Intelligent Analysis System in Time Series of Smart Health Home On-line Monitoring Data

Abstract

The conception of smart health home (SHH) is proposed in recent years with aging of population. Automatic processing of monitoring data becomes essential for SHH. Intelligent data analysis system based on autoregressive models (AR-models) is developed for SHH on-line monitoring data analysis. This system has three components, including AR-models identification, AR-models adjustment, and boundaries of the Prediction Interval (PI) computation. In this system, the order of AR-models is determined based on the Final Prediction Error (FPE) criterion, and then keep AR-models agreeing sufficiently well with the observed data. The parameters of AR-models are adjusted online based on adaptive filter algorithms, and then keep AR-models describe the true system of time series monitoring vital signs data. The vital signs data from PhysioBank biomedicine database are used for system test. The results proved that it can be used for vital signals data-processing on-line.