distribution management system are not interconnected, and therefore cannot provide decision support for high-level energy storage optimization impact analysis. Existing technologies still need improvement. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a method for optimizing the scheduling of shared energy storage in distribution substations through the collaboration of multiple flexible resources. This method considers the collaboration of electric vehicle loads, temperature-controlled loads, and energy storage batteries, and incorporates carbon emission costs into the optimization objective. The strategy aggregates distributed photovoltaic (PV) power, electric vehicle charging stations, temperature-controlled loads (air conditioning), and shared energy storage battery resources within the substation area through a Virtual Power Plant (VPP) platform. First, a day-night curve model of the current PV output is constructed, and a future incremental curve model is determined through PV forecasting. Then, combined with the quantified carbon emission cost results, the VPP issues collaborative control commands—on the one hand, guiding adjustable loads to operate during off-peak hours, and on the other hand, scheduling the day-night charging and discharging of shared energy storage batteries via a power storage converter (PCS). This aims to minimize both economic and carbon emission costs while ensuring stable grid operation and improving PV absorption rates. This addresses the aforementioned problems.

[0009] To achieve the above objectives, this invention provides the following technical solution: a method for optimized scheduling of shared energy storage in distribution substations using a multi-flexible resource coordination approach. This method is applied to residential substations equipped with electric vehicle charging stations, large-scale air conditioning (temperature-controlled loads), shared energy storage batteries, and connected to a distributed photovoltaic distribution network. The core objective is to achieve stable grid operation, full photovoltaic absorption, and minimization of both economic and carbon emission costs through the coordinated scheduling of "distributed photovoltaic - multiple flexible resources - shared energy storage." The solution uses "data acquisition - model building - constraint setting - real-time monitoring - dynamic control" as its core process, combined with the substation multi-load coordinated optimization logic in its working principle, and mainly includes the following steps: S1: Photovoltaic curve construction and basic parameter initialization; S1-1. Photovoltaic Data Acquisition and Curve Generation: Collect historical operating data and real-time output power of all distributed photovoltaic systems within the distribution area. Determine the daytime and nighttime photovoltaic output curves of the current distribution area through data fitting, distinguishing between peak output periods during the day and zero output periods at night. Simultaneously, adopt a photovoltaic prediction algorithm and a machine learning prediction model based on meteorological data to calculate the incremental curve of photovoltaic output in the next 24 hours, clarifying the fluctuation range and peak nodes of photovoltaic output in each period.

[0010] S1-2. Time Dimension and Basic Parameter Setting: The 24-hour control cycle is divided into 96 control periods, with each period having a time interval of Δt=15min (i.e., control cycle T={1, 2, ..., 96}), providing time granularity support for subsequent refined scheduling; at the same time, key basic parameters are initialized, including the benchmark price for charging and discharging of shared energy storage, the time-of-use electricity price range for each distribution area, and the unit carbon emission cost, which are determined based on regional carbon tax standards or carbon quota trading prices, laying the foundation for subsequent cost accounting and optimization modeling.

[0011] S2: System Decision Model Construction Based on the diverse resource characteristics within the distribution area, a system decision-making model covering the coordination of "load-energy storage-photovoltaics" is constructed to clarify the optimization objectives and resource operation characteristics of each entity.

[0012] S2-1. Quantification of Resource Characteristics: Electric vehicle charging station load: Statistically analyze the time distribution of user charging demand, such as the peak charging period from 18:00 to 22:00 after commuting, to determine the upper limit of single-pile charging power, charging time requirements, and the proportion of load that can be staggered. Shared energy storage batteries: Defining the charge and discharge efficiency of energy storage ( For charging efficiency, (for discharge efficiency), rated capacity ( ), the safe range of state of charge (SOC) (0≤S≤1), and the maximum charge and discharge power limits corresponding to different SOC ranges.

[0013] S2-2. Embedding of a Two-Dimensional Objective Function: User-side objective function: The core objective is to minimize the overall user cost. Cost components include the operating cost of electrical equipment and the cost of electricity (time-of-use pricing). The calculation and sharing of energy storage usage costs, and the innovative addition of carbon emission cost items (based on the quantitative calculation of indirect carbon emissions from users' use of non-clean energy and energy storage charging and discharging). The objective function of the distribution area is centered on "peak load reduction + full absorption of distributed photovoltaic power", while linking peak shaving costs (including compensation for user comfort loss and loss of revenue from energy storage peak charging) with carbon emission control targets (limiting carbon emissions from fossil fuel energy replenishment during peak hours), forming a coordinated optimization direction of "technical indicators - economic costs - low carbon targets".

[0014] S3: Multi-dimensional constraint settings ensure scheduling safety and low carbon emissions.

[0015] Based on the power grid operation requirements, resource physical characteristics, and carbon emission control needs of the distribution area, four types of core constraints are set to ensure that the dispatching plan is feasible and meets the objectives: S3-1. Photovoltaic power generation constraint: Real-time photovoltaic output of user i Must meet ( (This is the upper limit of photovoltaic output for users), and the total photovoltaic output of the distribution area must be matched with the load and energy storage charging and discharging power to avoid carbon emission losses caused by curtailment (i.e., photovoltaic output is prioritized for self-use or storage and not directly wasted). S3-2. Flexible Load and Energy Storage Constraints: Temperature control load: The air conditioner's operating power must be within the equipment's rated power range, and the temperature adjustment frequency must not exceed the user's comfort threshold (e.g., no more than once per hour). Electric vehicle charging: The charging power of a single charging pile shall not exceed the rated power of the charging pile, and the basic charging needs of users shall be guaranteed during off-peak scheduling (such as charging time not less than the minimum required time). Shared energy storage: The charging and discharging power of user-side energy storage must meet the following requirements. (The negative sign indicates discharge), the total energy storage charging and discharging power of the distribution area must meet the following requirements. (Avoid energy storage overload); at the same time, the energy storage SOC needs to be maintained. And at the end of the scheduling period It is necessary to return to the initial safe value (to avoid power shortage in energy storage the next day); S3-3. User Power Purchase and Sale Constraints: Power Purchased by User i Must meet (The negative sign indicates electricity sales), and a service fee must be deducted when selling electricity. To prevent users from excessively selling electricity and affecting the stability of the power grid; S3-4. Implicit Carbon Emission Constraints: Limiting total carbon emissions through indirect parameters, such as ensuring that peak-hour fossil fuel power replenishment in the distribution area does not exceed 10% of the total load, and that the carbon emission intensity during energy storage charging and discharging does not exceed the regional low-carbon standard.

