Day-ahead optimization regulation and control decomposition strategy and intraday-real-time precise control method

Abstract

The invention discloses a day-ahead optimization regulation and control decomposition strategy and an intra-day-real-time precise control method, and the method comprises the steps: constructing a hierarchical control architecture of a distributed photovoltaic and novel energy storage aggregator participation frequency modulation market operation mechanism, and carrying out the frequency modulation of the distributed photovoltaic and novel energy storage aggregator in different control intervals; a day-ahead market energy-frequency modulation bidding model considering light and energy storage loss and a real-time market rolling optimization model considering a Nash bargaining party to realize power mutual aid and deviation penalty of all aggregators are obtained, and decision optimization of distributed photovoltaic and novel energy storage aggregators participating in the energy frequency modulation market in a full time scale is realized. According to the method, a joint operation framework of the aggregator cluster and the dispatching center is constructed, and day-ahead and intra-day real-time optimization models are combined, so that collaborative optimization configuration and joint clearing of the electric energy market and the frequency modulation auxiliary service market are realized, and the overall market income of each aggregator is maximized; the method is suitable for the fields of intelligent power grids, distributed energy management, auxiliary service markets, collaborative optimization of aggregator clusters and the like.

Description

Technical Field

[0001] This invention relates to a technology in the field of power dispatching, specifically a day-ahead optimization control decomposition strategy and intraday-real-time precise control method for distributed photovoltaic and novel energy storage aggregators considering frequency regulation services. Background Technology

[0002] Currently, new power systems based on new energy sources have become a new form of future development in the power energy sector. my country's installed capacity of new energy power generation has reached a considerable scale and continues to maintain a high growth rate. The high proportion of new energy sources such as photovoltaic and wind power, which are intermittent and uncertain, places higher demands on the operational flexibility of the power system, and the important role of energy storage in new power systems is receiving increasing attention. Research on market participant decision-making under the existing energy-frequency regulation market coordination mechanism has the following limitations: First, the research perspective mainly focuses on using energy storage to assist in the absorption of photovoltaic power, without fully utilizing energy storage, and lacks in-depth discussion on the additional revenue obtained by the charging and discharging behavior of energy storage in the electricity market under time-of-use pricing; second, few studies incorporate power interaction constraints between aggregators into the decision-making model, failing to fully reflect the overall synergistic effect of the power system. Existing research, when using cooperative game theory to optimize the common interests of the group, mainly focuses on maximizing overall benefits, but fails to fully consider the increase and decrease in energy storage losses of each participating entity during the cooperation process, resulting in a lack of refinement and fairness in the benefit distribution mechanism, thus making it difficult to effectively stimulate the initiative and enthusiasm of the cooperating entities. Summary of the Invention

[0003] To address the aforementioned shortcomings of existing technologies, this invention proposes a day-ahead optimization control decomposition strategy and an intraday-real-time precise control method. By constructing a joint operation framework between aggregator clusters and the dispatch center, and combining day-ahead and intraday-real-time optimization models, it achieves coordinated optimization allocation and joint clearing of the electricity market and frequency regulation ancillary service market, maximizing the overall market revenue of each aggregator. This invention is applicable to fields such as smart grids, distributed energy management, ancillary service markets, and coordinated optimization of aggregator clusters.

[0004] This invention is achieved through the following technical solution:

[0005] This invention relates to a day-ahead optimization decomposition strategy and an intraday-real-time precision control method. By constructing a hierarchical control architecture for the participation of distributed photovoltaic (PV) and new energy storage aggregators in the frequency regulation market, and then applying frequency regulation to PV and new energy storage aggregators under different control intervals, a day-ahead market energy-frequency regulation bidding model considering light and energy storage losses and a real-time market rolling optimization model considering Nash bargaining parties to achieve power mutual assistance and deviation penalties among aggregators are obtained. This enables decision optimization for PV and new energy storage aggregators participating in the energy frequency regulation market across the entire time scale.

[0006] Technical effect

[0007] This invention first constructs a hierarchical control architecture for the frequency regulation market operation mechanism involving distributed photovoltaic (PV) power and new energy storage aggregators, establishing the main functions of each level and clarifying the data connection relationships between them. Based on this hierarchical control architecture, a trading method for distributed PV power and new energy storage aggregators participating in the energy-frequency regulation market is proposed. The first stage is the day-ahead market bidding optimization stage, where distributed PV power and new energy storage aggregators submit their bidding output for each time period of the following day before the day-ahead market closes, based on the predicted PV output for the previous 24 hours and the predicted electricity price in the day-ahead energy-frequency regulation market. The second stage is the intraday-real-time market rolling optimization bidding stage, where distributed PV power and new energy storage aggregators revise their day-ahead bids based on the latest PV output forecast information from the PV power plant and the real-time market electricity price. An intraday-real-time rolling optimization model is constructed, considering the uncertainty of PV output, intraday deviation penalties, and energy storage operating costs, to achieve coordinated operation and closed-loop control with the day-ahead plan.

[0008] It effectively solves the problems of insufficient consideration of differentiated contribution and underutilization of energy storage in traditional models, significantly improves the overall economic benefits and cooperation stability of aggregator clusters, and achieves coordinated allocation of electrical energy and frequency regulation power under the constraint of power margin consistency principle, providing new theoretical support and practical guidance for multi-entity collaborative optimization in the energy-frequency regulation market. Attached Figure Description

[0009] Figure 1 This is a flowchart of the present invention;

[0010] Figure 2 This is a schematic diagram of a daily scrolling window for an example.

