CN-121983994-A - Electric power energy storage day-ahead scheduling method

CN121983994ACN 121983994 ACN121983994 ACN 121983994ACN-121983994-A

Abstract

The invention relates to the technical field of power system optimization operation and discloses a power energy storage day-ahead scheduling method which comprises the following steps of collecting power data used for inputting a mixed integer programming model, wherein the power data comprises next-day spot market discharge clear electricity price and historical load data, and calculating a next-day maximum demand constraint value according to the historical load data The method comprises the steps of constructing a mixed integer planning model, solving the constructed mixed integer planning model, generating a next-day scheduling plan and sending the next-day scheduling plan to an energy storage management module if solving is successful, giving an alarm if solving is unsuccessful, executing the next-day scheduling plan by the energy storage management module, monitoring real-time demand in the execution process of the next-day scheduling plan on an operating day, updating input information of the mixed integer planning model by using the real-time demand, and combining the real-time demand with input information of the mixed integer planning model And (5) performing comparison. The invention solves the problems that the high-efficiency and accurate energy storage scheduling is difficult to realize in the prior art.

Inventors

Yuan Jushigang
Hu Xiabo
LU JINFENG
ZHANG JIE
TANG GUOLI
Zhou Bianchi
LI ZHANYU
WU XINGYU

Assignees

华润电力(湖北)销售有限公司

Dates

Publication Date: 20260505
Application Date: 20251203

Claims (10)

1. The power energy storage day-ahead scheduling method is characterized by comprising the following steps of: Collecting power data for inputting a mixed integer programming model, wherein the power data comprises next-day spot market electricity clearing price and historical load data; Calculating a next-day maximum demand constraint value according to the historical load data , The expression of (2) is: ; In the formula, Represents the average value of the historical same month maximum demand with the same current month, Representing the maximum demand average value of the set time period before the current month, and μ representing the weight coefficient; Constructing a mixed integer programming model, wherein the input information of the mixed integer programming model comprises power data and maximum demand constraint, and the objective function of the mixed integer programming model is to minimize the total electric quantity and electricity charge of the operation day; Wherein the expression of the maximum demand constraint is: ; In the formula, t represents a time number, T represents the total number of times of the scheduling period, The power purchased by the power grid at the time t is represented, Representing a next-day maximum demand constraint value; The method comprises the steps of establishing a mixed integer programming model, solving the established mixed integer programming model, generating a next-day scheduling plan and sending the next-day scheduling plan to an energy storage management module if the solution is successful, and sending an alarm if the solution is failed, wherein the next-day scheduling plan comprises one or more of energy storage and charging power at each moment, discharge energy storage power at each moment, power grid purchase power at each moment and energy storage charge state at each moment; The energy storage management module executes the next day scheduling plan, monitors real-time demand in the execution process of the next day scheduling plan on the running day, updates input information of the mixed integer programming model by using the real-time demand, and compares the real-time demand with the input information of the mixed integer programming model And (3) comparing, namely triggering intervention logic to adjust the real-time demand to be less than or equal to the set threshold value if the real-time demand is greater than the set threshold value.
2. The method of claim 1, wherein the expression of the objective function of the mixed integer programming model is: ; In the formula, The time interval is represented by a time interval, And the current market price of electricity discharged the next day before the time t is shown.
3. The method of claim 2, wherein the input to the mixed integer programming model further comprises one or more of the following daily load power predictions Predicted value of photovoltaic power generation power of next day Rated capacity of energy storage Maximum charge power of stored energy Maximum discharge power of stored energy Energy storage and charging efficiency Energy storage discharge efficiency Minimum state of charge of stored energy Maximum state of charge of stored energy 。
4. A method of day-ahead scheduling for power storage according to claim 3, wherein the input information of the mixed integer programming model further comprises a power balance constraint, and the power balance constraint is expressed as: ; In the formula, The energy storage discharge power at the time t is represented, And the stored charge power at the time t is indicated.
5. The method for day-ahead scheduling of power storage according to claim 3, wherein the input information of the mixed integer programming model further comprises a dynamic energy storage constraint, and the expression of the dynamic energy storage constraint is: ; In the formula, Representing the energy storage charge state at the end of the time t; Wherein: the boundary conditions of (a) are expressed as: , ; In the formula, The state of charge of the stored energy at the beginning of the day is indicated, Representing a state of charge at the end of the scheduling period; The expression of the update equation of (c) is: ; In the formula, The time interval is represented by a time interval, Represents the stored state of charge at the end of time t-1, The energy storage discharge power at the time t is represented, And the stored charge power at the time t is indicated.
6. The method for day-ahead scheduling of power storage according to claim 3, wherein the input information of the mixed integer programming model further includes a hill climbing power constraint, and the expression of the hill climbing power constraint is: ; In the formula, Indicating the maximum rate of ascent and descent, Indicating the maximum ramp down rate of the vehicle, The energy storage discharge power at the time t is represented, The stored charge power at time t is indicated, The energy storage discharge power at the time t-1 is shown, And the stored charge power at time t-1 is indicated.
7. The method for day-ahead scheduling of power storage according to claim 3, wherein the input information of the mixed integer programming model further includes a maximum charge-discharge cycle number constraint, and the expression of the maximum charge-discharge cycle number constraint is: ; In the formula, The start variable of the charge-discharge cycle at time t, , Indicating the maximum number of full transitions of the energy storage charge-discharge state allowed per day.
8. The method for day-ahead scheduling of power storage according to claim 3, wherein the input information of the mixed integer programming model further includes a charge-discharge mutual exclusion constraint, and the expression of the charge-discharge mutual exclusion constraint is: , ; A binary decision variable representing the state of operation of the stored energy, , The energy storage discharge power at the time t is represented, And the energy storage charging power at the time t is represented, and M represents a variable coefficient.
9. A method of day-ahead scheduling for power storage according to claim 3, wherein the input information of the mixed integer programming model further comprises a power limitation constraint, and the power limitation constraint is expressed as: , ; In the formula, The energy storage discharge power at the time t is represented, And the stored charge power at the time t is indicated.
10. The method for scheduling the day before power storage according to any one of claims 1 to 9, wherein the method for scheduling the day before power storage is characterized by solving the constructed mixed integer programming model, generating a scheduling plan and sending the scheduling plan to an energy storage management module if the solving is successful, sending an alarm if the solving is failed, and comprising the following steps: Invoking a mathematical programming solver to construct a mixed integer programming model, wherein the maximum solving time and the optimal gap threshold of the mathematical programming solver are set before solving; The mathematical programming solver performs branch processing according to the solving state: If the mathematical programming solver successfully solves the feasible solution in the solving time less than or equal to the maximum solving time and in the gap less than or equal to the optimal gap threshold, extracting the optimal solution from the feasible solution, generating a next day scheduling plan and sending the next day scheduling plan to the energy storage management module; and if the solving time of the mathematical programming solver is greater than the maximum solving time and the solving is not successful, the feasible solution is obtained, and an alarm is sent.