[0016] S4: Real-time Data Acquisition and Collaborative Computing Platform Monitoring and Analysis S4-1. Full-Dimensional Data Acquisition and Transmission: Through devices such as smart meters, photovoltaic inverters, energy storage monitoring terminals, and load controllers within the distribution area, the following data are collected in real time and transmitted to the collaborative computing platform: Photovoltaic power generation data: real-time output, cumulative power generation, and prediction deviation of each photovoltaic module; Load operation data: real-time power and temperature setpoint of temperature-controlled load, charging pile operation status and power of electric vehicle charging station, and time period data of residential basic load (non-adjustable); Energy storage device data: Real-time SOC, charging and discharging power, charging and discharging efficiency, and equipment fault information of shared energy storage; Cost and carbon emission data: real-time time-of-use electricity price, unit carbon emission cost, and user-incurred electricity purchase cost and carbon emission cost.

[0017] S4-2. Platform Data Analysis and Anomaly Identification: The collaborative computing platform performs three core processes on the collected data: Compliance verification: Compare the data with the S3 constraints and identify abnormal data that deviate from the constraints (such as energy storage SOC below 0.2, photovoltaic output exceeding the upper limit, and electric vehicle charging power overload). Low-carbon performance analysis: Calculate the total carbon emissions of the transformer area in the current period (based on fossil fuel power replenishment and carbon emissions from energy storage charging and discharging). If the emissions exceed the low-carbon target threshold, they are marked as abnormal carbon emission data. Optimization assessment: Analyze the power matching degree of the current load-PV-energy storage, identify efficiency problems such as PV curtailment, excessive peak load, and idle energy storage, and generate a report of nodes to be optimized.

[0018] S4-3. Reporting of Abnormal Data: Report data on compliance abnormalities, carbon emission abnormalities, and efficiency abnormalities to the area management terminal, and automatically trigger the early warning mechanism to ensure timely response to abnormal issues.

[0019] S5: Issuance of control instructions and coordinated execution of multiple resources Based on the analysis results from the collaborative computing platform, and combined with the hierarchical logic of "user-side iterative optimization - transformer area-side peak shaving optimization", control commands are issued and executed step by step: S5-1. User-side optimization instruction execution: Issue optimization instructions based on "minimum overall cost (including carbon cost)" to each user terminal to guide users to adjust their electricity consumption and energy storage strategies. During peak photovoltaic output periods (e.g., 10:00-15:00): users are instructed to prioritize self-consumption of photovoltaic power to reduce grid purchases; at the same time, shared energy storage is controlled to enter charging mode (using low-priced electricity and low-carbon energy for charging). During peak electricity consumption and carbon emission periods (e.g., 18:00-21:00): Instruct electric vehicle charging stations to stagger charging times (shift to charging after 22:00), adjust air conditioning temperature settings (e.g., raise by 1-2℃ in summer), and have shared energy storage enter discharge mode (to supplement load gaps and reduce fossil fuel energy replenishment). S5-2. Peak Shaving and Low-Carbon Control at the Distribution Center Side: Based on the execution results from the user side, the distribution center management end further conducts coordinated control during peak load periods. Temperature control load cluster regulation: Batch scheduling of air conditioning loads operating centrally within the distribution area, reducing peak loads through "rotational shutdown" or "power reduction" methods (such as timed start and stop of air conditioning in each building), while ensuring minimal loss of user comfort; Shared energy storage for coordinated peak shaving: Adjust the charging and discharging plans of public energy storage on the distribution area side. If the peak load still exceeds the upper limit, instruct the public energy storage to increase the discharge power; if the photovoltaic output is excessive, instruct the public energy storage to extend the charging time to avoid curtailment of solar power. S5-3. Closed-loop verification of control effect: After each distribution area executes the instruction, it feeds back the adjusted power generation (photovoltaic, supplementary power), storage power (shared energy storage), and load power data to the collaborative computing platform in real time. The platform verifies whether the three objectives of "full photovoltaic consumption, compliance with constraints, and minimum carbon emission cost" are met. If not, the instruction is iterated and optimized again until the optimal dispatch effect is achieved.

[0020] Furthermore, a system decision model is constructed to quantify the correlation characteristics of temperature control load temperature adjustment threshold, the load distribution characteristics of electric vehicle charging stations during different time periods, and the safe range of the state of charge of shared energy storage batteries. User-side objective functions and transformer-side objective functions are embedded. The user-side objective function includes electricity purchase cost, shared energy storage usage cost, and carbon emission cost, while the transformer-side objective function includes peak load reduction and carbon emission control.

[0021] Furthermore, the elastic load and energy storage constraints require sharing the energy storage state of charge to maintain a safe range, and the implicit carbon emission constraints limit the proportion of fossil energy supplementary power during peak hours.

[0022] Furthermore, the collaborative computing platform performs compliance verification, low-carbon analysis, and optimization assessment on the collected data, identifies abnormal data and reports it, and obtains the abnormal identification results.

[0023] Furthermore, the user-side optimization command guides self-consumption and shared energy storage charging during peak photovoltaic periods, while the distribution area-side control command adjusts the temperature-controlled load rotation and shared energy storage charging and discharging plans during peak load periods.

[0024] Furthermore, the feedback is sent to the collaborative computing platform.

[0025] Furthermore, if the conditions are not met, the instructions are iteratively adjusted and optimized to obtain the final scheduling scheme.

