The management of dynamics objects is significant to VR systems, especially its data management, which plays an important role in the efficiency and reality of the systems. By separating and classifying the position, orientation, geometry structure and other relative attributes of moving objects into static and dynamic categories, different methods are used to process them. Generally, two steps are taken. First, we set up the static model of the objects, and then add dynamic attributes to them to make them dynamic. This paper pays more attention to dynamic attributes, especially spatial state time series, which may take various forms when stored in database. A compression algorithm, Hyper-Rectangle Compression (HRC), is introduced to avoid over dilating of the storage space with rapid growing time series data.