摘要
针对电力相关数字文件数据融合方法的数据纠偏机制与数据的实时状况不相符,导致数据融合效率低、数据融合效果不佳的问题,提出了基于联合卡尔曼滤波方法的电力数字文件数据融合方法。所述方法从数字文件中获取数据源,按照数据的实时情况构建对应的数据纠偏机制,并对数据降噪处理实现数据的预处理,然后,对预处理后的数据进行特征提取并进行归一化处理,利用联合卡尔曼滤波算法对电力数字文件中的数据进行快速融合。仿真实验表明,相比于改进的BP神经网络的数字文件数据融合方法和基于离散小波分解与重构的数字文件数据融合方法,所述方法的平均融合效率达到402.3条/min,具有更好的计算和融合效率。
Aiming at the problem of low data fusion efficiency and poor data fusion performance due to the inconsistency between the constructed data correction mechanism and the real-time status of the data,this paper proposed a data fusion method for digital files of power system based on joint Kalman filter method.First,the data from the digital file and the real-time data are used to build data correction mechanism.Then a noise reduction calculation is used to pre-process the data.Based on normalization of the extracted data,the joint Kalman filter method is used to calculate the correlation,and realize the quick fusion of data.Simulation results show that compared to the improved BP neural network data fusion method and the discretized wavelet decomposition and reconstruction data fusion method,average fusion efficiency of the proposed method reaches 402.3 items/min,which shows that the proposed method has a higher data fusion speed and a better data fusion efficiency.
作者
肖建毅
蔡海滨
李宏亮
XIAO Jianyi;CAI Haibin;LI Hongliang(CSG Digital Enterprise Technology(Guangdong)Co.,Ltd.,Guangzhou,Guangdong 510000,China)
出处
《计算技术与自动化》
2025年第2期89-93,共5页
Computing Technology and Automation