[0026] Furthermore, the Pareto optimal frontier curves corresponding to peak shaving costs and peak loads are generated to determine the target value for comprehensive optimization of the transformer substation.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows: By incorporating four core resource types—distributed photovoltaic (PV) power, electric vehicle charging station loads, temperature-controlled loads, and shared energy storage batteries—into a unified scheduling system, a system decision-making model covering the characteristics of these four resource types is established. Multi-dimensional constraints are set to form a coordinated "source-storage-load" scheduling loop, effectively mitigating PV power output fluctuations and improving the power flow stability and voltage control accuracy of the distribution network. Integrating low-carbon target design, carbon emissions are indirectly converted into carbon costs and incorporated into the comprehensive cost objective function within the core optimization logic. A "unit carbon emission cost" is introduced, quantifying carbon emissions from fossil fuel supplementation due to insufficient PV absorption within the distribution area and indirect carbon emissions from energy storage charging and discharging as cost items, achieving the dual minimization of "economic cost + carbon cost." A hierarchical iterative optimization mechanism is constructed to balance the interests of both users and the power grid. An innovative hierarchical optimization architecture, "user-side-transformer-side," is adopted. On the user side, the goal is to minimize overall cost, generating personalized electricity consumption plans using the CPLEX tool. On the transformer-side, the core focus is on peak load reduction and full renewable energy integration, dynamically adjusting energy storage prices and regulating air conditioning load based on user plans to generate a Pareto optimal curve for peak shaving cost versus peak load. This avoids the one-sidedness of "single-dimensional optimization," reducing both user electricity and energy storage costs while minimizing peak-valley differences in the power grid and increasing the integration rate of renewable energy sources such as photovoltaics, thus achieving a balance of interests between users and the power grid. It ensures both user economic viability and meets the operational needs of the power grid. Real-time monitoring and dynamic control ensure system stability. A full-process mechanism of "real-time data acquisition - collaborative computing platform analysis - anomaly reporting - control command issuance" is established. This mechanism collects real-time data on photovoltaic power generation, energy storage devices, and load operation, quickly identifies and reports abnormal data deviating from constraints, and adjusts power generation and storage capacity through regional control commands. Real-time response effectively avoids the inability of traditional "static dispatch" to handle sudden operating conditions, ensuring stable system operation even under photovoltaic output fluctuations and sudden load changes, thus improving regional power supply reliability and resource utilization efficiency. Through multi-dimensional technological innovation, optimized dispatch is achieved, solving grid operation problems and responding to the national "dual-carbon" strategy, breaking through existing technological limitations and realizing multi-flexible resource collaborative dispatch. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the solution process of a multi-load collaborative optimization model for a distribution area shared energy storage optimization scheduling method based on multiple flexible resources, as described in this invention. Figure 2 This is a user-side load optimization data diagram for a multi-flexible resource collaborative distribution area shared energy storage optimization scheduling method according to the present invention; Figure 3 This is a multi-load collaborative optimization data diagram for a multi-flexible resource collaborative distribution area shared energy storage optimization scheduling method of the present invention; Figure 4The charging and discharging power and state of charge diagram of the energy storage on the distribution substation side in the present invention is a method for optimizing the scheduling of shared energy storage in distribution substations with multiple flexible resources. Figure 5 This is a hardware and software system architecture diagram of a distribution area virtual power plant (VPP) based on a method for optimizing the scheduling of shared energy storage in distribution areas using multiple flexible resources, as described in this invention. Detailed Implementation

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0030] This invention provides a technical solution: a method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources, using "data acquisition - model building - constraint setting - real-time monitoring - dynamic control" as the core process, combined with the multi-load collaborative optimization logic of the substation in the working principle. The specific steps are as follows: S1: Photovoltaic Curve Construction and Basic Parameter Initialization S1-1. Photovoltaic Data Acquisition and Curve Generation: Collect historical operating data and real-time output power of all distributed photovoltaic systems within the distribution area. Determine the daytime and nighttime photovoltaic output curves of the current distribution area through data fitting (distinguishing between peak daytime output periods and zero nighttime output periods). Simultaneously, use photovoltaic prediction algorithms (such as machine learning prediction models based on meteorological data) to calculate the incremental curve of photovoltaic output in the next 24 hours, clarifying the fluctuation range and peak nodes of photovoltaic output in each period.

[0031] S1-2. Time Dimension and Basic Parameter Setting: The 24-hour control cycle is divided into 96 control periods, with each period having a time interval Δt=15min (i.e., control cycle T={1, 2, ..., 96}), providing time granularity support for subsequent refined scheduling; at the same time, key basic parameters are initialized, including the benchmark price for charging and discharging of shared energy storage, the time-of-use electricity price range for each distribution area, and the unit carbon emission cost (determined based on regional carbon tax standards or carbon quota trading prices), laying the foundation for subsequent cost accounting and optimization modeling.

[0032] S2: System Decision Model Construction Based on the diverse resource characteristics within the distribution area, a system decision-making model covering the coordination of "load-energy storage-photovoltaics" is constructed to clarify the optimization objectives and resource operation characteristics of each entity.

[0033] S2-1. Quantification of Resource Characteristics: Temperature-controlled loads (such as air conditioners): Determine the correlation characteristics between "temperature adjustment threshold and operating power", quantify the power consumption range and start-stop response time of air conditioners in different temperature ranges, and clarify their adjustable time window and power elasticity space; Electric vehicle charging station load: Statistically analyze the time distribution of user charging demand (e.g., the peak charging period is 18:00-22:00 after commuting) to determine the upper limit of single-pile charging power, charging time requirements, and the proportion of load that can be scheduled during off-peak hours. Shared energy storage batteries: Defining the charge and discharge efficiency of energy storage (ξ) C For charging efficiency, ξ D (discharge efficiency), rated capacity (G) ESN ), the safe range of state of charge (SOC) (0≤S≤1), and the maximum charge and discharge power limits corresponding to different SOC ranges.

[0034] S2-2. Embedding of a Two-Dimensional Objective Function: User-side objective function: The core objective is to minimize the overall user cost. Cost components include the operating cost of electrical equipment and the cost of electricity purchase (time-of-use price π). t The calculation and sharing of energy storage usage costs, and the innovative addition of carbon emission cost items (based on the quantitative calculation of indirect carbon emissions from users' use of non-clean energy and energy storage charging and discharging). The objective function of the distribution area is centered on "peak load reduction + full absorption of distributed photovoltaic power", while linking peak shaving costs (including compensation for user comfort loss and loss of revenue from energy storage peak charging) with carbon emission control targets (limiting carbon emissions from fossil fuel energy replenishment during peak hours), forming a coordinated optimization direction of "technical indicators - economic costs - low carbon targets".

[0035] S3: Setting multi-dimensional constraints (ensuring scheduling safety and low carbon footprint) Based on the power grid operation requirements, resource physical characteristics, and carbon emission control needs of the distribution area, four types of core constraints are set to ensure that the dispatching plan is feasible and meets the objectives: S3-1. Photovoltaic power generation constraint: Real-time photovoltaic output of user i Must meet , (This is the upper limit of photovoltaic output for users), and the total photovoltaic output of the distribution area must be matched with the load and energy storage charging and discharging power to avoid carbon emission losses caused by curtailment (i.e., photovoltaic output is prioritized for self-use or storage and not directly wasted). S3-2. Flexible Load and Energy Storage Constraints: Temperature control load: The air conditioner's operating power must be within the equipment's rated power range, and the temperature adjustment frequency must not exceed the user's comfort threshold (e.g., no more than once per hour). Electric vehicle charging: The charging power of a single charging pile shall not exceed the rated power of the charging pile, and the basic charging needs of users shall be guaranteed during off-peak scheduling (such as charging time not less than the minimum required time). Shared energy storage: The charging and discharging power of user-side energy storage must meet the following requirements. (The negative sign indicates discharge), the total energy storage charging and discharging power of the distribution area must meet the following requirements. (Avoid energy storage overload); at the same time, the energy storage SOC needs to be maintained. And at the end of the scheduling period It is necessary to return to the initial safe value (to avoid power shortage in energy storage the next day); S3-3. User Power Purchase and Sale Constraints: Power Purchased by User i Must meet (The negative sign indicates electricity sales), and a service fee must be deducted when selling electricity. To prevent users from excessively selling electricity and affecting the stability of the power grid; S3-4. Implicit Carbon Emission Constraints: Limiting total carbon emissions through indirect parameters, such as ensuring that peak-hour fossil fuel power replenishment in the distribution area does not exceed 10% of the total load, and that the carbon emission intensity during energy storage charging and discharging does not exceed the regional low-carbon standard.

