摘要
针对电力市场随机性、多变量和时变性的特点导致电力客户服务需求预测值不准确的问题,提出了一种基于大数据分析的电力客户服务需求预测方法.该方法依托于贵州地区的智能电网大数据,从区域商业价值和区域宏观经济角度来采集数据并通过挖掘其中的关联信息,建立了电力客户的细分模型;并在客户细分模型的基础上,使用BP神经网络算法建立了电力客户的需求预测模型.在Matlab平台上的仿真与测试结果表明,所提出的方法能帮助电网公司更好地理解客户行为和服务需求,制定营销策略.
Aiming at the problem of inaccurate forecasting of power customer service demand due to the randomness,multivariate and time-varying characteristics of power market,a service demand forecasting method for power customers based on big data analysis was proposed.The as-proposed method relied on Guizhou smart grid big data,the data were collected from the perspective of regional business value and regional macroeconomy,and a subdivision model for the power customers was established by collecting related information.On the basis of customer subdivision model,a BP neural network algorithm was used to establish a demand forecasting model for power customers.The simulation and test results on the Matlab platform show that the as-proposed method can help the grid company to better understand customer behavior and service needs,so as to develop marketing strategies.
作者
朱州
ZHU Zhou(Information Center, Guizhou Power Grid Company of China Southern Power Grid, Guiyang 550000, China)
出处
《沈阳工业大学学报》
EI
CAS
北大核心
2020年第4期368-372,共5页
Journal of Shenyang University of Technology
基金
国家自然科学青年基金项目(61863113)。