Time series models

Abstract

This literature study is a part of my study at University of Twente. The purpose of my literature study is to explore information in the area of time series model and some curve fitting, and also to learn how to search and to select the relevant literature. This literature report will be used as basic knowledge before I am doing my final project at Centraal Bedieningssysteem Distributiepompstattions(CBD) Gemeentewaterleidingen Amsterdam. The huge amount of measurement from 120 channel is available (more or less 3456000 measurement per day) at CBD, from this heap information we want to learn the way the measurements vary overtime and to construct the time patterns of past period without storing all the original measurements. My final project deal with storing the time patterns in an efficient way and controlling the whole process of the waterworks. Base on this fact, time series model and curve fitting are chosen as a topic of my literature study. My report is organized in the following way. First part of my report deal with the literature search process that I do and its result. The second part deal with my literature survey on time series and curve fitting. In this part I try to give an overview on the time series models (in general), linear stationary models, model identification, parameter estimation, diagnostic checking, regression model with time series error, some smoothing and curve fitting methods.