[0036] S4: Real-time Data Acquisition and Collaborative Computing Platform Monitoring and Analysis S4-1. Full-Dimensional Data Acquisition and Transmission: Through devices such as smart meters, photovoltaic inverters, energy storage monitoring terminals, and load controllers within the distribution area, the following data are collected in real time and transmitted to the collaborative computing platform: Photovoltaic power generation data: real-time output, cumulative power generation, and prediction deviation of each photovoltaic module; Load operation data: real-time power and temperature setpoint of temperature-controlled load, charging pile operation status and power of electric vehicle charging station, and time period data of residential basic load (non-adjustable); Energy storage device data: Real-time SOC, charging and discharging power, charging and discharging efficiency, and equipment fault information of shared energy storage; Cost and carbon emission data: real-time time-of-use electricity price, unit carbon emission cost, and user-incurred electricity purchase cost and carbon emission cost.

[0037] S4-2. Platform Data Analysis and Anomaly Identification: The collaborative computing platform performs three core processes on the collected data: Compliance verification: Compare the data with the S3 constraints and identify abnormal data that deviate from the constraints (such as energy storage SOC below 0.2, photovoltaic output exceeding the upper limit, and electric vehicle charging power overload). Low-carbon performance analysis: Calculate the total carbon emissions of the transformer area in the current period (based on fossil fuel power replenishment and carbon emissions from energy storage charging and discharging). If the emissions exceed the low-carbon target threshold, they are marked as abnormal carbon emission data. Optimization assessment: Analyze the power matching degree of the current load-PV-energy storage, identify efficiency problems such as PV curtailment, excessive peak load, and idle energy storage, and generate a report of nodes to be optimized.

[0038] S4-3. Reporting of Abnormal Data: Report data on compliance abnormalities, carbon emission abnormalities, and efficiency abnormalities to the area management terminal, and automatically trigger the early warning mechanism to ensure timely response to abnormal issues.

[0039] S5: Issuance of control instructions and coordinated execution of multiple resources Based on the analysis results from the collaborative computing platform, and combined with the hierarchical logic of "user-side iterative optimization - transformer area-side peak shaving optimization", control commands are issued and executed step by step: S5-1. User-side optimization instruction execution: Issue optimization instructions based on "minimum overall cost (including carbon cost)" to each user terminal to guide users to adjust their electricity consumption and energy storage strategies. During peak photovoltaic output periods (e.g., 10:00-15:00): users are instructed to prioritize self-consumption of photovoltaic power to reduce grid purchases; at the same time, shared energy storage is controlled to enter charging mode (using low-priced electricity and low-carbon energy for charging). During peak electricity consumption and carbon emission periods (e.g., 18:00-21:00): Instruct electric vehicle charging stations to stagger charging times (shift to charging after 22:00), adjust air conditioning temperature settings (e.g., raise by 1-2℃ in summer), and have shared energy storage enter discharge mode (to supplement load gaps and reduce fossil fuel energy replenishment). S5-2. Peak Shaving and Low-Carbon Control at the Distribution Center Side: Based on the execution results from the user side, the distribution center management end further conducts coordinated control during peak load periods. Temperature control load cluster regulation: Batch scheduling of air conditioning loads operating centrally within the distribution area, reducing peak loads through "rotational shutdown" or "power reduction" methods (such as timed start and stop of air conditioning in each building), while ensuring minimal loss of user comfort; Shared energy storage for coordinated peak shaving: Adjust the charging and discharging plans of public energy storage on the distribution area side. If the peak load still exceeds the upper limit, instruct the public energy storage to increase the discharge power; if the photovoltaic output is excessive, instruct the public energy storage to extend the charging time to avoid curtailment of solar power. S5-3. Closed-loop verification of control effect: After each distribution area executes the instruction, it feeds back the adjusted power generation (photovoltaic, supplementary power), storage power (shared energy storage), and load power data to the collaborative computing platform in real time. The platform verifies whether the three objectives of "full photovoltaic consumption, compliance with constraints, and minimum carbon emission cost" are met. If not, the instruction is iterated and optimized again until the optimal dispatch effect is achieved.

[0040] Working principle: 1. Establish a multi-load collaborative optimization model for transformer substations. Assuming the number of users in the control area is N, and the control period is 24 hours, it is divided into 96 time periods with a time interval Δt = 15 minutes, that is, the control period T = {1, 2, ..., 96}.

[0041] User-side optimization model: 1) Objective function.

[0042] The user aims to minimize overall cost, and the objective function is expressed as: (1) In the formula: The overall cost for user i; The total operating power of user i's electrical equipment. The total operating power of electrical equipment for user i during time period t; For user i, the carbon emission cost item Cost per unit of carbon emissions For user carbon emissions during period t; The power consumption that user i purchased from the transformer substation. The power consumption of user i from the distribution area during time period t; The electricity purchase cost for user i; Shared energy storage charging and discharging power for user i This represents the charging and discharging power of the shared energy storage for user i during time period t. A positive value indicates charging, and a negative value indicates discharging. Share the cost of energy storage for users.

[0043] a) The user's electricity purchase cost.

[0044] Electricity purchase cost for user i It can be represented as:

[0045]

[0046] In the formula: The electricity purchase cost for user i during time period t; The time-of-use electricity price for users purchasing electricity from the distribution area during time period t. These are the upper and lower limits of the electricity price, respectively. The service fee charged by management when selling electricity to users, and .

[0047] b) Cost of using shared energy storage.