[0011] Figure 3 This is a schematic diagram of day-ahead power dispatch for an aggregator, as shown in the example.

[0012] Figure 4 This is a schematic diagram illustrating the day-ahead bidding strategy in the FM market;

[0013] Figure 5 This is an example of the day-ahead clearing results for frequency modulation capacity;

[0014] Figure 6 This example illustrates the planned SOC trajectory of a commercial energy storage system up to the day.

[0015] Figure 7 This is a schematic diagram illustrating an intraday bidding strategy in the FM market.

[0016] Figure 8 This example aggregates the intraday SOC trajectory of a commercial energy storage system.

[0017] Figure 9 This is a schematic diagram of the daily power dispatching of an aggregator, as shown in the example. Detailed Implementation

[0018] like Figure 1 As shown, this embodiment relates to a day-ahead optimization control decomposition strategy and an intraday-real-time precise control method, including:

[0019] Step 1: Construct a hierarchical control architecture for the frequency regulation market operation mechanism involving distributed photovoltaic (PV) and new energy storage aggregators. This architecture includes: a capacity assessment and reporting unit, a dispatch instruction issuance unit, a frequency regulation and energy reporting unit, and an energy interaction and intraday rolling execution unit. Specifically: The capacity assessment and reporting unit processes aggregator information to obtain adjustable capacity assessment results. This information includes at least: distributed PV predicted output, load baseline and planned power, energy storage rated power and rated energy, and the current energy storage SOC and its upper and lower limits. Based on the above information, the capacity assessment and reporting unit calculates the upper limit of available frequency regulation capacity for each aggregator during each intraday period and generates a capacity assessment report as the basis for subsequent dispatch and reporting. The dispatch instruction issuance unit processes the frequency regulation demand information issued by the power grid or dispatch center and the available capacity information reported by the capacity assessment and reporting unit to obtain dispatch instructions. The frequency regulation demand information represents the dispatch center's frequency regulation capacity requirements or execution requirements for aggregators. The dispatch instruction issuing unit outputs and distributes to aggregators: frequency regulation capacity instructions / allocation results for each time period and corresponding execution time information. The frequency regulation and electricity application unit processes the capacity assessment results from the capacity assessment and reporting unit and the dispatch demand information from the dispatch instruction issuing unit to obtain the electricity and frequency regulation market application results. The frequency regulation and electricity application unit only generates and submits the frequency regulation capacity application quantity for each time period, without price quotation. To ensure the absorption of new energy photovoltaics, the electricity application quantity is the winning bid quantity. At the same time, it outputs the participation status / execution quantity plan corresponding to the application quantity for subsequent intraday rolling execution. The energy interaction and intraday rolling execution unit processes the frequency regulation capacity instructions issued by the dispatch instruction issuing unit, the application quantity plan generated by the frequency regulation and electricity application unit, and the intraday updated forecast information, and generates 15-hour executable control quantities using a phased rolling optimization and mixed integer programming solution strategy. The output includes: the power grid's power purchase and sale and the execution control quantity of frequency regulation capacity. Specifically, the energy interaction and intraday rolling execution unit adopts a SOC cross-window inheritance mechanism: only the initial period of the first window is subject to an anchoring constraint of SOC = 0.1 times the energy storage capacity, and the end period of the last window is subject to a regression constraint of SOC = 0.1 times the energy storage capacity. The starting SOC of the remaining windows is taken from the end SOC of the previous window, and a boundary constraint of 0.1 times the energy storage capacity ≤ SOC ≤ 0.9 times the energy storage capacity is applied to the entire period within the window to ensure the continuous feasibility of SOC during intraday rolling execution.

[0020] Step 2: Frequency regulation of distributed photovoltaic and new energy storage aggregators under different control intervals, specifically including:

[0021] In the first phase, during the day-ahead phase, the electricity purchased and sold by distributed photovoltaic (PV) and energy storage aggregators in the electricity market are directly considered as market-cleared electricity. To improve local PV consumption and curb curtailment, a penalty term for PV declaration deviation is explicitly set in the optimization model. Based on a comprehensive consideration of PV output forecast information and energy storage charging and discharging characteristics, the aggregators collaboratively optimize their electricity purchase and sale bids and frequency regulation capacity bids. Under the goal of maximizing their own profits, they determine their day-ahead bid levels for participating in the electricity market and frequency regulation ancillary service market for each time period.

[0022] In the day-ahead market clearing process, with the objective function of minimizing the total clearing cost of intraday transactions, the day-ahead frequency regulation capacity declared by each aggregate is centrally optimized and cleared, thereby forming the corresponding day-ahead frequency regulation capacity clearing result.

[0023] In the second phase, the intraday phase, each aggregate bases its operations on the aforementioned day-ahead frequency regulation clearing results and its day-ahead power purchase and sale declarations. To strengthen the binding force of plan execution and the contractual constraints of market transactions, an intraday deviation penalty mechanism is further implemented in the model. Through a rolling optimization strategy, each aggregate solves for the intraday bidding capacity that maximizes its own revenue within each settlement interval. Since intraday rolling optimization has a higher time resolution (15 hours), and the accuracy of photovoltaic output prediction significantly improves as the operation time approaches, the optimal solutions for power purchase and sale and frequency regulation bidding obtained in this phase can be directly used as control and execution instructions for on-site operation, thereby achieving a close coupling between market mechanisms and physical dispatch.

