CN-119891377-B - Water, wind and light storage capacity planning method considering high-dimensional uncertainty

CN119891377BCN 119891377 BCN119891377 BCN 119891377BCN-119891377-B

Abstract

The invention belongs to the field of storage capacity planning, and provides a water-wind-solar storage capacity planning method considering high-dimensional uncertainty, which comprises the following steps: time correlation model construction, space correlation model construction, high-dimensional uncertainty scene set generation, water, wind and light storage capacity planning model construction, income matrix construction, multi-objective optimization function construction and Pareto front solution. The method generates the scene set covering high-dimensional uncertainty through the transition probability matrix C-VineCopula method, can comprehensively capture the space-time correlation and multi-dimensional random characteristics of natural conditions such as illumination intensity, wind power and the like, realizes ideal balance between economical efficiency and safety targets by introducing the Utobang line optimization strategy, improves the efficiency and quality of multi-target optimization solution, and provides diversified balance decision support for the system.

Inventors

ZHAO WENCHENG
LIU HONGXU
Gu Faying
YANG XUEWEI
WANG LI
YUAN QIANG
LI YANG

Assignees

国家能源集团青海电力有限公司
国能信控技术股份有限公司

Dates

Publication Date: 20260512
Application Date: 20241226

Claims (7)

1. A water, wind and light storage capacity planning method considering high-dimensional uncertainty is characterized by comprising the following steps of: Dividing the water-wind-solar historical data into a plurality of state intervals according to the fluctuation condition of month data, constructing a transition probability matrix according to the transition frequency of the variable in each month in different state intervals, and integrating all the transition probability matrices to obtain a time correlation model; fitting the edge distribution of each variable in the water-wind-solar historical data by using a non-parameter kernel density estimation method to obtain a cumulative distribution function; Calculating Kendall correlation coefficients of each variable in the water-wind-solar historical data, selecting the variable with the largest Kendall correlation coefficient as a root node of a rattan structure, describing the correlation between the root node and the residual variable in the water-wind-solar historical data through a binary Copula function to construct a first layer of the rattan structure, and recursively describing the interdependence relationship among the residual variables layer by layer to obtain a multi-dimensional joint probability density function; Fusing the cumulative distribution function, the multidimensional joint probability density function and the hybrid Copula function to obtain a spatial correlation model; Generating an initial scene set considering time correlation according to the time correlation model and the space correlation model, generating a first random variable meeting uniform distribution through cumulative probability of an initial scene corresponding to a variable with the largest sum absolute value of Kendall correlation coefficients in the initial scene set, and taking the first random variable as a first sample point; Generating a second random variable and a third random variable which meet uniform distribution through the accumulated probability, calculating a second sample point according to the first sample point and the second random variable, and calculating a third sample point according to the second sample point and the third random variable; mapping the first sample point, the second sample point and the third sample point to actual distribution by using an inverse transformation sampling method to obtain a high-dimensional uncertainty scene set; the method comprises the steps of constructing a water-wind-solar storage capacity planning model, wherein the water-wind-solar storage capacity planning model comprises an upper capacity configuration model and a lower capacity configuration model, the upper capacity configuration model comprises an upper objective function, a device sub-model and capacity constraint, and the lower capacity configuration model comprises a lower objective function, constraint conditions and power balance constraint; Converting the upper layer objective function and the lower layer objective function into standard forms, constructing a benefit matrix according to the converted upper layer objective function and lower layer objective function, and carrying out normalization processing on the benefit matrix; Extracting a Utropina line according to the normalized benefit matrix, calculating the distance between a target point in a standardized space and the Utropina line to obtain the Utropina line distance, and constructing a multi-objective optimization function according to the Utropina line distance; Optimally solving the multi-dimensional joint probability density function according to the high-dimensional uncertainty scene set and the water-wind-solar storage capacity planning model to obtain a Pareto front, and updating the Pareto front into a target system control strategy; the transition probability matrix is: ; Wherein, the ; A transition probability matrix is used as the transition probability matrix; And Respectively represent And A light state of the month; representing