[0048] The cost of using shared energy storage includes capacity cost and power cost, which can be expressed as follows:

[0049] In the formula: The price of shared energy storage capacity; Shared energy storage capacity for user i; These represent the charging and discharging prices for shared energy storage during time period t; These are the minimum and maximum charging prices for shared energy storage, respectively. These represent the lowest and highest discharge prices for shared energy storage, respectively.

[0050] 2) Constraints The user-side constraints are as follows:

[0051]

[0052]

[0053]

[0054]

[0055]

[0056] In the formula: Contribute to user i's photovoltaic power. Photovoltaic power output for user i during time period t; This represents the upper limit of the power consumption of user device i; , which are the maximum charging and discharging power of the shared energy storage for user i, respectively; These are the maximum charging and discharging power of the shared energy storage system within the transformer area, respectively. The upper limit of iPhotovoltaic's output for users; These represent the maximum electricity purchase and sale power of user i, respectively; The state of charge of the shared energy storage for user i during time period t; These are the charging and discharging efficiencies of the shared energy storage for user i, respectively.

[0057] 1.2 Optimization Model for Transformer Area On the distribution transformer side, while ensuring full absorption of distributed power sources, the load is optimized by combining the transformer's own energy storage equipment and the user-side elastic load. Taking peak load reduction as an example, the optimization objective function on the distribution transformer side is:

[0058]

[0059] In the formula: J is the value of the comprehensive optimization objective function; Total load of the transformer area; Contribute to renewable energy in the Taiwan area; To adjust the energy storage charging and discharging power on the front-end area side, a positive value indicates charging, and a negative value indicates discharging; Peak shaving costs on the transformer substation side; These are the weighting coefficients; The power change for user i; The charging and discharging power of the energy storage equipment on the transformer substation side is changed due to peak shaving demand.

[0060] Peak shaving costs on the distribution platform side include the revenue loss from energy storage peak-loading during user-side adjustments to flexible loads, which can be specifically expressed as:

[0061] In the formula: This refers to the retail electricity price in the affected area.

[0062] The state of charge of energy storage on the transformer substation side can be expressed as:

[0063] In the formula: The state of charge of the energy storage on the platform side during time period t; The charging and discharging power of the energy storage on the platform side during the time period; These are the charging and discharging efficiencies of the energy storage on the transformer substation side, respectively. The capacity for energy storage on the transformer substation side.

[0064] 2. Solution steps and methods for coordinated optimization of multiple loads in a distribution area The above-mentioned multi-load collaborative optimization model for distribution transformer areas based on shared energy storage mainly includes three core stages in its solution process: data initialization, user-side load iterative optimization, and distribution transformer area peak shaving optimization. The corresponding solution process can be found in the following references. Figure 2 The specific operational steps for each stage are as follows: Phase 1: Data Initialization. This phase requires completing two key tasks: first, conducting predictive analysis of user electricity load and distributed power output; and second, initially setting the benchmark prices for charging and discharging the shared energy storage system for the following day. Phase Two: User-Side Load Iterative Optimization Process. First, at the user level, with the core objective of minimizing overall comprehensive costs, the user optimizes the configuration of their diverse electricity loads and uses the CPLEX optimization tool to calculate the user's daily electricity usage plan. Subsequently, based on the load optimization results submitted by the user and combined with the output forecast data of distributed photovoltaic power sources, the substation management end dynamically adjusts the charging and discharging prices of shared energy storage at different peak and valley times. Phase 3: Implementation of Peak Shaving Optimization at the Distribution Center Level. After obtaining the user-side load optimization results, the distribution center focuses on optimizing the load control of air conditioning equipment that operates intensively during peak load periods. At the same time, it coordinates with the energy storage resources configured in the distribution center to carry out collaborative optimization calculations, ultimately generating the Pareto optimal frontier curve corresponding to the peak load and the peak shaving cost. Based on this, the optimal collaborative optimization implementation scheme is determined by comprehensively balancing the peak load control effect and the peak shaving investment cost.

[0065] Figure 1This is a schematic diagram illustrating an application scenario of a low-carbon collaborative optimization method provided in an embodiment of this application. Figure 1 In the scenario depicted, photovoltaic power generation, flexible loads, and energy storage devices within the transformer substation need to operate collaboratively to match power generation with load demand while controlling carbon emissions. This application optimizes energy allocation within the substation by setting various constraints and combining real-time data acquisition and analysis, thereby reducing carbon emissions and improving operational efficiency.

[0066] Based on the updated charging and discharging prices and load control strategies, a Pareto optimal frontier curve corresponding to peak shaving cost and peak load is generated. Specifically, the collaborative computing platform uses the updated data to construct a relationship model between peak shaving cost and peak load, and generates the Pareto optimal frontier curve through a multi-objective optimization algorithm. This curve shows the distribution of the optimal solution for peak shaving cost under different peak load levels, providing a decision-making basis for transformer area managers. This embodiment achieves a balanced optimization of peak shaving cost and load management through the generation of the Pareto optimal frontier curve.

[0067] The overall optimization target value for the power distribution area is determined based on the Pareto optimal frontier curve. After generating the Pareto optimal frontier curve, the collaborative computing platform selects the most suitable optimization point from the curve as the overall optimization target value for the power distribution area, taking into account the actual needs and operational objectives of the area. If the power distribution area prioritizes cost control, the optimization point with lower peak shaving costs is selected; if it prioritizes load stability, the optimization point with lower peak loads is selected.

[0068] The above embodiments utilize a collaborative computing platform to perform multi-objective verification and iterative optimization of photovoltaic (PV) grid integration, constraints, and carbon emission costs, generating a final scheduling scheme. By combining shared energy storage charging and discharging price updates and load regulation strategy optimization, a Pareto optimal frontier curve is constructed, ultimately determining the comprehensive optimization target value for the distribution area. This method achieves a balance among multiple objectives, improving the accuracy and reliability of scheduling optimization and providing technical support for the efficient operation of distribution areas.

