Search

CN-122001017-A - Day-ahead optimization regulation decomposition strategy and day-in-real-time precise control method

CN122001017ACN 122001017 ACN122001017 ACN 122001017ACN-122001017-A

Abstract

A day-ahead optimization regulation decomposition strategy and a day-in-real-time accurate control method are provided, after a layered control framework of a frequency modulation market operation mechanism is participated by a distributed photovoltaic and novel energy storage aggregator, the distributed photovoltaic and novel energy storage aggregator in different control intervals are subjected to frequency modulation, a day-ahead market energy-frequency modulation bidding model considering light and energy storage loss and a real-time market rolling optimization model considering Nash bargaining party to realize mutual power and deviation punishment of each aggregator are obtained, and decision optimization of the distributed photovoltaic and novel energy storage aggregator in the whole time scale of the energy frequency modulation market is realized. The invention realizes the collaborative optimization configuration and the joint clearing of the electric energy market and the frequency modulation auxiliary service market by constructing the joint operation framework of the aggregation business cluster and the dispatching center and combining the daily and daily-real-time optimization model, maximizes the overall market income of each aggregation business, and is suitable for the fields of intelligent power grid, distributed energy management, auxiliary service market, collaborative optimization of the aggregation business cluster and the like.

Inventors

  • YANG XINGANG
  • WANG DI
  • YANG ZHONGGUANG
  • LIANG WEIPENG
  • DU YANG
  • GUO LINGYU
  • WU SIMIN
  • DU SHUZE
  • AI QIAN
  • LI JIAMEI

Assignees

  • 国网上海市电力公司
  • 上海交通大学

Dates

Publication Date
20260508
Application Date
20260123

Claims (10)