the status Transition to State Frequency of (2); is the total number of the state intervals; the benefit matrix is: ; Wherein, the Is the benefit matrix; And Respectively representing the converted upper layer objective function and the converted lower layer objective function; to optimize only Decision variables corresponding to the optimal solution; to optimize only Decision variables corresponding to the optimal solution; The multi-objective optimization function is: ; Wherein, the ; ; ; ; ; ; Representing coordinates of the target point; A normal unit vector representing the Utobang line; the normalized objective function vector is obtained; And Are all constraint conditions of optimization problems; The mixed Copula function is formed by weighting and superposing a plurality of Copula functions according to weight parameters, estimating the weight parameters and dependent parameters by adopting an expected maximization algorithm, generating a first random number which obeys zero to one uniform distribution for the current month illumination state when generating an initial scene set considering time correlation, comparing the first random number with corresponding row data of a current month accumulated frequency matrix to determine a next month illumination state, then generating a second random number which obeys zero to one uniform distribution, determining the illumination intensity of the next month in an illumination intensity interval corresponding to the determined illumination state according to a mode of adding a product of the second random number and the difference between the illumination intensity upper limit and the illumination intensity lower limit to the illumination intensity lower limit, circulating until the sequence generation of the illumination state and the illumination intensity of twelve months is completed, and repeating the preset times to obtain the initial scene set, and optimizing under different Ubbelopsis line normal unit vector combinations to obtain a plurality of different non-dominant solutions when solving the multi-objective optimization problem, and forming the non-dominant solution Paro by the non-dominant edge.
2. The method for planning the storage capacity of the water, the wind and the light according to claim 1, wherein the calculation formula of the Kendall correlation coefficient is as follows: ; Wherein, the ; The Kendall correlation coefficient is obtained; Representing a probability density function; Representing the total number of variables of the water-wind-solar historical data; And Respectively represent the first of the water, wind and solar historical data And (d) And a variable.
3. The method for planning the storage capacity of the water, the wind and the light according to claim 1, wherein the calculation formula of the second sample point is as follows: ; Wherein, the Is the second random variable; For the first sample point; Is the second sample point.
4. The method for planning a storage capacity of a water-wind-solar system taking high-dimensional uncertainty into consideration according to claim 1, wherein the upper objective function comprises: 、、 And ; Wherein, the To minimize total annual costs; Initial investment cost for equipment; The cost is maintained for the operation of the equipment in the later period; for upper-level electric network the income generated by electricity selling; Is a scene index; Is a scene Probability of (2); 、 And The constant annual value coefficients of wind power, photovoltaic and storage equipment are respectively; 、 And Investment cost of unit capacity of each device is respectively; 、 And Respectively, scenes The investment total capacity of each device; 、 And The unit operation and maintenance cost of each device is respectively; 、、 And Generating power of each device respectively; And The charging and discharging power of the storage system is respectively; Is that Time unit electricity price; And selling electricity to an upper power grid for the system.
5. A method of planning a storage capacity of a wind and water plant taking into account high dimensional uncertainty as defined in claim 1, wherein said plant submodel comprises: 、、 And ; Wherein, the Is photovoltaic power; The power is the hydroelectric power; efficiency as photovoltaic panel; Is the conversion coefficient of the photovoltaic panel; is the intensity of solar radiation; a photovoltaic panel temperature coefficient; Is that The ambient temperature at the moment; Reference temperature for the photovoltaic panel; Generating coefficients for the hydropower station; Is the flow of the generated water; The height of the power generation water head is; Is the runoff amount; Is the water discarding quantity; And Respectively correspond to Time of day and time of day The energy state at the moment; 、 charging power and discharging power respectively; 、 respectively a charging coefficient and a discharging coefficient; 、 A charge flag and a discharge flag, respectively.
6. The method for planning a storage capacity of a water-wind-solar system taking into account high-dimensional uncertainty as claimed in claim 1, wherein the lower objective function comprises: And ; Wherein, the For upper-level electric network expected value of electricity selling.
7. A method of planning a storage capacity of a wind and water plant taking into account high dimensional uncertainty as defined in any one of claims 4 and 5, wherein the power balance constraint is: 。