[0069] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources, characterized in that, Includes the following steps: S1: Photovoltaic curve construction and basic parameter initialization; S1-1. Photovoltaic Data Acquisition and Curve Generation: Collect historical operating data and real-time output power of all distributed photovoltaic systems within the distribution area, and determine the daytime and nighttime photovoltaic output curves of the current distribution area through data fitting (distinguishing between peak daytime output periods and zero nighttime output periods); at the same time, use photovoltaic prediction algorithms (such as machine learning prediction models based on meteorological data) to calculate the incremental curve of photovoltaic output in the next 24 hours, and clarify the fluctuation range and peak nodes of photovoltaic output in each period. S1-2. Time Dimension and Basic Parameter Setting: The 24-hour control cycle is divided into 96 control periods, with each period having a time interval of Δt=15min, i.e., the control cycle T={1, 2, ..., 96}. At the same time, key basic parameters are initialized, including the charging and discharging benchmark price of shared energy storage, the time-of-use electricity price range of the distribution platform, and the unit carbon emission cost, which is determined based on regional carbon tax standards or carbon quota trading prices. S2: System decision-making model construction. Based on the characteristics of diverse resources within the distribution area, a system decision-making model covering the coordination of "load-energy storage-photovoltaics" is constructed to clarify the optimization objectives and resource operation characteristics of each entity. S2-1. Quantification of Resource Characteristics: Temperature-controlled loads such as air conditioners: Determine the correlation characteristics between "temperature adjustment threshold and operating power", quantify the power consumption range and start-stop response time of air conditioners in different temperature ranges, and clarify their adjustable time window and power elasticity space; Electric vehicle charging station loads: Statistically analyze the time distribution of user charging demand, such as the charging peak from 18:00 to 22:00 after commuting, determine the upper limit of single-pile charging power, charging time requirements, and the proportion of load that can be staggered. Shared energy storage batteries: Determine the charge and discharge efficiency of energy storage. For charging efficiency, For discharge efficiency and rated capacity 1. The safe range of State of Charge (SOC) is 0 ≤ S ≤ 1, and the maximum charge / discharge power limits corresponding to different SOC ranges; S2-2. Embedding of a two-dimensional objective function: The user-side objective function is centered on "minimizing the overall user cost," and the cost components include the operating cost of electrical equipment, the cost of purchasing electricity, and the time-of-use electricity price. The system calculates and shares the cost of using energy storage, and innovatively incorporates a carbon emission cost item, which is based on the quantitative calculation of the indirect carbon emissions from the user's use of non-clean energy and the charging and discharging of energy storage. The objective function of the distribution area is centered on "peak load reduction + full absorption of distributed photovoltaic power", while also considering peak shaving costs, including compensation for user comfort loss, loss of revenue from energy storage peak charging, and carbon emission control targets that limit the carbon emissions of fossil fuel supplementation during peak hours, forming a coordinated optimization direction of "technical indicators - economic costs - low carbon targets". S3: Multi-dimensional constraint settings ensure scheduling safety and low carbon emissions; Based on the power grid operation requirements, resource physical characteristics, and carbon emission control needs of the distribution area, four types of core constraints are set: S3-1. Photovoltaic power generation constraint: Real-time photovoltaic output of user i Must meet The upper limit of photovoltaic output for users is set, and the total photovoltaic output of the distribution area must be matched with the load and energy storage charging and discharging power to avoid carbon emission losses caused by curtailment. In other words, photovoltaic output is prioritized for self-use or storage and is not directly wasted. S3-2. Flexible Load and Energy Storage Constraints: Temperature control load: The air conditioner's operating power must be within the equipment's rated power range, and the temperature adjustment frequency must not exceed the user's comfort threshold, such as starting and stopping no more than once per hour; Electric vehicle charging: The charging power of a single charging pile shall not exceed the rated power of the charging pile, and the basic charging needs of users shall be guaranteed during off-peak scheduling, such as the charging time not being less than the minimum required time. Shared energy storage: The charging and discharging power of user-side energy storage must meet the following requirements. The negative sign indicates discharge, and the total energy storage charging and discharging power of the transformer area must meet the following requirements. To avoid energy storage overload; at the same time, the energy storage SOC needs to be maintained. And at the end of the scheduling period It is necessary to return to the initial safe value to avoid energy storage shortages the next day; S3-3. User Power Purchase and Sale Constraints: Power Purchased by User i Must meet The negative sign indicates electricity sales, and a service fee must be deducted when selling electricity. To prevent users from excessively selling electricity and affecting the stability of the power grid; S3-4. Implicit carbon emission constraints: Limiting total carbon emissions through indirect parameters, such as ensuring that peak-hour fossil fuel power replenishment in the distribution area does not exceed 10% of the total load, and that the carbon emission intensity during energy storage charging and discharging does not exceed the regional low-carbon standard. S4: Real-time data acquisition and collaborative computing platform monitoring and analysis; S4-1. Full-Dimensional Data Acquisition and Transmission: Through devices such as smart meters, photovoltaic inverters, energy storage monitoring terminals, and load controllers within the distribution area, the following data are collected in real time and transmitted to the collaborative computing platform: Photovoltaic power generation data: real-time output, cumulative power generation, and prediction deviation of each photovoltaic module; Load operation data: real-time power and temperature setpoint of temperature-controlled load, charging pile operation status and power of electric vehicle charging stations, and data on unadjustable time periods of residential basic load; Energy storage device data: Real-time SOC, charging and discharging power, charging and discharging efficiency, and equipment fault information of shared energy storage; Cost and carbon emission data: real-time time-of-use electricity price, unit carbon emission cost, and user-incurred electricity purchase cost and carbon emission cost; S4-2. Platform Data Analysis and Anomaly Identification: The collaborative computing platform performs three core processes on the collected data: Compliance verification: Compare the data with the S3 constraints to identify abnormal data that deviate from the constraints, such as energy storage SOC below 0.2, photovoltaic output exceeding the upper limit, and electric vehicle charging power overload; Low-carbon performance analysis: Calculate the total carbon emissions of the transformer area during the current period, based on the amount of electricity replenished by fossil fuels and the carbon emissions from energy storage charging and discharging. If the emissions exceed the low-carbon target threshold, they are marked as abnormal carbon emission data. Optimization assessment: Analyze the power matching degree of the current load-PV-energy storage, identify efficiency problems such as PV curtailment, excessive peak load, and idle energy storage, and generate a report on nodes to be optimized; S4-3. Reporting of Abnormal Data: Report compliance abnormalities, carbon emission abnormalities, and efficiency abnormalities to the area management terminal, and automatically trigger the early warning mechanism to ensure timely response to abnormal issues; S5: Issuance of control instructions and coordinated execution