Short-term Load Forecasting Based On Chaos Theory And Support Vector Machine

Power load forecasting has become one of the key factors affecting the development of power system with the development of industry and agriculture,the increase of electricity demand.The accuracy of power load forecasting directly affects the grid planning,scheduling and safe operation.Based on the combination of chaos theory and support vector regression machine,this paper analyzes the characteristics of power load time series to identify the chaotic characteristics of load sequence and builds the prediction model on the basis of phase space reconstruction.Firstly,on the basis of introducing the research background and significance of power load forecasting,the research contents of power load forecasting and the status quo at home and abroad are discussed.The definition,characteristics and principle of power load forecasting are discussed,and the steps of power load forecasting are expounded.Secondly,the classification and characteristic analysis index of power load are introduced to analyze the annual,monthly and daily load characteristics.In order to excavate the chaotic characteristics of time series of power load,the principle and reconstruction method of phase space reconstruction are discussed.Based on the phase space reconstruction,the chaos recognition of the power load time series is carried out according to the Poincaré section method and the maximum Lyapunov exponent method.Thirdly,through the introduction of support vector regression theory,the advantages of time series regression analysis are analyzed,and the chaos is combined with support vector regression method to avoid the premature local optimal solution and simplify the calculation step,And the power load time series for more accurate prediction.Finally,the gray relational analysis theory is used to analyze the influencing factors such as temperature and date type of power load time series.At the same time,the multivariate phase space reconstruction of the load time series is carried out by introducing the influence factor,and a multivariate chaos prediction model is constructed.Finally,the practical load sequence is analyzed by load forecasting on the working day and rest day,and the practicability of the multivariable chaos prediction model is verified.