Description

Water, wind and light storage capacity planning method considering high-dimensional uncertainty Technical Field The invention relates to the field of storage capacity planning, in particular to a water-wind-solar storage capacity planning method considering high-dimensional uncertainty. Background With the increasing global energy demand and the increasing prominence of environmental problems, clean renewable energy is becoming the main stream of energy development, with hydropower, wind power and photovoltaics being of great concern. However, the renewable energy output characteristics show strong uncertainty, which can lead to fluctuation of the power grid frequency and instability of the voltage, and further affect the safety of the power system. Therefore, it is important to reasonably plan a high-proportion renewable energy access system and improve the economical efficiency and reliability of the system. Currently, a great deal of research has been conducted on system planning for high-proportion renewable energy access. The method comprises the steps of establishing an energy storage optimization model, analyzing the influence of time variation of photovoltaic and load on a power grid, providing a capacity configuration and optimization method of the storage system under high photovoltaic permeation, designing an integrated demand response program, discussing potential interaction capacity of electric power, gas, heat, cold and flexible loads of electric vehicles, providing an optimized scheduling method of a wind power-photovoltaic-gas-electric vehicle community comprehensive energy system, improving coordination among various energy converters by establishing a digital twin model of the comprehensive energy system and utilizing the prediction capacity of the digital twin, thereby improving energy efficiency, reducing cost and reducing carbon emission, and establishing a system planning method based on the game theory, so that the system not only meets the demands of cold, heat and electric loads, but also realizes optimal economical efficiency of the system. Although the above-described research has made some progress in system planning for high-rate renewable energy access, uncertainty in renewable energy has not been fully accounted for. The method comprises the steps of describing wind-light output uncertainty through a maximum uncertainty set, providing an optimal scheduling method for accounting for uncertainty and equipment coupling, providing a double-layer optimal configuration method by adopting an interval linear programming method to process uncertainty problems in consideration of wind-light output and uncertainty of electric coupling equipment, establishing an economic scheduling model of a comprehensive energy system by considering the uncertainty of photovoltaic output and considering economic cost and environmental protection and emission reduction, and providing an economic operation strategy by adopting a vector machine method to predict photovoltaic output in consideration of the influence of the uncertainty of photovoltaic output on system economy. However, the above-mentioned researches still do not fully consider the inherent time correlation of various weather conditions with seasonal variation in the system with high-proportion renewable energy access and the high-dimensional uncertainty caused by the spatial correlation of the weather conditions in the same area. The traditional experience planning method estimates the capacity of the equipment according to past experience, is simple to operate and low in early-stage cost, but lacks of uncertainty consideration, is easy to cause unreasonable in capacity configuration, and is high in long-term operation cost. The deterministic optimization planning method is based on deterministic data, does not consider uncertainty, is relatively simple to calculate, can be implemented when the data is limited, but has poor practical adaptability, cannot cope with wind and light resource fluctuation, and cannot balance economy and safety. The partial uncertainty considered planning method only considers partial uncertainty factors, the calculation complexity and the data requirement are between the two, and the method has certain adaptability, but the planning is not accurate enough because the high-dimensional uncertainty is not fully considered, and the system synergistic effect cannot be fully exerted. Disclosure of Invention In order to overcome the defects of the prior art, the invention aims to provide a water-wind-solar storage capacity planning method considering high-dimensional uncertainty, and solve the problems that the traditional method does not deeply process high-dimensional dependency, has insufficient depth for describing complex dependency and dynamic characteristics, lacks a scheduling optimization strategy of a system and omits multi-objective balance in optimal scheduling. In order to achieve the above object, the pres