Research And Application Of Power Demand Forecasting Technology For Intelligent City

The phenomenon of regional power supply imbalance in China is widespread.In order to change this phenomenon,it is necessary to accurately predict the power supply.Accurate power demand forecasting plays an important role in both power supply enterprises and power customers.It can make the power supply enterprises prepare for power generation,transmission and distribution in advance,upgrade the equipment and establish a reasonable annual power supply plan.What’s more,it can meet the demand of electricity customers well.At present,the accurate prediction of power demand is mainly through the improvement and optimization of the algorithm model,but the relevant data shows that:it is difficult to achieve high precision only by improving the algorithm,because the electric power demand forecasting value is not only related with the demand of historical data,it is also affected by various external factors,such as macroeconomic forecasting area population size and structure,economic structure,government policy and climate etc.Since 2009,the development of smart city in China has made the acquisition of cross department big data possible.In the thesis,the prediction and analysis of electricity demand not only refer to the internal data of power,but also refer to various external indicators that affect the demand of electricity.Among them,the short-term prediction of electricity demand mainly refers to the weather factors such as temperature,humidity and rainfall.The medium and long term prediction of electricity demand takes into account the external factors such as population,legal person and macro economy.In the prediction analysis,these impact indexes are used together with the historical data of power demand as the input of the model,and the future power demand is predicted and analyzed.For the selection of prediction model algorithm,the BP neural network optimized by genetic algorithm is used to optimize the initial weight and threshold of the BP neural network by genetic algorithm.In the long term prediction of power demand is divided into two kinds of methods of analysis,the first one is the power demand of the city as a whole are predicted using partial least squares regression method as the prediction model,the model can make full use of the input and output data of the model,establish the corresponding regression equation,to achieve accurate forecasts of electricity demand for second;the hierarchical forecast for the region,and then classified according to the different types of electricity customers,to achieve the power demand of a partition of an electricity customers accurate analysis,and then according to all kinds of electricity customers accounted for electricity and the area accounted for regional prediction of total electricity consumption proportion change,determine the relevant ratio coefficient the integration of regional electric power demand forecasting model.