基于关系探索和KTBoost的暂态稳定裕度评估  

Transient Stability Margin Assessment Based on Relationship Exploration and KTBoost

作　　者：王强 刘炼 陈浩 WANG Qiang;LIU Lian;CHEN Hao(College of Electrical Engineering and New Energy,China Three Gorges University,Yichang 443002,China)

机构地区：[1]三峡大学电气与新能源学院,宜昌443002

出　　处：《电力系统及其自动化学报》2022年第8期35-43,共9页Proceedings of the CSU-EPSA

基　　金：国家自然科学基金资助项目(52077120);国网江西省电力有限公司科技项目(5218F0180049)。

摘　　要：为了充分挖掘电网运行数据中的信息,进一步提高暂态稳定预测的精度,提出了一种基于关系探索和核-树联合提升KTBoost(combined kernel and tree boosting)的暂态稳定裕度评估方法。首先,考虑特征与暂态稳定裕度指标之间的相关性以及特征之间的冗余性和协同性,将信息论和施密特正交化相结合,以此筛选出关键电网运行特征,降低无关变量的干扰。然后,利用KTBoost算法建立关键特征量和暂态稳定裕度指标的映射关系,生成KTBoost驱动的暂态稳定裕度评估模型。最后,通过算例分析表明,所提方法不仅能够实现高精度的暂态稳定评估,且具有较强的鲁棒性和泛化能力。To fully mine the information in the power grid operation data and further improve the accuracy of transient stability prediction,a transient stability margin assessment method based on relationship exploration and the combined kernel and tree boosting(KTBoost)is proposed. First,considering the relationship between features and the transient stability margin index,as well as the redundancy and synergy between features,the information theory and the GramSchmidt orthogonalization are combined to select the key grid operation features and reduce the interference of irrelevant variables. Then,the KTBoost algorithm is used to construct a mapping relationship between key features and the transient stability margin index,and a transient stability margin assessment model driven by KTBoost is generated. Finally,the analysis of examples show that the proposed method can not only achieve the transient stability assessment with high accuracy,but also has a strong robustness and a strong generalization capability.

关 键 词：暂态稳定裕度 关系探索 信息论 施密特正交化 核-树联合提升 

分 类 号：TM712[电气工程—电力系统及自动化]