基于结构化数据的用电需求预测研究

Research on demand forecasting based on structured data

针对电力供需不平衡将会影响供用电质量,严重时会影响电力系统的稳定性问题,对用电需求进行了预测研究。随着智能电网技术的飞速发展,对采集的海量电力数据进行分析能够反映出电力负荷的需求情况,基于此,提出了电力需求预测的原理及其特点,并采用PCA方法,在海量电力数据中提取出用于电力需求预测的主要影响因子：宏观经济,电价,温度,湿度,作为用电需求预测的输入数据。对ARMA的数学模型进行推导分析,并建立了基于ARMA的用电需求预测模型,给出了相应的预测步骤。最后根据某市的用电信息情况,进行用电需求预测的实验仿真,仿真结果表明,在总需求预测上ARMA方法预测误差均小于1%,相比于BP和SVM算法,ARMA方法在综合考虑预测精度和预测时间上,优势明显。并开发了相应的ARMA预测模型系统,该系统在某市电力公司运行效果良好,功能稳定,为电力公司制定相应的用电决策提供了依据。

In view of the imbalance of power supply and demand,the power supply quality will be caused,and the stability of the power system will be affected seriously.This paper predicts the demand for electricity.With the rapid development of smart grid technology,the analysis of large amount of power data collected can reflect the demand of power load.Based on this,the principle and characteristics of power demand forecasting are put forward.And using PCA method,in the massive power data,used in electric power demand forecasting is analyzed with the main influencing factor for the macro economy,electricity,temperature,humidity,t he input data for electricity demand forecasting.The mathematical model of ARMA is derived and the prediction model based on ARMA is established.The corresponding prediction steps are given.Finally,according to the electricity information situation of a city,the experiment simulation of demand prediction is carried out.The simulation results show that the prediction error of ARMA method is less than 1% in total demand forecasting.Compared with BP and SVM algorithm,the ARMA method has obvious advantages in comprehensive consideration of precision and prediction time.The corresponding ARMA prediction model system is developed and the system works well and is stable in a certain city.The method and system mentioned in this paper provide the basis for the power company to make the corresponding electric power decision.