A Typical Scenario Generation Method For Renewable Energy In Power System Analysis

Compared with conventional energy sources,renewable energy sources such as wind power and photovoltaics have obvious randomness,volatility and intermittent characteristics.With the continuous increase of wind power and photovoltaic penetration rate,medium and long-term planning and operation and short-term scheduling of power systems The uncertainty of wind power and photovoltaic output must be considered.Power system scenario analysis is widely used because it can probabilistically describe uncertainties of wind power and photovoltaic output,but accurately characterizes the randomness and volatility of wind power and PV output,and how to solve large-scale scenarios that cause time for power system planning.The problem of high complexity is still an aspect that needs to be solved in current scenario analysis.Aiming at the above problems,this paper proposes a wind power and photovoltaic typical scene set generation method.The method is divided into two aspects: scene generation and scene reduction.Firstly,the wind power and photovoltaic generation scene generation technology based on predictive output is studied to accurately describe The uncertainty characteristics of wind power and photovoltaic power generation;secondly,a power time scene reduction method for time series analysis is studied to reduce and reduce the amount of wind power,photovoltaic and load scenarios and reduce the amount of calculation and reduce the amount of calculation.purpose.In this paper,the theory of power system scene analysis including wind power and photovoltaic is studied in detail,and the classical scene analysis methods are summarized.Secondly,in terms of scene generation,this paper focuses on a wind power and photovoltaic scene generation method based on unequal prediction box technology and multivariate standard normal distribution random number and inverse transform sampling.The method firstly establishes wind and photovoltaic historical statistics.The prediction error power prediction box of the data distribution characteristics,and the appropriate distribution function is used to fit the error data in the prediction box,and then the error data in the prediction box is inversely sampled by the multivariate standard normal distribution random number,and finally superimposed to Predicting the output to generate wind power and PV output scenarios,this paper verifies the rationality of the proposed method by using morphological analysis and related index evaluation.In terms of scene reduction,this paper studies a power timing scenario aggregation method based on improved spectral clustering and genetic algorithm.Firstly,based on the structural characteristics of different data of wind power and photovoltaic load,the optimized spectral clustering algorithm is used to construct the original data.The eigenvector matrix is used,and the preferred eigenvectors are clustered by genetic clustering optimization algorithm,and the final segmentation results are mapped back to the original data.In this paper,the classification results of the proposed scene algorithm are evaluated by standard time series,wind power,photovoltaic and load measured data,and the effectiveness of the proposed reduction method is verified.Finally,this paper applies the proposed scene reduction and generation algorithms to the optimization calculation of medium-and long-term deterministic planning operation and short-term random scheduling for large-scale wind power access,and compares and analyzes the relevant evaluation indicators between different algorithms.The application value of the proposed scene generation and reduction algorithm in practical engineering is verified.