  1. 1. A day-ahead optimization regulation decomposition strategy and a day-in-real-time precise control method are characterized in that after a layered control framework of a frequency modulation market operation mechanism is participated by a distributed photovoltaic and novel energy storage aggregator is constructed, the distributed photovoltaic and the novel energy storage aggregator are subjected to frequency modulation 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 Nash bargaining party to realize mutual power and deviation punishment of each aggregator are obtained, and decision optimization of the distributed photovoltaic and the novel energy storage aggregator under the whole time scale of the energy frequency modulation market is realized.
  2. 2. The day-ahead optimal regulation decomposition strategy and the intra-day-real-time precise control method according to claim 1, wherein the day-ahead 24-hour-level plan is discretized into 96-point sequences according to 15 hours, so as to form intra-day rolling optimization considering deviation penalty.
  3. 3. The day-ahead optimal regulation decomposition strategy and the intra-day-real-time precise control method according to claim 1 are characterized in that 15 hours are taken as time step discrete time domain in the day, a rolling step length is set to be 4 hours, when rolling to the end of the day, the predicted time domain is respectively shortened to 3 hours/2 hours/1 hour to cover the remaining time period, and the predicted information is updated during each rolling to realize online correction.
  4. 4. The daily optimization regulation and control decomposition strategy and daily-real-time precise control method according to claim 1 is characterized in that in each rolling window, electricity purchase mutual exclusion, charge-discharge mutual exclusion and frequency modulation participation states are modeled as 0-1 decision variables, power balance, frequency modulation feasibility constraint, SOC dynamic and SOC boundary constraint are uniformly expressed as linear or piecewise linear constraint, and an absolute value deviation term in an objective function is introduced into an auxiliary variable for linearization processing, so that a mixed integer linear programming sub-problem is constructed and solved, and an optimal control sequence comprising electricity purchase power, electricity sale power, charging power, discharging power and frequency modulation capacity in the window is obtained.
  5. 5. The day-ahead optimization regulation decomposition strategy and the day-in-real-time precise control method according to claim 1 are characterized in that an operation boundary constraint of 0.1E is set for SOC and is smaller than or equal to 0.9E, an anchoring constraint of SOC=0.1E is applied only to a first window starting period, a regression constraint of SOC=0.1E is applied to a last window end period, and the rest window starting SOCs take the last window end SOCs to realize continuous transmission of SOCs between adjacent windows, wherein E is the energy storage rated energy capacity.
  6. 6. The day-ahead optimal regulation decomposition strategy and the intra-day-real-time precise control method according to any one of claims 1 to 5, which are characterized by comprising the following steps: Step 1, constructing a hierarchical control architecture of a frequency modulation market operation mechanism, which comprises a capacity evaluation and reporting unit, a scheduling instruction issuing unit, a frequency modulation and electric energy reporting unit and an energy interaction and daily rolling execution unit; Step two, modulating frequency of the distributed photovoltaic and novel energy storage polymers under different control intervals, and specifically comprising the following steps: In the first stage, in the day-ahead stage, the electricity purchasing and selling quantity of distributed photovoltaic and energy storage aggregators declared in the electric energy market is directly regarded as the market output clear quantity, in order to improve the photovoltaic on-site consumption level and inhibit light rejection, a photovoltaic declaration deviation penalty item is explicitly set in an optimization model, the aggregation synergistically optimizes the scheme of electricity purchasing and selling price and frequency modulation capacity quotation of the electric energy on the basis of comprehensively considering photovoltaic output prediction information and constraint of energy storage charging and discharging characteristics, and under the goal of self income maximization, day-ahead bidding levels of participation in the electric energy market and frequency modulation auxiliary service market in each period are respectively determined; In the daily market clearing link, the total clearing cost of daily transactions is minimized as an objective function, and only the daily frequency modulation capacity declared by each polymer is subjected to centralized optimization and clearing calculation, so that a corresponding daily frequency modulation capacity clearing result is formed; In the second stage, in the intra-day stage, each polymer is based on the previous frequency modulation clearing result and the previous electricity purchasing and selling declaration thereof, and further through an intra-day deviation punishment mechanism in a model, the polymer solves the intra-day bidding quotation capacity which maximizes the self income in each settlement interval through a rolling optimization strategy, and as the intra-day rolling optimization has higher time resolution, and the photovoltaic output prediction precision is obviously improved along with the approach of the running time, the energy electricity purchasing and frequency modulation bidding quotation optimal solution obtained in the stage can be directly used as a control and execution instruction of on-site running, thereby realizing the tight coupling of the market mechanism and physical dispatching; step three, modeling through daily deviation punishment and energy storage operation cost, specifically comprising the following steps: 3.1 day-ahead optimization regulation decomposition strategy, namely determining day-ahead bidding capacity of the energy and frequency modulation auxiliary service market of the distributed photovoltaic and novel energy storage aggregators in each period according to real-time prediction data uploaded by the distributed photovoltaic and novel energy storage and aiming at the maximum profit of the energy and frequency modulation auxiliary service market of the aggregators; 3.2 According to ultra-short-term prediction data of distributed photovoltaic and novel energy storage and daily bid-in capacity of an aggregator, determining the real-time bid-in capacity of the distributed photovoltaic and the novel energy storage aggregator in the energy market and the frequency modulation market in each period by a real-time bid-in deviation punishment mechanism and with the aim of participating in the real-time energy market and the maximum profit of the frequency modulation auxiliary service market by the joint aggregator; and 4, solving the model constructed in the step 3 by adopting a staged rolling optimization and mixed integer programming solving strategy.