Description

Electric power energy storage day-ahead scheduling method Technical Field The invention relates to the technical field of power system optimization operation, in particular to a power energy storage day-ahead scheduling method. Background The industrial user electricity charge commonly adopts a two-part electricity price structure, and the total electricity charge is jointly formed by the electricity charge calculated according to the electricity consumption and the basic electricity charge calculated based on the maximum monthly demand. Along with the construction and development of the power spot market, the electric quantity and electricity charge is determined by the market electricity price fluctuating in real time, which obviously increases the complexity of the cost optimization at the user side. Under the background, the use of an energy storage system for peak clipping and valley filling to reduce the total electricity charge has become a core requirement and an important means for industrial users. However, the existing energy storage system optimization scheduling method generally has a plurality of technical limitations, and is difficult to meet the application requirements in the current complex market environment. When the existing industrial user energy storage participates in the electric power spot market, three core bottlenecks of difficult minimization of total electricity cost and insufficient model solving efficiency and practicality under complex constraint caused by insufficient response to dynamic electricity price signals so as to limit arbitrage space, demand cost and electricity quantity cost and lack of cooperation in daily planning are faced. In China patent with publication number CN118537160A, an energy management method and system for an optical storage park based on maximum demand estimation, the energy management method for the optical storage park based on maximum demand estimation comprises the steps of collecting photovoltaic historical data and load historical data of the optical storage park, performing clustering analysis on the photovoltaic historical data and the load historical data through a clustering algorithm to determine photovoltaic-load typical days, dividing the photovoltaic-load typical days into a plurality of preset time periods, collecting the number of electric vehicles reaching and leaving the optical storage park in the time periods, updating a V2G resource pool of the optical storage park according to the number of the electric vehicles, constructing a rolling optimization model according to the photovoltaic-load typical days and the updated V2G resource pool, combining time-of-use electricity prices with two electricity prices, performing energy management solution through the rolling optimization model, and performing energy management according to the solving result. The electric automobile can improve new energy consumption and reduce park electric charge and create benefits for users in idle time while the automobile using experience is not affected. However, the method regards the demand control as a hard constraint which cannot be exceeded, but is not a decision variable which can be balanced and optimized with the electric quantity and electricity charge, lacks prospective collaborative optimization capability for the maximum demand, is difficult to realize the basic minimization of the total cost of electricity purchase, and can be in suboptimal dilemma of saving the electric quantity and electricity charge and increasing the higher electric quantity and electricity charge. In the Chinese patent with the publication number of CN118195211A, namely, an industrial park energy management-oriented method based on an improved NSGA-II and GA-BP combined algorithm, an optimized scheduling model based on an improved NSGA-II algorithm is realized based on a power prediction model based on a GA-BP algorithm, an industrial park micro-grid model and an optimized scheduling model based on an improved NSGA-II algorithm, and the optimized coordinated scheduling of the energy management of a commercial park is realized by taking the model with the characteristics of the industrial park as a core and combining predicted distributed energy power generation and load power. However, the optimal model is established on the basis of an electricity price curve formed by fixed moments such as peaks, flat, valleys and the like, and the model cannot effectively respond to dynamic price signals which are nodes in the spot market in 15 minutes or less. Although uncertainty is processed to a certain extent, the method is essentially a random search algorithm, has the problems of low convergence speed, unstable solving result and the like, and is difficult to ensure that a high-quality and executable day-ahead plan is obtained in the time of engineering requirements, so that the dual high requirements of spot market on decision speed and accuracy cannot be met. Disclosur