CN-122026504-A - Water-wind-light output deficiency analysis method considering forecast uncertainty

CN122026504ACN 122026504 ACN122026504 ACN 122026504ACN-122026504-A

Abstract

The invention relates to the crossing field of clean renewable energy utilization and reservoir dispatching, and discloses a water-wind-light output deficiency analysis method considering forecast uncertainty. The method comprises the following steps of (1) representing wind and light forecast uncertainty based on a copula method and a K-means clustering method, (2) establishing a water and wind and light short-term optimization scheduling model considering insufficient output, solving by adopting a three-layer nesting method, and (3) carrying out insufficient output analysis according to scheduling results. According to the invention, the uncertainty characteristic of wind-light output in short-term prediction is considered, a set of short-term optimization scheduling model of the water-wind-light with insufficient output is constructed, the risk of the short-term scheduling of the water-wind-light with insufficient output is solved and analyzed, the relation between the generated energy and the insufficient output is revealed, the reliability is improved while the economy of the system is ensured, and the safe and stable operation of the water-wind-light complementary system can be effectively guided.

Inventors

WU DI
ZHOU HUA
LI DACHENG
NIE BINGBING
XIANG HUAWEI
LIU YUNHAN
ZHANG SHENGJIE
LIU PAN

Assignees

中国电建集团贵阳勘测设计研究院有限公司
华能澜沧江水电股份有限公司
武汉大学

Dates

Publication Date: 20260512
Application Date: 20251216

Claims (8)

1. The method for analyzing the insufficient water-wind-light output by considering the forecast uncertainty is characterized by comprising the following steps of: (1) Characterizing wind and light forecast uncertainty based on copula method and K-means clustering method; (2) Establishing a water-wind-solar short-term optimization scheduling model considering insufficient output, and solving by adopting a three-layer nesting method; (3) And carrying out insufficient output analysis according to the dispatching result.
2. The method for analyzing the insufficient water-wind-light output considering the forecast uncertainty as claimed in claim 1, wherein in the step (1), based on the predicted values of wind power and photovoltaic output before the day, a copula method is adopted to generate a plurality of groups of wind power and photovoltaic uncertainty output scenes, and a typical scene is reduced and extracted according to a K-means clustering method.
3. The method for analyzing insufficient water-wind power output considering prediction uncertainty as claimed in claim 2, wherein the method comprises the following steps: assuming that the relative prediction errors of wind power and photovoltaic output are subject to unbiased and normal distribution: ; wherein: representing a normal distribution; And Respectively representing the relative prediction errors of wind power and photovoltaic output; And Standard deviation of relative prediction errors of wind power and photovoltaic output are respectively represented; Adopting t-copula to construct the joint distribution of wind-light output relative prediction error: ; wherein: is a linear correlation coefficient, and ; Is a unitary distribution function Is the inverse function of (a), both degrees of freedom are 。
4. The method for analyzing the insufficient water-wind-light output considering the forecast uncertainty is characterized in that after a plurality of groups of wind power and photovoltaic uncertainty output scenes are obtained, a K-means clustering method is adopted to reduce and extract typical scenes, a K-means algorithm divides sample categories based on Euclidean distance, and an optimal clustering number is determined based on a Dyssenbut index, so that several wind-light output scenes and occurrence probability thereof are finally obtained.
5. The method for analyzing insufficient water-wind power output considering prediction uncertainty as claimed in claim 1, wherein in the step (2), The water-wind-solar short-term optimization scheduling model gives consideration to the generation benefit and the insufficient output risk, and the objective function is to enable the maximum generating capacity and the minimum insufficient output risk of the water-wind-solar complementary system on the premise of planning water consumption on a given day; the maximum system generating capacity is as follows: ; wherein: The daily actual net quantity of the water-wind-solar complementary system is represented; Representing the total number of wind-light output scenes; And Respectively representing the total time period number and the time period step length of short-term scheduling; Represent the first Complementary system in time period under wind-light output scene Internet surfing output; Represent the first Probability of occurrence of individual scenes; Representing the total number of hydroelectric generating sets; Represent the first Hydroelectric generating set under scene of wind-light output In the time period Is a force of the (a); And Respectively represent the first Wind power station and photovoltaic power station in time period under wind-light output scene Is a force of the (a); Represent the first In a period of time under a scene of wind-light output Is discharged from the battery; the risk of insufficient output is minimal: ; wherein: The total quantity of insufficient output of the water-wind-solar complementary system is represented; representing a period of time for a day-ahead power generation plan And (2) planned output of ; To facilitate efficient model solving, the insufficient output is converted into a "penalty function" construction objective function: ; wherein: representing the daily power generation capacity of the water-wind-solar complementary system considering the penalty term; Is a punishment coefficient of insufficient output.
6. The method for analyzing the water-wind-solar power output deficiency considering the forecast uncertainty as recited in claim 5, wherein the constraint conditions of the model comprise water quantity deviation constraint, water quantity balance constraint, load balance constraint, reservoir capacity constraint, power generation head constraint, hydroelectric generating set power characteristics, water level-reservoir capacity relation, tail water level-delivery flow relation, minimum start-stop constraint of the set, hydroelectric generating set output constraint and transmission capacity constraint.
7. The method for analyzing the insufficient water-wind-light output considering the forecast uncertainty, which is disclosed by claim 6, is characterized in that a three-layer nested solving method based on CSO and DP is adopted to solve a water-wind-light short-term optimizing scheduling model, and the solving process is divided into three parts, namely the following specific steps: ① The outer layer inputs a load characteristic curve and daily planned water quantity, and searches an optimal power generation plan based on a CSO algorithm; ② The middle layer inputs warehouse-in runoff, wind and light output forecast values and a system power generation plan, and the minimum starting platform and starting and stopping states of the unit in each period are optimized based on a CSO algorithm; ③ And the inner layer inputs a power generation plan and the minimum starting number, the optimal load distribution and the actual water consumption of the hydro-generator set are determined based on a DP algorithm, and finally the model solving is completed through iterative circulation.
8. The method for analyzing insufficient water-wind power output considering prediction uncertainty as claimed in claim 1, wherein the step (3) comprises the following steps: ① The correlation analysis of the generated energy and the insufficient output of the system comprises the steps of taking the occurrence probability of the insufficient output and the depth of the insufficient output as evaluation indexes, and setting different punishment coefficients to explore the relationship between the generated energy and the insufficient output; And (3) constructing an evaluation index: ; wherein: The probability of insufficient output is expressed as the proportion of the number of days with insufficient output to the total number of days; representing the number of time periods when the actual force is less than the planned force; The depth of the insufficient output is expressed as the proportion of the total amount of the insufficient output to the actual power generation; ② And analyzing the risk of insufficient output based on the historical long-sequence scheduling result.