[0024] Step 3: Modeling intraday deviation penalties and energy storage operating costs, specifically including:

[0025] 3.1 Optimized Control Decomposition Strategy: Based on real-time forecast data uploaded by distributed photovoltaic and new energy storage companies, and aiming to maximize the revenue of aggregators participating in the energy and frequency regulation ancillary services market, the day-ahead bidding capacity of distributed photovoltaic and new energy storage aggregators in the energy and frequency regulation ancillary services market for each time period is determined, specifically including:

[0026] 3.1.1 The objective function is constructed as follows: ,in: , Photovoltaic and energy storage aggregator Time period The day-to-day revenue from the electricity market. Photovoltaic and energy storage aggregator During the period The recent FM market revenue. Photovoltaic and energy storage aggregator Time period The cost of curtailing solar power in the current phase. Photovoltaic and energy storage aggregator Time period The day-ahead operating costs of energy storage (such as charging and discharging losses). and These refer to the day-ahead electricity sales price and electricity purchase price for the aggregator during the i-th time period t scenario. and These represent the electricity sales and purchases of aggregator i in the day-ahead electricity market during time period t. The timescale is 1 hour from the previous day. and They are photovoltaic and energy storage aggregators respectively Time period The current frequency regulation market's frequency regulation capacity price and frequency regulation mileage price. This is the proportional coefficient for the frequency modulation process. Photovoltaic and energy storage aggregator Time period The frequency modulation output of the application; This refers to frequency modulation performance indicators. The cost of penalties for abandoning light. Photovoltaic and energy storage aggregator Time period The recent forecast has been released. Photovoltaic and energy storage aggregator Time period Contribute to the application for photovoltaic power. This represents the energy storage loss coefficient. and They are photovoltaic and energy storage aggregators respectively Time period The discharge and charging power in the current electricity market.

[0027] Since each charge and discharge cycle of energy storage affects battery life, and considering the frequent and difficult-to-predict frequency regulation commands from the power grid, energy storage needs to respond to the frequency regulation signal in each control cycle through charge and discharge during actual frequency regulation operations. Therefore, it is set to... This is the frequency regulation power factor, representing the charging (discharging) amount of energy storage that occurs during actual operation for every 1MW of frequency regulation power provided. The energy is MWh, and the average value of the frequency modulation signal is approximately 0 within one control cycle, therefore it uses... This represents the average total charge and discharge volume of energy storage.

[0028] 3.1.2 Set the constraints for the objective function, including: a) Aggregate merchant purchase and sale of electricity constraints: , b) Photovoltaic output constraints: c) Energy storage operation constraints: d) Frequency modulation capacity constraints: e) Power balance constraints: , of which: Let A be a binary variable representing the electricity purchase / sale status of aggregator i in time period t (X). =1 indicates purchasing electricity or receiving electricity from the grid. =0 represents power sold to the grid or power transmitted to the grid. , These represent the maximum electricity purchase capacity and the maximum electricity sales (external transmission) capacity allowed by the aggregator, respectively. , They are photovoltaic and energy storage aggregators respectively Maximum energy storage charging and discharging power; , For the charge and discharge states of energy storage, there are 0-1 variables; Photovoltaic and energy storage aggregator Time period Energy storage SOC. Photovoltaic and energy storage aggregator Time period The current energy storage SOC. , : Photovoltaic and energy storage aggregators Energy storage charging and discharging efficiency; Photovoltaic and energy storage aggregator The lower limit of the energy storage SOC; Photovoltaic and energy storage aggregator The upper limit of the energy storage SOC; The SOC of aggregator i's energy storage as of time t; and These are the energy storage SOC settings for aggregator i at the start and end times, respectively, set to 0.1 times the energy storage capacity. The minimum frequency regulation capacity required by the market; Let i be the state variable indicating whether aggregator i participates in frequency regulation during the day-ahead time period t. It is a sufficiently large positive number.

[0029] 3.1.3 The day-ahead clearing model for the frequency regulation ancillary services market aims to minimize day-ahead transaction clearing costs. The objective function is set as follows: The constraints are: ,in: For time period The frequency modulation output recently won by aggregator i For a moment The frequency regulation capacity demand of the dispatch center is currently... For a moment The frequency regulation mileage demand of the dispatch center.

[0030] 3.2 Intraday - Real-time Precise Control Strategy: Based on ultra-short-term forecast data of distributed photovoltaic (PV) and new energy storage, and the day-ahead winning bid capacity of aggregators, a real-time bidding deviation penalty mechanism is used to determine the real-time bidding capacity of distributed PV and new energy storage aggregators in the energy market and frequency regulation ancillary service market for each time period, with the goal of maximizing the revenue of joint aggregators in participating in the real-time energy market and frequency regulation ancillary service market. Specifically, this includes:

[0031] 3.2.1 Distributed PV + New Energy Storage Intraday-Real-Time Coordination Bidding Strategy: A bidding deviation penalty mechanism is used to incentivize improved day-ahead forecast accuracy. The bidding deviation penalty cost includes energy market deviation penalties and frequency regulation market deviation penalties. Specifically, the energy market deviation penalty for PV and energy storage aggregator i during intraday time period t is... When photovoltaic and energy storage aggregator i purchases and sells electricity during the daytime period t Compared with the previously declared electricity purchase and sale volume When deviations occur, photovoltaic and energy storage aggregators will be penalized for deviations in the electricity market. Among them, the electricity purchased and sold by photovoltaic and energy storage aggregator i during the day-ahead period t was The electricity purchased and sold by photovoltaic and energy storage aggregator i during the daytime period t is , The penalty coefficient for aggregators' intraday output failing to meet the day-ahead bid volume in the energy market. The penalty factor for aggregators' intraday output exceeding the day-ahead bid volume in the energy market. For the aggregator, the electricity purchase and sale price during the period t of the day is... hour, , hour, , The timescale is 15 minutes within the day. The penalty for frequency regulation market deviation for photovoltaic and energy storage aggregator i during the intraday time period t is... When the frequency regulation output provided by photovoltaic and energy storage aggregator i during intraday period t Compared to the frequency modulation output that was won in the previous bid, When deviations occur, photovoltaic and energy storage aggregators will be penalized by frequency regulation market deviations. , The deviation penalty coefficient for aggregators' daily output failing to meet the bidding volume in the FM market; The deviation penalty coefficient for the intraday frequency modulation market bid volume of aggregators. and They are photovoltaic and energy storage aggregators respectively Time period The frequency regulation capacity price and frequency regulation mileage price in the intraday frequency regulation market.

[0032] 3.2.2 Constructing the objective function: ,in: Photovoltaic and energy storage aggregator Time period Intraday electricity market revenue. Photovoltaic and energy storage aggregator During the period Daily frequency modulation market revenue. This refers to the daily operating costs of energy storage (such as charging and discharging losses). Cost of abandoning light during the day. and These represent the electricity purchased and sold within the day and the electricity sold before the previous day, respectively. and The penalty for the deviation between the power supply and frequency regulation service of the aggregator i during the period t day. and These refer to the electricity sales price and purchase price for electricity during the same day in the scenario of time period t for aggregator i. and These represent the electricity sales and purchases of aggregator i in the intraday electricity market during time period t. For time period The frequency modulation output provided by the aggregator i within the day For time period Aggregator i Photovoltaic Unit's intraday power output forecast. We will contribute to the photovoltaic application process for aggregators. and These represent the discharge power and charging power of the aggregator's energy storage in the intraday electricity market, respectively.

[0033] 3.2.3 Set the constraints for the objective function, including: a) Photovoltaic output constraints: b) Charge / discharge power limits: , c) SOC dynamic equation constraints: d) SOC upper and lower limit constraints: e) Initial / Termination SOC Constraints: f) Frequency modulation capacity constraints: g) Power balance constraints: ,in: The SOC of aggregator i's energy storage within day t; and These are the energy storage SOC settings for aggregator i at the start and end times, respectively, set to 0.1 times the energy storage capacity. The minimum frequency regulation capacity required by the market; Let i be the state variable indicating whether aggregator i participates in frequency regulation during intraday period t. It is a sufficiently large positive number.

[0034] Step 4: Solve the model constructed in Step 3 using a phased rolling optimization and mixed integer programming solution strategy, specifically including:

[0035] 4.1 Phased Rolling Optimization Framework: The model described in step 3 is solved in phases: day-ahead optimization – intraday rolling optimization and execution. Specifically: In the day-ahead phase, the benchmark bidding and winning plans for each aggregate in the electricity / frequency regulation market are determined at an hourly granularity; in the intraday phase, at a 15-hour granularity, after obtaining the latest forecast information and operating status, the day-ahead plan is rolled over and re-optimized to generate executable scheduling instructions; during instruction execution, the intraday instructions are tracked and corrected online based on actual operating deviations and planned deviations, outputting final control quantities such as charging and discharging power, purchased and sold power, and frequency regulation capacity.

[0036] 4.2 Intraday scrolling window construction and SOC cross-window inheritance: such as Figure 2 As shown, the intraday time domain is discretized with a 15-hour time step, and the rolling step is set to 1 hour (4 15-hour periods), with a prediction time domain length of 4 hours (16 15-hour periods). When rolling to the end of the day, the prediction time domain is shortened to 3 hours / 2 hours / 1 hour respectively to cover the remaining periods. Each rolling window obtains a SOC trajectory, where only the initial period of the first window is subject to an anchoring constraint of SOC=0.1E, and the final period of the last window is subject to a regression constraint of SOC=0.1E. The starting SOC of the remaining windows is taken from the ending SOC of the previous window, realizing the continuous transfer of SOC between adjacent windows. At the same time, the operating boundary constraint of 0.1E≤SOC≤0.9E is applied to all periods within the window, where E is the rated energy capacity of the energy storage.

[0037] 4.3 Mixed Integer Programming Solution Strategy: Within each rolling window, the mutual exclusion of power purchase and sale, mutual exclusion of charging and discharging, and frequency regulation participation are modeled as 0-1 decision variables. Power balance, frequency regulation feasibility constraints, SOC dynamics, and boundary constraints are uniformly transformed into linear / piecewise linear constraints. The absolute value deviation penalty term in the objective function is linearized by introducing auxiliary variables, thus constructing a mixed integer linear programming subproblem. Each window is solved using a MILP solver (such as Gurobi or CPLEX). After obtaining the optimal control sequence for the current window, only decisions within the rolling step size are executed, and the final SOC is used as the initial SOC for the next window.