of multiple resources; Based on the analysis results from the collaborative computing platform, and combined with the hierarchical logic of "user-side iterative optimization - transformer area-side peak shaving optimization", control commands are issued and executed step by step: S5-1. User-side optimization instruction execution: Issue optimization instructions based on "minimum overall cost, including carbon cost" to each user terminal to guide users to adjust their electricity consumption and energy storage strategies. During peak photovoltaic output periods, such as 10:00-15:00: users are instructed to prioritize self-consumption of photovoltaic power to reduce the purchase of electricity from the grid; at the same time, shared energy storage is controlled to enter charging mode to utilize low-priced electricity and low-carbon energy for charging. During peak electricity consumption and carbon emission periods, such as 18:00-21:00: Instruct electric vehicle charging stations to shift charging to 22:00, adjust air conditioning temperature settings (e.g., raise by 1-2℃ in summer), and have shared energy storage enter discharge mode to supplement load gaps and reduce fossil fuel energy replenishment. S5-2. Peak Shaving and Low-Carbon Control at the Distribution Center Side: Based on the execution results from the user side, the distribution center management end further conducts coordinated control during peak load periods. Temperature control load cluster regulation: Batch scheduling of air conditioning loads operating in a centralized area, reducing peak loads through "rotational shutdown" or "power reduction" methods, such as staggered start and stop times for air conditioning in each building, while ensuring minimal loss of user comfort; Shared energy storage for coordinated peak shaving: Adjust the charging and discharging plans of public energy storage on the distribution area side. If the peak load still exceeds the upper limit, instruct the public energy storage to increase the discharge power; if the photovoltaic output is excessive, instruct the public energy storage to extend the charging time to avoid curtailment of solar power. S5-3. Closed-loop verification of control effect: After each distribution area executes the instruction, it provides real-time feedback on the adjusted power generation. Photovoltaic, supplementary power, and storage power share energy storage and load power data to the collaborative computing platform. The platform verifies whether the three objectives of "full photovoltaic consumption, compliance with constraints, and minimum carbon emission cost" are met. If not, the instruction is iterated and optimized again until the optimal dispatch effect is achieved. Carbon emissions are converted into "carbon costs" and embedded into the user-side and transformer-side objective function, with resource data support provided by VPP. User-side objective function construction: The user's objective is clearly defined as "minimizing overall cost," and its mathematical expression is as follows: （1） In equation (1), For user i's overall cost, For electricity purchase costs, To share the cost of energy storage usage, For carbon emission costs, Cost per unit of carbon emissions For user carbon emissions during period t; Objective function correlation on the transformer substation side: The objective on the transformer substation side is "peak load reduction + carbon emission control", and the mathematical model is as follows: （2） （3） In equations (2) and (3): J is the value of the comprehensive optimization objective function; Total load of the transformer area; Contribute to renewable energy in the Taiwan area; To adjust the energy storage charging and discharging power on the front-end area side, a positive value indicates charging, and a negative value indicates discharging; Peak shaving costs on the transformer substation side; These are the weighting coefficients; The power change for user i; The charging and discharging power of the energy storage equipment on the transformer substation side is changed due to peak shaving demand; Peak shaving costs on the distribution platform side include the revenue loss from energy storage peak-loading during user-side adjustments to flexible loads, which can be specifically expressed as: （4） In equation (4): For the retail electricity price in the distribution area, The state of charge of energy storage on the transformer substation side can be expressed as: （5） In the formula: The state of charge of the energy storage on the platform side during time period t; The charging and discharging power of the energy storage on the platform side during the time period; These are the charging and discharging efficiencies of the energy storage on the transformer substation side, respectively. The capacity for energy storage on the transformer substation side; Relying on the VPP platform, real-time operational data of distributed photovoltaic, flexible loads, and energy storage devices within the distribution area (such as...) , ), which is the objective function These parameters provide precise input, ensuring the accuracy of cost quantification and optimization calculations; Guided by carbon emission control, and combining the CPLEX optimization algorithm with VPP data interaction in a hierarchical iterative optimization mechanism, the specific process and mathematical constraints are as follows: User-side iterative optimization: The CPLEX tool is used to solve the user-side objective function while satisfying multiple constraints. Key mathematical constraints include: Power balance constraints: ; Energy storage SOC constraints: ； Implicit constraints on carbon emissions: , The upper limit of carbon emission coefficient per unit of electricity purchased; By solving the constrained objective function using CPLEX, optimal electricity consumption and energy storage charging / discharging strategies for users can be generated, such as prioritizing the absorption of photovoltaic power to reduce [electricity consumption and energy storage charging / discharging costs]. , Peak shaving optimization on the transformer substation side: Based on the optimization results on the user side, the objective function on the transformer substation side is solved and the Pareto optimal frontier curve of "peak shaving cost - peak load - carbon emissions" is generated. The key is to adjust the power deviation of temperature-controlled loads. Deviation from the energy storage power of the distribution area Ensure that peak-hour carbon emissions from fossil fuel replenishment do not exceed 10% of the total load; VPP assists in completing user-side optimization results, such as The interaction and transmission of control commands with the distribution area avoids iterative optimization deviations caused by data silos. The combination of this hierarchical iterative mechanism and the CPLEX solution algorithm is the core operational path for achieving carbon emission targets. The "data acquisition-constraint verification" technology system, which relies on VPP to achieve data integration and provides support for carbon emission quantification, specifically includes: Multi-dimensional data acquisition specifications: By integrating real-time data from multiple devices within the distribution area through the VPP platform, the core data to be collected and their mathematical relationships are as follows: Photovoltaic data: For real-time photovoltaic power output, This is the upper limit of photovoltaic output, used for accounting. Carbon emission reductions from "photovoltaic power substitution for electricity purchase" in China Energy storage data: For charging and discharging power, State of charge The charge / discharge efficiency is used to calculate the indirect carbon emissions from energy storage charging and discharging. , The carbon emission factor for energy storage, Load data: For temperature control / electric vehicle load power, This is used to verify the compliance of implicit carbon emission constraints regarding load regulation deviation. Constraint Verification Mathematical Logic: The VPP platform automatically verifies three core constraints based on collected data: Photovoltaic output constraints: To avoid carbon emission losses caused by curtailment of solar power; Energy storage charging and discharging power constraints: To avoid inefficient carbon emissions caused by energy storage overload; Carbon emission limits: This is the upper limit for carbon emissions during the designated time period in the distribution area; By enabling unified data collection and constraint verification through VPP, a fundamental guarantee is provided for the quantification of carbon emission costs and the implementation of optimization targets.

