基于用户分类的电动汽车充电站需求预测

Demand forecast of electric vehicle charging stations based on customer classification

阐述了电动汽车充电站的需求问题,根据某区域内电动汽车用户的性质将其分为规律性较强的A类用户、规律性一般的B类用户以及随机性C类用户。在历史数据的基础上,运用多元线性回归法预测出区域内未来某时间的汽车保有量,然后根据国家相关部门的政策情况、调查情况以及Bass扩散模型预测各类用户所占比例和充电需求。根据各类用户的数量、对充电时间的要求并考虑日常维护备用容量、特殊事件等因素对区域内的常规充电站和快速充电站的需求量进行了合理预测。算例分析结果表明该方法具有较强实用性。

In this paper, the demandof electric vehicle charging stations is described. According to electric vehicle customers are classified into three categories, among which Class A is of strong regular characters, the , Class B weak while Class C random. On the basis of historical data, the tenure in future is easily to be predicted with the applica-tion of multiple - linear regression. Then according to the relevant national policies, the investigation and the Bass diffusion model, the proportion and number of customers, the request of charging needs of all types of customers can be predicted. Finally, based on the charging time, the reserve capacity of routine maintenance, special events and other factors, a reasonable prediction of the results of the example show that the method conventional charging stations in the region is reasonably made. The enjoys strong practicality.