针对最大信息系数计算与传统支持向量机进行电力系统动态频率预测时计算或训练时间过长、精度较低、模型泛化性能较差等问题,通过改进最大信息系数计算的网格划分方法,并改造原有SVR模型的不等式约束条件以及模型损失函数,提出基于统计信息系数与多输出最小二乘支持向量机(multi-output least squares support vector regreesion,MLSSVR)的频率响应曲线预测模型,实现对电力系统动态频率的特征提取、整体输出与链式预测,从而提升原有机器学习模型速度、精度和准确性。利用灰狼优化算法对MLSSVR算法中的核函数宽度以及惩罚因子进行寻优以提高模型综合性能。基于IEEE16机68节点的仿真实验证明,相较SVR算法及其他机器学习模型,本文所提模型的准确度、精度与训练预测速度均有提升,模型性能更强。
Abstract
关键词
Keywords
references
Mohammad Alizadeh,Turaj Amraee.Adaptive scheme for local prediction of post-contingency power system frequency[J].Electric Power Systems Research,2014.
Yan Xu,Yuanyu Dai,Zhao Yang Dong,Rui Zhang,Ke Meng.Extreme learning machine-based predictor for real-time frequency stability assessment of electric power systems[J].Neural Computing and Applications,2013.
David N. Reshef,Yakir A. Reshef,Hilary K. Finucane,Sharon R. Grossman,Gilean McVean,Peter J. Turnbaugh,Eric S. Lander,Michael Mitzenmacher,Pardis C. Sabeti.Detecting Novel Associations in Large Data Sets[J].Science,2011.
Tony van Gestel,Johan A.K. Suykens,Bart Baesens,Stijn Viaene,Jan Vanthienen,Guido Dedene,Bart de Moor,Joos Vandewalle.Benchmarking Least Squares Support Vector Machine Classifiers.[J].Machine Learning,2004.