Research On Forecast Of Urban Electricity Consumption

Today,the macroeconomics and power industry is developing rapidly.The domestic electricity market has entered a deepening reform stage.Power enterprise at all levels need to make predictions on the electricity demand market.Urban electricity consumption and load characteristic forecasting are the basic work in the electrical engineering.Relative to the short-term electricity demand forecast,mid-and long-term electricity demand forecast can be more reflects on the future direction of the national policy,the global economic situation,the direction of global energy resources cooperation and the environmental changes.While long-term electricity demand is more likely to predicts the power saturation.Therefore,this paper studies the city's mid-and long-term electricity demand.First of all,this paper collects previous economic performance and the electricity data among various departments of a city in central China and analyzes the macroeconomics,the adjustment of industrial structure,the development of critical user,the climates and other factors that have impacts on power demand and load characteristics.Aiming the several main factors,this paper established four respectively predictions models of the following: simple linear regression,industrial output value of unit consumption method,large user analysis and climates analysis.For the higher prediction accuracy,this paper has presented an optimal allocation from each individual model's constant weight obtained by the constrained optimization evolutionary algorithm.Constrained optimization evolutionary algorithm's search engine was designing by the Differential Evolution,and adding objective function's information to prevent population from getting into local optimums of the infeasible region,therefore,to achieve the purpose of balance constraint's conditions and objective functions.The variable weights acquired from adaptive recursive algorithm,turning the constant weight combinatorial model into the variable weight combinatorial forecast model.The establishedcombinatorial forecast model in this paper was more likely to reflect the influence of variation factors in the mid-and long-term forecasting on the electricity consumption.On the basis of the electricity consumption forecasting,this paper also analyzes the city's load characteristics.At the end of this paper,based on the city's "13th Five-Year Plan",this paper using the previous model to predict the electricity consumption and load characteristics of the whole society from 2015 to 2020,providing a reliable basis for this regional grid plan and power system peaking dispatch.