2. The method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources as described in claim 1, characterized in that, Constructing a system decision model integrating multiple elastic resource characteristics, and a technical solution for integrating and modeling the characteristics of three types of elastic resources in residential areas—temperature-controlled loads such as air conditioners, electric vehicle charging station loads, and shared energy storage batteries—is the fundamental support for achieving optimized scheduling with carbon emission considerations. In step S2, by quantitatively analyzing the core operating characteristics of the three types of resources—the correlation between the "temperature regulation range and power consumption" of temperature-controlled loads, such as the correspondence between air conditioner start / stop thresholds and operating power; the "charging demand period and charging power fluctuation" characteristics of electric vehicle charging stations, such as the concentrated power demand during peak charging periods; and the "charge / discharge efficiency, state of charge (SOC), and power limitation" characteristics of shared energy storage batteries, such as the differences in the upper limit of charge / discharge power in different SOC ranges—a unified system decision model is constructed. This transforms the operating boundaries and constraints of the three types of resources into a computable parameterized model, achieving the synergistic integration of multiple resource characteristics.

3. The method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources as described in claim 2, characterized in that, The hierarchical solution process of "user-side iterative optimization - distribution area peak shaving optimization," aimed at achieving synergy between "economic cost, grid efficiency, and carbon emission reduction," is a key operational path for realizing core carbon emission optimization goals. The specific process corresponds to three core steps in the solution phase: The first phase completes user load and distributed photovoltaic output prediction, and initializes the benchmark price for shared energy storage, providing a data foundation for subsequent optimization. The second phase, on the user side, aims to "minimize the overall cost including carbon costs," using the CPLEX optimization tool to generate electricity consumption and shared energy storage charging / discharging schemes. The distribution area management end dynamically adjusts the charging / discharging price of shared energy storage based on these schemes, such as guiding users to charge energy storage during low-carbon periods. The third phase, on the distribution area side, aims to achieve "full absorption of distributed photovoltaic power + peak load reduction," combining user schemes to regulate the temperature-controlled load operation status and generate the Pareto optimal frontier curve of "peak shaving cost - peak load," ultimately determining an execution scheme that balances multiple objectives. This hierarchical iterative process, through the two-way interaction of "user-autonomous optimization - distribution area coordinated control," ensures the synergistic achievement of core carbon emission goals with economic and grid objectives, and its process is unique and operable.

4. The method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources as described in claim 3, characterized in that, The real-time data-driven anomaly monitoring and dynamic control mechanism, specifically in step S4, involves real-time collection of photovoltaic power generation, energy storage device operation data (charging and discharging power, SOC, efficiency), and user load data (temperature control load, electric vehicle charging load) within the distribution area. This data is then transmitted to a collaborative computing platform. The platform compares this data with preset constraints such as photovoltaic power limits, energy storage charging and discharging power limits, and SOC safety ranges. It automatically identifies abnormal data deviating from these constraints, such as overcharging of energy storage or a sudden drop in photovoltaic output leading to carbon emission risks, and reports these in real-time. Based on the anomaly analysis results, the platform issues control commands to each distribution area, adjusting power generation (e.g., limiting non-clean energy replenishment) and storage power (e.g., adjusting energy storage discharge strategies). This ensures the system operates within the constraints, avoiding the limitations of static scheduling in handling sudden operating conditions and providing operational assurance for the stable execution of the core carbon emission optimization strategy.

5. The method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources as described in claim 4, characterized in that, The design of multi-dimensional constraints for shared energy storage charging and discharging is proposed. This design scheme supports multi-dimensional constraints for shared energy storage devices, including scheduling considerations related to carbon emissions, and covers both user-side and distribution area-side constraints. The user-side constraints explicitly define the maximum charging and discharging power constraint of the shared energy storage. ), Overall charging and discharging power constraint of the transformer area ( ), State of Charge (SOC) constraints ) and charge / discharge efficiency constraints ( This ensures the efficiency and safety of user-side energy storage operation. In the constraints on the distribution area side, it further supplements the SOC constraints at the beginning and end of the energy storage period. For example, the SOC is maintained within a reasonable range at the end of the dispatch cycle to avoid increased carbon emissions the next day. It also includes charging and discharging constraints linked to photovoltaic consumption, such as prioritizing energy storage charging during peak photovoltaic output to reduce carbon emission losses caused by curtailment. This multi-dimensional constraint design limits the energy storage operation boundary to avoid inefficient energy storage operation, such as frequent charging and discharging increasing indirect carbon emissions, and provides equipment-level constraint guarantees for achieving the core carbon emission optimization goals.

6. The method for optimized scheduling of shared energy storage in distribution substations with multiple flexible resources as described in claim 5, characterized in that, Acquire historical operating data and real-time power output of distributed photovoltaic power within the transformer area, and determine the current day and night photovoltaic power output curves through data fitting; A predictive algorithm is used to calculate the future photovoltaic power output increment curve to determine the fluctuation range and peak node of photovoltaic power output. The control cycle is divided into multiple regulation periods, and the benchmark price for shared energy storage charging and discharging, the time-of-use electricity price range for each distribution area, and the unit carbon emission cost are initialized. Construct a system decision model to quantify the correlation characteristics of temperature control load temperature regulation threshold, the time period distribution characteristics of electric vehicle charging station load, and the safe range of state of charge of shared energy storage battery. The system incorporates user-side objective functions and distribution area-side objective functions, wherein the user-side objective function includes electricity purchase cost, shared energy storage usage cost, and carbon emission cost, and the distribution area-side objective function includes peak load reduction and carbon emission control. The system sets constraints on photovoltaic power generation, flexible load and energy storage, user electricity purchase and sale, and implicit carbon emission. The photovoltaic power generation constraint requires the total photovoltaic output of the distribution area to match the load and energy storage charging and discharging power. The flexible load and energy storage constraint requires the shared energy storage state of charge to maintain a safe range. The device collects photovoltaic power generation data, load operation data, energy storage device data, and cost and carbon emission data in real time and transmits them to a collaborative computing platform. The collaborative computing platform performs compliance verification, low-carbon analysis, and optimization evaluation on the collected data, identifies abnormal data, and generates anomaly identification results. Based on the anomaly identification results and the system decision model, user-side optimization instructions and transformer area control instructions are generated. The user-side optimization instructions guide self-consumption and shared energy storage charging during peak photovoltaic periods. The transformer area control instructions adjust the temperature-controlled load rotation and shared energy storage charging and discharging plans for peak load periods. The user-side optimization command and the transformer-side control command are sent to the corresponding terminals for execution. After execution, the power generation, storage power and load power data are obtained and fed back to the collaborative computing platform. The collaborative computing platform verifies whether the data after execution meets the requirements of full photovoltaic consumption and compliance with constraints. The verification result is judged. If it does not meet the requirements, the instructions are iteratively adjusted and optimized to form the final scheduling scheme. Based on the final scheduling scheme, the shared energy storage charging and discharging price and load control strategy are updated, the optimization frontier curve corresponding to peak shaving cost and peak load is generated, and the comprehensive optimization target value of the transformer area is determined.

Images