  7. 7. The day-ahead optimal regulation and control decomposition strategy and day-time accurate control method according to claim 6 is characterized in that the hierarchical control architecture of the frequency modulation market operation mechanism participated by the distributed photovoltaic and novel energy storage aggregators comprises a capacity assessment and reporting unit, a scheduling instruction issuing unit, a frequency modulation and electric energy reporting unit and an energy interaction and day-ahead rolling execution unit, wherein the capacity assessment and reporting unit processes according to the information of an aggregation provider to obtain an adjustable capacity assessment result, the information at least comprises distributed photovoltaic predicted output, load baseline and planning power, energy storage rated power and rated energy, current value of energy storage SOC and upper limit constraint thereof, the capacity assessment and reporting unit calculates the upper limit of available frequency modulation capacity of each aggregation provider in each time period based on the information and forms the basis of subsequent scheduling and reporting, the scheduling instruction issuing unit processes a scheduling instruction according to the demand information of a power grid or a scheduling center and the available capacity information on the capacity assessment and the frequency modulation unit, the scheduling instruction issuing unit processes the demand information of the frequency modulation and the frequency modulation, the scheduling unit sends out the demand information of the frequency modulation and the corresponding to the frequency modulation and electric energy reporting unit in each time period, the price of the scheduling unit sends out the demand information of the frequency modulation and the electric energy reporting unit in order to ensure that the new frequency modulation capacity is required to be distributed by the scheduling unit and the electric energy demand of the aggregation unit and the electric energy has no corresponding to the current value of the frequency modulation capacity of the frequency modulation and the energy reporting unit, and the current value of the frequency modulation capacity of the frequency modulation and the energy is distributed to be distributed by the energy reporting unit, and the current demand of the energy has the energy of the frequency modulation and the energy supply unit, and the current demand of the energy is distributed by the energy of the energy supply unit, the electric energy report quantity is the winning bid quantity, a participation state/execution quantity plan corresponding to the reporting quantity is output at the same time and used for subsequent daily rolling execution, an energy interaction and daily rolling execution unit processes according to a frequency modulation capacity command issued by a scheduling command issuing unit, a Shen Baoliang plan formed by the frequency modulation and electric energy reporting unit and prediction information updated in the day, a staged rolling optimization and mixed integer programming solving strategy is adopted to generate an executable control quantity for 15 hours, the execution control quantity comprising the purchase and sales quantity of a power grid and the frequency modulation capacity is output, wherein the energy interaction and daily rolling execution unit adopts an SOC cross-window inheritance mechanism, namely an anchoring constraint of SOC=0.1 times of energy storage capacity is only applied to a first window starting period, a regression constraint of SOC=0.1 times of energy storage capacity is applied to a last window end period, the rest window starting SOCs are used for obtaining the last window end SOCs, and a boundary constraint of SOC is applied to the whole time of SOC which is not more than 0.1 times of energy storage capacity and not more than 0.9 times of energy storage capacity in the window so as to ensure continuous feasibility in the daily rolling execution process.
  8. 8. The method for optimizing, regulating and decomposing strategy and real-time precise control in the day according to claim 1, wherein the step 3.1 specifically comprises: 3.1.1 constructing an objective function as: Wherein: , For photovoltaic and energy storage polymers Time period of Is a daily electric energy market benefit, For photovoltaic and energy storage polymers In the time period Is a daily frequency modulation market benefit, For photovoltaic and energy storage polymers Time period of The light-discarding cost at the early stage of (a), For photovoltaic and energy storage polymers Time period of Energy storage operation cost (such as charge and discharge loss) before the day, And The electricity selling price and the electricity purchasing price of the electric energy before the day under the scene of the period t of the aggregation business i are respectively, And The electricity selling power and the electricity purchasing power of the aggregation business i participating in the day-ahead electric energy market in the period t are respectively, For a time scale of 1 hour before day, And Photovoltaic and energy storage respectively Time period of Frequency-modulated capacity electricity price and frequency-modulated mileage electricity price of the future frequency-modulated market, For the scaling factor of the frequency modulation process, For photovoltaic and energy storage polymers Time period of The declared fm output; is the index of the frequency modulation performance, In order to discard the light penalty costs, For photovoltaic and energy storage polymers Time period of Is used for predicting the output of the machine, For photovoltaic and energy storage polymers Time period of Is used for reporting the output force of the photovoltaic system, As the coefficient of the energy storage loss, And Photovoltaic and energy storage respectively Time period of Discharge power and charge power in the electric energy market before date (a); 3.1.2 setting constraints on the objective function, including a) aggregating the commercially available electricity vending constraints: , b) photovoltaic output constraint: c) energy storage operation constraint: D) frequency modulation capacity constraint: e) power balancing constraints: Wherein lambda is Binary variable (lambda) for the purchase/sale status of the aggregator i during period t =1 Is either purchased or received from the grid, =0 Is selling electricity or sending power up to the grid); 、 Maximum purchase power and maximum electricity (delivery) power allowed by the aggregator, 、 Photovoltaic and energy storage respectively Maximum energy storage charge and discharge power of (2); 、 0-1 variable which is the charge and discharge state of energy storage; For photovoltaic and energy storage polymers Time period of Is used for storing the energy storage SOC of the battery, For photovoltaic and energy storage polymers Time period of Is used for storing the energy in the future, 、 Photovoltaic and energy storage polymer respectively Is a battery charging and discharging efficiency; For photovoltaic and energy storage polymers Lower energy storage SOC limit; For photovoltaic and energy storage polymers An upper energy storage SOC limit; SOC for stored energy by the aggregator i before time t day; And The energy storage SOC set values of the aggregation quotient i at the starting time and the ending time are respectively set to be 0.1 times of energy storage capacity, Minimum frequency modulation capacity required for the market; A state variable for whether the aggregator i participates in frequency modulation in a day-ahead period t; Is a sufficiently large positive number; 3.1.3 the day-ahead clearing model of the frequency modulation auxiliary service market takes the lowest day-ahead transaction clearing cost as an optimization target, and sets an objective function as follows: The constraint conditions are: Wherein: For a period of time Winning bid before day of the polymeric business i is used for the frequency modulation output of the transformer, For the moment of time The demand for frequency modulated capacity of the day-ahead dispatch center, For the moment of time Frequency modulation mileage requirement of day-ahead dispatch center.
