基于信度网络推理的风电监控数据缺失值恢复算法

Wind Power Monitoring Data Missing Value Recovery Algorithm Based on Belief Network

构建各级各类电力数据仓库是支撑电力系统数据平台建设的基础,针对电力调度数据仓库的风电监测数据的缺失问题,文章提出一种基于信度网络推理的风电监控数据缺失值恢复算法。首先根据数据集属性相关性构建信度网络,然后基于信度网络进行概率推理,最终找出最大概率下的候选值来对缺失记录进行恢复。通过使用MapReduce技术对某风电监控数据进行缺失值恢复实验,验证了算法的有效性。

Building various kinds of power data warehouses at different levels is the foundation of supporting the construction of power data platform.Aiming at the problem of wind power monitoring data missing in power dispatching data warehouse,this paper presents a method of filling missing wind power monitoring data based on belief network reasoning.Firstly,a belief network is constructed according to the relevance of attributes of data sets.Then,probabilistic reasoning is carried out based on the belief network.Finally,candidate values with the maximum probability are found to recover missing records.By using MapReduce technology to fill the missing values of wind power monitoring data,the validity of the method is verified.