Processing Method for Missing Data in Digital Twin System
摘要
The current digital twin system is extremely dependent on the integrity of the data when performing data analysis, and in the process of data collection and transmission, the phenomenon of data loss is very easy to occur. In this paper, the types of missing data are classified according to the size of missing data. For long-term data missing, it is processed in segments. For short-term data missing, the training of random forest algorithm model and the prediction of missing values are carried out according to different missing data sizes, And filled in the prediction using missing values from the wind power data twin system, which shows a good filling effect and helps to solve the problem of missing data filling in the digital twin system.