  9. 9. The method for optimizing, regulating and decomposing strategy and real-time precise control in the day according to claim 1, wherein the step 3.2 specifically comprises: 3.2.1 The distributed photovoltaic and novel energy storage intra-day-real-time coordination bidding strategy comprises the steps of improving the daily prediction precision through a bidding deviation punishment mechanism, wherein bidding deviation punishment cost comprises energy market deviation punishment and frequency modulation market deviation punishment, and specifically comprises the following steps that the energy market deviation punishment of a photovoltaic and energy storage aggregator i in a intra-day period t is as follows When the photovoltaic and energy storage polymer i purchases electricity quantity in the time period t in the day Purchase and sales electricity quantity reported before date When the deviation occurs, the photovoltaic and energy storage polymer i will be punished by the deviation of the electric energy market Wherein the purchase and sales electricity quantity of the photovoltaic and energy storage polymer i in the day-ahead period t is The purchase and sales electric quantity of the photovoltaic and energy storage polymer i in the time period t in the day is , For the aggregate intra-day output to fail to meet the bias penalty factor for the energy market day-ahead scalar projection, A deviation penalty factor for aggregate intra-day output exceeding the energy market day-ahead scalar, Electricity purchase price for electric energy in period t days of the aggregator i, when In the time-course of which the first and second contact surfaces, , In the time-course of which the first and second contact surfaces, , For 15 minutes of time scale in the day, the market deviation penalty of the frequency modulation of the photovoltaic and energy storage polymer i in the time period t in the day is as follows Frequency modulation output provided by photovoltaic and energy storage polymer i during time period t Frequency modulation output of winning bid in the daytime before the day When deviation occurs, the photovoltaic and energy storage polymer i will be punished by frequency modulation market deviation , The method comprises the steps that a deviation penalty coefficient of a projected quantity of a frequency modulation market cannot be met for the daily output of an aggregator; The deviation penalty factor for the aggregate intra-day market bid amount, And Photovoltaic and energy storage respectively Time period of Frequency-modulation capacity electricity price and frequency-modulation mileage electricity price of the frequency-modulation market in the daytime; 3.2.2 Constructing an objective function: Wherein: For photovoltaic and energy storage polymers Time period of Is a daily electric energy market benefit, For photovoltaic and energy storage polymers In the time period Is a daily frequency modulation market benefit, For the operation cost (such as charge and discharge loss) in the energy storage day, The cost of discarding light at the daily stage is shown, And The power is purchased and sold in the day and before the day respectively, And (3) with Penalizing the deviation of the electric energy service from the frequency modulation service of the aggregate i in the period t day, And The electricity selling price and the electricity purchasing price of the daily electric energy under the scene of the period t of the aggregate i are respectively, And The electricity selling power and the electricity purchasing power of the electric energy market of the aggregation business i in the day in the period t are respectively, For a period of time The frequency modulated output provided by the aggregate merchant i during the day, For a period of time The daily predicted output of the photovoltaic machine set of the polymerization quotient i, For photovoltaic declaration of output in the aggregator, And Respectively storing discharge power and charging power of the energy stored by the polymerizer in an electric energy market in the day; 3.2.3 setting constraints of the objective function, including a) photovoltaic output constraints: B) charge-discharge power limit: , c) SOC dynamic equation constraint: d) SOC upper and lower limit constraints: e) initial/termination SOC constraints: f) frequency modulation capacity constraint: g) power balance constraint: Wherein: SOC for stored energy for aggregator i over time t day; And The energy storage SOC set values of the aggregation quotient i at the starting time and the ending time are respectively set to be 0.1 times of energy storage capacity, Minimum frequency modulation capacity required for the market; a state variable for whether the aggregator i participates in frequency modulation in the time period t in the day; is a sufficiently large positive number.
  10. 10. The method for optimizing, regulating and decomposing strategy and real-time precise control in the day according to claim 1, wherein the step 4 specifically comprises: 4.1, a staged rolling optimization framework, namely solving the model in the step 3 in stages according to daily optimization-daily rolling optimization and execution, wherein the daily stage uses hours as granularity to determine the standard declaration and the winning plan of each polymer in the electric energy/frequency modulation market, the daily stage uses 15 hours as granularity to roll and re-optimize the daily plan and generate an executable scheduling instruction after the latest prediction information and the running state are acquired, and the daily instruction is subjected to online tracking correction by combining the actual running deviation and the planning deviation in the instruction execution process to output the final charge and discharge power, the purchase and sale power and the frequency modulation capacity control quantity; 4.2 days of rolling window construction and SOC cross-window inheritance, wherein as shown in figure 2, 15 hours are taken as time step to discrete time domains in days, the rolling step length is set to be 1 hour, the predicted time domain length is set to be 4 hours, when rolling to the end of the day, the predicted time domain is respectively shortened to be 3 hours/2 hours/1 hour so as to cover the rest time period, each rolling window is solved to obtain a section of SOC track, wherein the anchoring constraint of SOC=0.1E is only applied to the initial time period of the first window, the regression constraint of SOC=0.1E is applied to the end time period of the last window, the rest window initial SOC is taken as the last window end SOC, continuous transmission of the SOC between adjacent windows is realized, and meanwhile, the running boundary constraint of SOC which is more than or equal to 0.1E and less than or equal to 0.9E is applied to all time periods in the window, wherein E is the energy storage capacity; 4.3, a mixed integer programming solving strategy is that in each rolling window, the mutual exclusion of electricity purchase and selling, mutual exclusion of charge and discharge and frequency modulation participation state are modeled as 0-1 decision variables, power balance, frequency modulation feasibility constraint, SOC dynamic and boundary constraint are uniformly converted into linear/piecewise linear constraint, linearization is realized by introducing auxiliary variables to absolute value deviation penalty items in an objective function, so that a mixed integer linear programming sub-problem is constructed, each window is solved by adopting an MILP solver, after the optimal control sequence of the current window is obtained, only decision in the rolling step length range is executed, and the executing end SOC is used as the initial SOC of the next window.

