利索-全球专利检索

  1. 登录
  2. 注册
发明公开
CN108665108A 一种基于大数据的大区域用电量预测方法及系统 无效 - 驳回

一种基于大数据的大区域用电量预测方法及系统
摘要：
本发明涉及一种基于大数据的大区域用电量预测方法及系统，包括预测服务器、接口设备和数据库；所述预测服务器中的处理器首先获得参数向量的集合，接着获得参数向量的最优先行时间向量，然后获得最优先行参数和用电量之间的关联向量，最后获得下一个时间周期内大区域的预测用电量。本发明实时性强，并且充分考虑到了大区域用电量相对稳定的特点，将用电量看成包括基本不变量和可变量两者，使用用电量变化率间接预测用电量，使得用电量变化率的误差类似于用电量的误差，基本避免了不变量造成的干扰。
摘要（英）：

The invention relates to a big-data-based electric quantity predicting method and system for a large area. The system comprises a predicting server, interface equipment and a database; a processor inthe predicting server firstly obtains a set of parameter vectors, then obtains optimal preceding time vectors, then obtains incident vectors between optimal preceding parameters and the electric quantity and finally obtains the predicted electric quantity for the large area within the next time period. The big-data-based electric quantity predicting method and system are high in practicability, and fully considers the feature that the electric quantity of the large area is relatively stable, the electric quantity comprises fundamental invariables and variables, the electric quantity is indirectly predicted through the change rate of the electric quantity, so that the error of the change rate of the electric quantity is similar to that of the electric quantity, and interference caused by the invariables are basically avoided.

0/0