Several Electricity Load Forecasting Methods and Their Comparison

Abstract: Trend analysis, regression analysis, exponential smoothing, single consumption, gray model, load density and elasticity coefficient are introduced as methods of power load forecasting, and the annual electricity consumption of Zhuhai as an example is analyzed and compared in terms of conditions of applicability, data form, difficulty of computation and time of application. analysis method is suitable for large samples, and the past, present and future development patterns are consistent prediction, gray model method is suitable for poor information conditions of prediction; gray system theory using generator series modeling, regression analysis method, trend analysis method using raw data modeling, exponential smoothing method is through the combination of index weighting of the raw data to directly predict the future value of the; regression analysis and trend analysis method of the relatively simple calculation; single-consumption method, exponential smoothing method, the calculation difficulty and application time of these methods. Simple; single-consumption method, exponential smoothing method, gray model method is more suitable for near-term forecasting, regression method, trend analysis and improved gray model is more suitable for medium- and long-term forecasting.