Description

Day-ahead optimization regulation decomposition strategy and day-in-real-time precise control method Technical Field The invention relates to a technology in the field of power dispatching, in particular to a day-ahead optimization regulation decomposition strategy and a day-in-real-time accurate control method for a distributed photovoltaic and novel energy storage aggregator to consider frequency modulation service. Background New power systems mainly using new energy have become new forms of future development in the field of electric power. The new energy power generation installed capacity of China has a scale, and the higher growth rate is kept. The high-proportion access of new energy sources such as photovoltaic, wind power and the like with intermittence and uncertainty brings higher requirements to the operation flexibility of a power system, and the important role of energy storage in the novel power system is increasingly focused. The research on the main decision of the market under the existing energy-frequency modulation market cooperative mechanism has the following limitations that firstly, the research view is mainly focused on the absorption of the photovoltaic by utilizing the energy storage assistance, the energy storage is not fully utilized, the deep discussion of obtaining extra benefit in the electric power market through the charge and discharge behaviors of the energy storage under the condition of time-sharing electricity price is lacking, and secondly, the research is very few, the power interaction constraint among the aggregators is incorporated into a decision model, and the integral cooperative effect of the electric power system cannot be fully reflected. When the community common interests are optimized by using the cooperative game theory, the conventional research is mainly focused on the maximization of the overall benefits, but the increase and decrease of the energy storage loss of each participating main body in the cooperation process cannot be fully considered, so that the benefit distribution mechanism lacks fineness and fairness, and the initiative and the enthusiasm of the cooperative main body are difficult to be effectively stimulated. Disclosure of Invention Aiming at the defects in the prior art, the invention provides a day-ahead optimization regulation and control decomposition strategy and a day-in-real-time accurate control method, and the cooperative optimization configuration and the joint clearing of an electric energy market and a frequency modulation auxiliary service market are realized by constructing a joint operation framework of an aggregator cluster and a dispatching center and combining day-ahead and day-in-real-time optimization models, and the overall market benefit 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 the aggregation business clusters and the like. The invention is realized by the following technical scheme: The invention relates to a day-ahead optimization regulation decomposition strategy and a day-in-real-time accurate control method, which is characterized in that after a layered control framework of a frequency modulation market operation mechanism is participated by a distributed photovoltaic and novel energy storage aggregator is constructed, the distributed photovoltaic and novel energy storage aggregator in different control intervals are subjected to frequency modulation, a day-ahead market energy-frequency modulation bidding model considering light and energy storage loss and a real-time market rolling optimization model considering Nash bargaining party to realize mutual power and deviation punishment of each aggregator are obtained, and decision optimization of the distributed photovoltaic and novel energy storage aggregator in the whole time scale of the energy frequency modulation market is realized. Technical effects The method comprises the steps of firstly constructing a layered control framework of a frequency modulation market operation mechanism participated by a distributed photovoltaic and novel energy storage aggregator, establishing main functions of each layer, determining data connection relations among the layers, and providing a transaction method of participated energy-frequency modulation market by the distributed photovoltaic and novel energy storage aggregator on the basis of the layered control framework, wherein the first stage is a day-ahead market bidding optimization stage, the distributed photovoltaic and novel energy storage aggregator predicts the capacity according to the photovoltaic of 24 hours in front of day and the electricity price prediction condition of the day-ahead energy-frequency modulation market, the bidding capacity of each time period of day is reported in advance before the day-ahead market transaction is closed, the second stage is a