基于随机森林的电网实时运行风险评估方法

Real-time risk assessment method for power grid operation based on random forest

为解决传统电网实时运行风险评估模型复杂、流程耗时等问题,提出了一种基于随机森林的电网实时运行风险评估方法,通过挖掘电网运行特征实现对运行风险的评估。介绍了随机森林的基本概念和实施要点。结合电网实时运行风险评估问题实际,构建了运行风险相关因素库,利用历史数据对随机森林智能体训练,实现了对电网实时运行风险等级的高效辨识。基于我国某省电网实际数据构造算例,通过对过去1年的历史数据挖掘分析,所训练得到的运行风险智能体预测准确率能够达到97%以上,评估耗时在2分钟以内,能有效满足电网实时运行风险评估的要求。

In order to solve the problems such as the complexity of the traditional real-time operation risk assessment model and the time consuming of the process,a real-time operation risk assessment method based on random forest is proposed.The basic concept and implementation of random forest are introduced.Combined with the actual situation of risk assessment of real-time operation of power grid,the database of operational risk related factors is constructed,and the random forest agents are trained with historical data,so as to realize the efficient identification of risk level of real-time operation of power grid.Based on the actual data construction example of a power grid in a certain province of China,through the historical data mining analysis of the past year,the trained prediction accuracy of operational risk agent can reach more than 97%,and the evaluation time is less than 2 minutes,which can effectively meet the requirements of real-time operation risk assessment of power grid.