基于改进的插补扩散模型与LSTM的风电数据清洗方法
边文远,火久元,常琛

Wind power data cleaning method based on improved imputation diffusion model and LSTM
Wenyuan BIAN,Jiuyuan HUO,Chen CHANG
表 3 不同方法异常值识别效果对比
Tab.3 Performance comparison of different methods in outlier detection
方法总数据量正常数据量$ \varphi $/%$ {\rho }_{v{\text{-}}P} $$ {\rho }_{n{\text{-}}P} $
无(原始数据)6 3006 3000.000.93420.9423
LOF6 3006 0484.000.93940.9448
DBSCAN6 3005 8826.630.94390.9481
DBSCAN+IF6 3005 9216.010.94730.9522
IMDiffusion6 3005 9864.980.94790.9577
TranAD6 3005 9914.900.94730.9533
TimeADDM6 3005 9675.290.95010.9592
IDM6 3005 9545.490.95110.9604