Description

Water-wind-light output deficiency analysis method considering forecast uncertainty Technical Field The invention belongs to the crossing field of clean renewable energy utilization and reservoir dispatching, and relates to a water-wind-light output deficiency analysis method considering prediction uncertainty. Background Wind and light resources are affected by natural climate conditions, and the output has the characteristics of randomness, intermittence and fluctuation, and the direct incorporation of the wind and light resources into a power grid can not only aggravate the peak regulation pressure of the power grid, but also is unfavorable for the consumption of wind and light energy sources by the power grid. The water-wind-solar multi-energy complementary system can effectively cope with the instability of wind power and photoelectric output, and becomes an internationally recognized key measure for solving the problem of energy consumption. Benefits and risks are two major targets for making scheduling decisions of the water-wind-solar complementary system, and how to improve the power generation benefits while guaranteeing the reliability of the system becomes a hot research problem in the field of energy. Because wind and light output has stronger uncertainty in a short period, a power generation plan established in the day before can not be ensured in the day, and the risk of insufficient output is increased. At present, the risk research of the water-wind-solar complementary system pays much attention to the power-off or water-off risk, the analysis of the problem of insufficient output is less, the relation between the generated energy and the insufficient output of the system is not clear, and the safe and stable operation of the water-wind-solar complementary system is difficult to guide. Based on the method, the invention provides the water-wind-light output deficiency analysis method considering prediction uncertainty. Disclosure of Invention The invention aims to solve the problems that the existing water-wind-solar complementary system risk research pays attention to the power-off or water-off risk, the key risk of insufficient output is considered, the relation between the generated energy and the insufficient output of the system is not clear, and the system is difficult to be guided to develop short-term dispatching operation which takes the power generation benefit and the risk of insufficient output into consideration. Therefore, the invention provides the water-wind-light output deficiency analysis method considering the forecast uncertainty, a water-wind-light short-term optimization scheduling model considering the output deficiency is constructed, the risk of the output deficiency existing in the water-wind-light short-term scheduling is solved and analyzed, the relation between the generated energy and the output deficiency is revealed, and the economic and stable operation of the water-wind-light complementary system can be effectively guided. The technical scheme is that the method for analyzing the insufficient water-wind-light output by considering the prediction uncertainty comprises the following steps: (1) Wind and light forecast uncertainty is represented based on copula method and K-means clustering method The wind power and the photovoltaic output before the day have larger prediction errors, and generate representative wind power and photovoltaic output scenes, the method is based on the predicted values of the wind power and the photovoltaic output before the day, and generating a plurality of groups of wind power and photovoltaic uncertainty output scenes by adopting a copula method, and reducing and extracting typical scenes according to a K-means clustering method. Assuming that the relative prediction errors of wind power and photovoltaic output are subject to unbiased and normal distribution: ; wherein: representing a normal distribution; And Respectively representing the relative prediction errors of wind power and photovoltaic output; And And respectively representing standard deviations of relative prediction errors of wind power and photovoltaic output. Adopting t-copula to construct the joint distribution of wind-light output relative prediction error: ; wherein: is a linear correlation coefficient, and ;Is a unitary distribution functionIs the inverse function of (a), both degrees of freedom are。 And after obtaining a plurality of groups of wind power and photovoltaic uncertainty output scenes, reducing and extracting typical scenes by adopting a K-means clustering method. The K-means algorithm divides sample categories based on Euclidean distance, determines optimal clustering numbers based on the Dyson Bao Ding Zhishu (DBI), and finally obtains several wind-light output scenes and occurrence probability thereof. (2) Establishing a water-wind-solar short-term optimization scheduling model considering insufficient output, and solving by adopting a three-