The Study Of Time Series Data Storage Management And Query Methodine-Healthcare Platform

This research work is a part of the home e-healthcare platform research project. It focuses on the establishing of data management system. The main research object of this research work is physiological time series data. The physiological time series data exists in the home e-healthcare platform research project. Analyzing and querying this kind of data is the base of data utilization. This research work mainly contributes on three aspects. Firstly, it established the main structure of physiological time series database which stores file pointers in the relational data table that pointed to the physiological signal file.And then it defined the similarity querying algorithm and implemented an extension of SQL which is based on the traditional relational database. It can efficiently find out the similar subsequence from a signal sequence in some certain similarity level. The matching method between given sequence and subsequence is average values methods and algorithm based on wavelet transform. It is easy to implement the similarity grading with these two methods.At last, it evaluated performance of this physiological time series data management system by testing three core parts, data compression ratio, table query performance and similarity query performance. The results show that this system works efficiently and accurately.