[0038] Through practical application experiments, the day-ahead optimization control decomposition strategy and intraday-real-time precise control method of this invention were simulated using Matlab in a Mac IM4 (hardware and software) environment, targeting the participation strategies of three distributed photovoltaic and energy storage aggregators in the day-ahead and intraday electricity markets. Each aggregator is equipped with independent photovoltaic modules and energy storage systems. Key parameters of the energy storage systems of each aggregator are shown in Table 1, and photovoltaic output is shown in Table 2. The core operating parameters and market pricing mechanism are as follows: the electricity price adopts the time-of-use (TOU) model. In the electricity market, the purchase price of electricity is 0.31 yuan / kW from 00:00–06:00 and 17:00–24:00, 1.07 yuan / kW from 07:00–10:00 and 16:00, and 0.64 yuan / kW from 11:00–15:00. The sales price of electricity is also set synchronously according to time of day, with the corresponding prices for the above time periods being 0.16 yuan / kW, 0.53 yuan / kW, and 0.32 yuan / kW, respectively.

[0039] Table 1. Key parameters of energy storage for each aggregator.

[0040] Table 2 Photovoltaic Power Output of Each Aggregator

[0041] Consider three distributed photovoltaic (PV) and energy storage aggregators, operating independently but jointly participating in the electricity market and frequency regulation ancillary services market. Each aggregator aims to maximize profits, leveraging its own day-ahead electricity and frequency regulation capabilities to rationally utilize energy storage. Their overall plan for participating in the electricity market is as follows: Figure 3 As shown, the day-ahead bidding strategy in the FM market is as follows: Figure 4 As shown, the corresponding day-to-day clearing results are as follows: Figure 5 As shown. The dispatching agency coordinates the power output and frequency regulation capacity of each aggregator at different time periods based on day-ahead electricity and frequency regulation demand to minimize total clearing costs. To ensure full utilization of energy storage, the state of charge (SOC) trajectory of each aggregator is shown below. Figure 6 As shown. After the day-ahead market clearing is completed, aggregators need to consider the deviation penalty relative to the day-ahead winning bid volume / capacity when submitting intraday energy and frequency regulation market quotations. Given the uncertainty of photovoltaic output and the day-ahead to intraday forecasting error, under the condition that the day-ahead bidding results are determined, each aggregator adopts a rolling optimization bidding strategy to jointly plan the energy and frequency regulation clearing to maximize intraday total revenue; specifically: the frequency regulation bid capacity for each bidding period is dynamically adjusted according to a 15-hour time scale, such as... Figure 7 As shown. To improve availability and reduce deviation risk, the energy storage SOC is configured in a synchronous rolling manner, and its evolution process is as follows. Figure 8As shown; the intraday collaborative planning results for the electricity market are as follows: Figure 9 As shown.

[0042] Under the constraints of frequency regulation and power deviation penalties, the frequency regulation command tracking accuracy of the three aggregators reached 94.48%, 94.62%, and 100.00%, respectively, and the photovoltaic absorption rate (non-curtailment rate) was 99.44%, 99.44%, and 99.99%, respectively. The results show that this invention can achieve high-quality renewable energy absorption while providing ancillary services, and ensure the safe and stable operation of each aggregator.

[0043] In summary, the technical effects of this invention include:

[0044] 1) Systematic Solution: This invention constructs a joint operation framework for distributed photovoltaic + new energy storage aggregator clusters and dispatch centers, providing a systematic solution for multi-entity interaction in the energy-frequency regulation collaborative market and improving overall operational efficiency.

[0045] 2) Differentiated contribution consideration: Through the day-to-day two-stage collaborative optimization model and the design of deviation penalty and frequency regulation performance index, the differentiated contribution of each member in the distributed photovoltaic + new energy storage aggregator cluster in the transaction of electricity and frequency regulation services is fully considered, and the resource allocation is optimized.

[0046] 3) Fair profit distribution mechanism: By embedding frequency regulation performance indicators and deviation penalty mechanisms into the revenue model, it is ensured that the revenue of the aggregator cluster is matched with the power and frequency regulation service quality it provides, thereby enhancing the stability and sustainability of the cooperation.

[0047] 4) System safety and economic operation in tandem: Based on the synergistic consideration of electrical energy and frequency regulation clearing results, power margin consistency constraints, and energy storage charging and discharging and SOC constraints, the safety and economy of the power system operation with the participation of aggregator clusters are guaranteed.

[0048] The above-described specific implementations can be partially adjusted by those skilled in the art in different ways without departing from the principles and purpose of the present invention. The scope of protection of the present invention is defined by the claims and is not limited to the above-described specific implementations. All implementation schemes within the scope of the claims are bound by the present invention.

Claims

1. A day-ahead optimization control decomposition strategy and intraday-real-time precise control method, characterized in that, By constructing a hierarchical control architecture for the participation of distributed photovoltaic (PV) and new energy storage aggregators in the frequency regulation market, and then applying frequency regulation to PV and new energy storage aggregators under different control intervals, a day-ahead market energy-frequency regulation bidding model considering light and energy storage losses and a real-time market rolling optimization model considering Nash bargaining parties to achieve power mutual assistance and deviation penalties for each aggregator are obtained. This enables decision optimization for the participation of distributed PV and new energy storage aggregators in the energy frequency regulation market across the entire time scale.

2. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, The planned volume for the previous 24 hours is discretized into a 96-point sequence over 15 hours to form an intraday rolling optimization that takes into account deviation penalties.

3. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, Discretize the intraday time domain with a time step of 15 hours, and set the rolling step size to a length of 4 hours; When rolling to the end of the day, the forecast time domain is shortened to 3 hours / 2 hours / 1 hour respectively to cover the remaining time period, and the forecast information is updated with each roll to achieve online correction.

4. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, Within each rolling window, the mutual exclusion of power purchase and sale, mutual exclusion of charging and discharging, and frequency regulation participation are modeled as 0-1 decision variables; power balance, frequency regulation feasibility constraints, SOC dynamics, and SOC boundary constraints are uniformly expressed as linear or piecewise linear constraints; and the absolute value deviation term in the objective function is linearized by introducing auxiliary variables, thereby constructing a mixed integer linear programming subproblem and solving it to obtain the optimal control sequence within the window, including power purchase, power sale, charging, discharging, and frequency regulation capacity.

5. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, The operating boundary constraint for SOC is set to 0.1E≤SOC≤0.9E; only the anchoring constraint of SOC=0.1E is applied to the initial period of the first window, and the regression constraint of SOC=0.1E is applied to the final period of the last window; the starting SOC of the remaining windows is taken as the ending SOC of the previous window, so as to realize the continuous transfer of SOC between adjacent windows, where E is the rated energy capacity of energy storage.

6. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to any one of claims 1-5, characterized in that, specifically include: Step 1: Construct a hierarchical control architecture for the frequency regulation market operation mechanism involving distributed photovoltaic and new energy storage aggregators, including capacity assessment and reporting units, dispatch instruction issuance units, frequency regulation and electricity energy reporting units, and energy interaction and intraday rolling execution units. Step 2: Frequency regulation of distributed photovoltaic and new energy storage aggregators under different control intervals, specifically including: In the first phase, during the day-ahead phase, the electricity purchase and sale volumes declared by distributed photovoltaic and energy storage aggregators in the electricity market are directly regarded as market-cleared volumes. To improve the local consumption level of photovoltaics and suppress curtailment, a photovoltaic declaration deviation penalty item is explicitly set in the optimization model. Based on a comprehensive consideration of photovoltaic output forecast information and energy storage charging and discharging characteristics constraints, the aggregators collaboratively optimize the electricity purchase and sale price and frequency regulation capacity price scheme. Under the goal of maximizing their own profits, they determine the day-ahead bidding price level for participating in the electricity market and frequency regulation ancillary service market for each time period. In the day-ahead market clearing process, with the objective function of minimizing the total clearing cost of intraday transactions, the day-ahead frequency regulation capacity declared by each aggregate is centrally optimized and cleared, thereby forming the corresponding day-ahead frequency regulation capacity clearing result. In the second stage, during the intraday phase, each aggregate, based on the aforementioned day-ahead frequency regulation clearing results and its day-ahead power purchase and sale declarations, further strengthens the binding force of plan execution and the contractual constraints of market transactions by implementing an intraday deviation penalty mechanism in the model. Through a rolling optimization strategy, the aggregate solves for the intraday bidding capacity that maximizes its own revenue within each settlement interval. Since intraday rolling optimization has a higher time resolution and the accuracy of photovoltaic output prediction is significantly improved as the operation time approaches, the optimal solutions for power purchase and sale and frequency regulation bidding obtained in this stage can be directly used as control and execution instructions for on-site operation, thereby achieving a close coupling between market mechanisms and physical dispatch. Step 3: Modeling intraday deviation penalties and energy storage operating costs, specifically including: 3.1 Optimize the decomposition strategy of regulation and control before the day: Based on the real-time forecast data uploaded by each distributed photovoltaic and new energy storage, determine the day-ahead bidding capacity of the energy and frequency regulation ancillary service market for distributed photovoltaic and new energy storage aggregators in each time period with the goal of maximizing the revenue of aggregators participating in the energy and frequency regulation ancillary service market. 3.2 Intraday - Real-time Precise Control Strategy: Based on the ultra-short-term forecast data of distributed photovoltaic and new energy storage and the aggregator's winning bid capacity in the day-to-day, a real-time bidding deviation penalty mechanism is used to determine the real-time bidding capacity of distributed photovoltaic and new energy storage aggregators in the energy market and frequency regulation market for each time period, with the goal of maximizing the revenue of aggregators participating in the real-time energy market and frequency regulation ancillary service market. Step 4: Solve the model constructed in Step 3 using a phased rolling optimization and mixed integer programming solution strategy.

7. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 6, characterized in that, The hierarchical control architecture of the frequency regulation market operation mechanism involving distributed photovoltaic and new energy storage aggregators includes: a capacity assessment and reporting unit, a dispatch instruction issuance unit, a frequency regulation and electricity energy reporting unit, and an energy interaction and intraday rolling execution unit. Specifically: the capacity assessment and reporting unit processes aggregator information to obtain adjustable capacity assessment results; the information includes at least: distributed photovoltaic predicted output, load baseline and planned power, energy storage rated power and rated energy, current energy storage SOC value and its upper and lower limits; the capacity assessment and reporting unit calculates the adjustable capacity of each aggregator at different times of the day based on the above information. The frequency regulation capacity upper limit is used to form a capacity assessment and reporting quantity as the basis for subsequent scheduling and application. The scheduling instruction issuing unit processes the frequency regulation demand information issued by the power grid or dispatch center and the available capacity information reported by the capacity assessment and reporting unit to obtain the scheduling instruction. The frequency regulation demand information is the frequency regulation capacity demand or execution requirement of the dispatch center to the aggregator. The scheduling instruction issuing unit outputs and issues to the aggregator: the frequency regulation capacity instruction / allocation result for each time period and the corresponding execution time period information. The frequency regulation and power energy application unit uses the capacity assessment result of the capacity assessment and reporting unit and the scheduling instructions of the scheduling instruction issuing unit to form the scheduling instruction. Demand information is processed to obtain the results of electricity and frequency regulation market bidding. The frequency regulation and electricity bidding units only generate and submit the frequency regulation capacity bidding quantity for each time period, without price quotation. To ensure the absorption of new energy photovoltaic, the electricity bidding quantity is the winning bid quantity. At the same time, the participation status / execution quantity plan corresponding to the bidding quantity is output for subsequent intraday rolling execution. The energy interaction and intraday rolling execution unit processes the frequency regulation capacity instructions issued by the dispatch instruction unit, the bidding quantity plan generated by the frequency regulation and electricity bidding units, and the intraday updated forecast information, and uses staged rolling optimization and mixed integer programming to solve the problem. The strategy generates 15 hours of executable control quantities; the output includes: the power purchase and sale of the power grid and the execution control quantities of frequency regulation capacity; among them: the energy interaction and intraday rolling execution unit adopts the SOC cross-window inheritance mechanism: only the anchoring constraint of SOC=0.1 times the energy storage capacity is applied to the starting period of the first window, the regression constraint of SOC=0.1 times the energy storage capacity is applied to the ending period of the last window, the starting SOC of the remaining windows is taken as the ending SOC of the previous window, and the boundary constraint of 0.1 times the energy storage capacity ≤ SOC ≤ 0.9 times the energy storage capacity is applied to the entire period within the window to ensure the continuous feasibility of SOC during intraday rolling execution.

8. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, Step 3.1 specifically includes: 3.1.1 The objective function is constructed as follows: ,in: , Photovoltaic and energy storage aggregator Time period The recent electricity market revenue, Photovoltaic and energy storage aggregator During the period The recent FM market revenue, Photovoltaic and energy storage aggregator Time period The cost of curtailing solar power in the current phase Photovoltaic and energy storage aggregator Time period The day-ahead operating costs of energy storage (such as charging and discharging losses), and These refer to the day-ahead electricity sales price and electricity purchase price under the aggregator's time period t scenario. and These represent the electricity sales and purchases of aggregator i in the day-ahead electricity market during time period t. The timescale is 1 hour from the previous day. and They are photovoltaic and energy storage aggregators respectively Time period The current frequency regulation market's frequency regulation capacity price and frequency regulation mileage price. This is the proportional coefficient for the frequency modulation process. Photovoltaic and energy storage aggregator Time period The frequency modulation output of the application; For frequency modulation performance indicators, To incur the cost of penalties for abandoning light, Photovoltaic and energy storage aggregator Time period The recent forecast has been effective. Photovoltaic and energy storage aggregator Time period Contributing to photovoltaic application efforts This is the energy storage loss coefficient. and They are photovoltaic and energy storage aggregators respectively Time period The discharge and charging power in the current electricity market; 3.1.2 Set the constraints for the objective function, including: a) Aggregate merchant purchase and sale of electricity constraints: , b) Photovoltaic output constraints: c) Energy storage operation constraints: d) Frequency modulation capacity constraints: e) Power balance constraints: , of which: Let A be a binary variable representing the electricity purchase / sale status of aggregator i in time period t (X). =1 indicates purchasing electricity or receiving electricity from the grid. =0 represents power sold to the grid or power transmitted to the grid. , These are the maximum electricity purchase capacity and maximum electricity sales (external transmission) capacity allowed by the aggregator, respectively. , They are photovoltaic and energy storage aggregators respectively Maximum energy storage charging and discharging power; , For the charge and discharge states of energy storage, there are 0-1 variables; Photovoltaic and energy storage aggregator Time period Energy storage SOC, Photovoltaic and energy storage aggregator Time period The current energy storage SOC, , : Photovoltaic and energy storage aggregators Energy storage charging and discharging efficiency; Photovoltaic and energy storage aggregator The lower limit of the energy storage SOC; Photovoltaic and energy storage aggregator The upper limit of the energy storage SOC; The SOC of aggregator i's energy storage as of time t; and These are the energy storage SOC settings for aggregator i at the start and end times, respectively, set to 0.1 times the energy storage capacity. The minimum frequency regulation capacity required by the market; Let aggregator i be the state variable indicating whether it participates in frequency regulation during the day-ahead time period t. It is a sufficiently large positive number; 3.1.3 The day-ahead clearing model for the frequency regulation ancillary services market aims to minimize day-ahead transaction clearing costs. The objective function is set as follows: The constraints are: ,in: For time period The frequency modulation output recently won by aggregator i For a moment The frequency regulation capacity demand of the dispatch center is currently... For a moment The frequency regulation mileage demand of the dispatch center.

9. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, Step 3.2 specifically includes: 3.2.1 Distributed PV + New Energy Storage Intraday-Real-Time Coordination Bidding Strategy: A bidding deviation penalty mechanism is used to incentivize improved day-ahead forecast accuracy. The bidding deviation penalty cost includes energy market deviation penalties and frequency regulation market deviation penalties. Specifically, the energy market deviation penalty for PV and energy storage aggregator i during intraday time period t is... When photovoltaic and energy storage aggregator i purchases and sells electricity during the daytime period t Compared with the previously declared electricity purchase and sale volume When deviations occur, photovoltaic and energy storage aggregators will be penalized for deviations in the electricity market. Among them, the electricity purchased and sold by photovoltaic and energy storage aggregator i during the day-ahead period t was The electricity purchased and sold by photovoltaic and energy storage aggregator i during the daytime period t is , The penalty coefficient for aggregators' intraday output failing to meet the day-ahead bid volume in the energy market. The penalty factor for aggregators' intraday output exceeding the day-ahead bid volume in the energy market. For the aggregator, the electricity purchase and sale price during the period t of the day is... hour, , hour, , The frequency regulation market deviation penalty for photovoltaic and energy storage aggregator i during the intraday time period t is calculated on a 15-minute timescale. When the frequency regulation output provided by photovoltaic and energy storage aggregator i during intraday period t Compared to the frequency modulation output that was won in the previous bid, When deviations occur, photovoltaic and energy storage aggregators will be penalized by frequency regulation market deviations. , The deviation penalty coefficient for aggregators' daily output failing to meet the bidding volume in the FM market; The deviation penalty coefficient for the intraday frequency modulation market bid volume of aggregators. and They are photovoltaic and energy storage aggregators respectively Time period The intraday frequency regulation market's frequency regulation capacity price and frequency regulation mileage price; 3.2.2 Constructing the objective function: ,in: Photovoltaic and energy storage aggregator Time period Intraday electricity market revenue, Photovoltaic and energy storage aggregator During the period Intraday frequency modulation market revenue, The daily operating costs of energy storage (such as charging and discharging losses), Intraday curtailment cost and These represent the electricity purchase and sale volume for the day and the day before, respectively. and The penalty for the deviation between the power supply and frequency regulation service of the aggregator i during the period t day. and These refer to the intraday electricity sales price and electricity purchase price under the aggregator's time period t scenario. and These represent the electricity sales and purchases of aggregator i in the intraday electricity market during time period t, respectively. For time period The frequency modulation output provided by the aggregator i within the day For time period Aggregator iPV unit's intraday power output forecast. Contribute to the photovoltaic application process for aggregators. and These represent the discharge power and charging power of aggregator energy storage in the intraday electricity market, respectively. 3.2.3 Set the constraints for the objective function, including: a) Photovoltaic output constraints: b) Charge / discharge power limits: , c) SOC dynamic equation constraints: d) SOC upper and lower limit constraints: e) Initial / Termination SOC Constraints: f) Frequency modulation capacity constraints: g) Power balance constraints: ,in: The SOC of aggregator i's energy storage within day t; and These are the energy storage SOC settings for aggregator i at the start and end times, respectively, set to 0.1 times the energy storage capacity. The minimum frequency regulation capacity required by the market; Let i be the state variable indicating whether aggregator i participates in frequency regulation during intraday period t. It is a sufficiently large positive number.

10. The day-ahead optimization control decomposition strategy and intraday-real-time precise control method according to claim 1, characterized in that, Step 4 specifically includes: 4.1 Phased Rolling Optimization Framework: The model described in step 3 is solved in phases: day-ahead optimization – intraday rolling optimization and execution. Specifically: the day-ahead phase uses hourly granularity to determine the benchmark bidding and winning plans for each aggregate in the electricity / frequency regulation market; the intraday phase uses 15-hour granularity, and after obtaining the latest forecast information and operating status, the day-ahead plan is rolled over and re-optimized to generate executable scheduling instructions; during instruction execution, the intraday instructions are tracked and corrected online based on actual operating deviations and planned deviations, outputting the final charging and discharging power, purchased and sold power, and frequency regulation capacity control quantity. 4.2 Intraday Rolling Window Construction and SOC Cross-Window Inheritance: As shown in Figure 2, the intraday time domain is discretized with a time step of 15 hours, the rolling step is set to 1 hour, and the prediction time domain length is 4 hours. When rolling to the end of the day, the prediction time domain is shortened to 3 hours / 2 hours / 1 hour respectively to cover the remaining time period. Each rolling window solves for a SOC trajectory. Only the initial time period of the first window is subject to an anchoring constraint of SOC=0.1E, and the final time period of the last window is subject to a regression constraint of SOC=0.1E. The starting SOC of the remaining windows is taken from the final SOC of the previous window, realizing the continuous transfer of SOC between adjacent windows. At the same time, the operating boundary constraint of 0.1E≤SOC≤0.9E is applied to all time periods within the window, where E is the rated energy capacity of the energy storage. 4.3 Mixed Integer Programming Solution Strategy: Within each rolling window, the mutual exclusion of power purchase and sale, mutual exclusion of charging and discharging, and frequency regulation participation are modeled as 0-1 decision variables. Power balance, frequency regulation feasibility constraints, SOC dynamics, and boundary constraints are uniformly transformed into linear / piecewise linear constraints. The absolute value deviation penalty term in the objective function is linearized by introducing auxiliary variables, thereby constructing a mixed integer linear programming subproblem. Each window is solved using the MILP solver. After obtaining the optimal control sequence for the current window, only decisions within the rolling step size are executed, and the SOC at the end of the execution is used as the initial SOC for the next window.

Images