Research of System Identification for Ni/MH Battery State of Charge Based on a Short Sequence and Multi-Sample Process

Abstract

In the battery state of charge of a systematic analysis, the observed data has a property that is the direction of time t (referred to as vertical) for a limited length, and number of samples obtained by N (called horizontal) for the infinite data set , it is called as a short sequence and multi-sample time series. By studying the characteristic of this time series, a new system identification method has been proposed, and the system identifiability for this process has been demonstrated. Through practice simulations, a satisfactory application results have been obtained. This feature of the time series identification problem is the same in other areas have a certain reference value.