基于证据推理的电力负荷建模

Electric Load Modeling Based on Evidence Reasoning

从信息融合的角度提出一种新的电力负荷建模方法。首先,以证据推理算法为基本理论工具,在传统分类建模方法的基础上提出一种面向不同应用环境的电力负荷建模方法框架;其次,基于证据推理算法建立具体的电力负荷模型,通过对不同来源、不同层次、不同时间的电力负荷信息的融合利用,以更准确地描述负荷特性的复杂变化规律;再次,详细讨论了证据体的选择及基本信度分配等具体问题;最后,通过算例说明了所提方法在在线仿真分析和预测仿真分析两种应用环境下的有效性。

A novel electric load modeling method is presented from the view of information fusion. First, the framework of electric load modeling is given with the evidence reasoning algorithm as a basic theoretical tool, which not only develops the traditional classification and modeling method in a new way, but also satisfies the needs of simulations in various environments. Secondly, the electric load model is built in detail based on the evidence reasoning algorithm. With the model, information of electric load from various sources, levels, and times can be fully utilized and the complex load characteristics are described accurately. The specific links, e. g. the choice of evidences and the assignment of basic reliability, are also analyzed. The simulation results demonstrate its effectiveness in the online simulation and predictive simulation, respectively.