한국해양대학교

Detailed Information

Metadata Downloads

A Study on the Compensation Algorithm for Inertial Navigation System

Title
A Study on the Compensation Algorithm for Inertial Navigation System
Alternative Title
A Study on the Compensation Algorithm for Inertial Navigation System
Author(s)
NGUYENDUYANH
Publication Year
2006
Publisher
한국해양대학교 대학원
URI
http://kmou.dcollection.net/jsp/common/DcLoOrgPer.jsp?sItemId=000002174188
http://repository.kmou.ac.kr/handle/2014.oak/8144
Abstract
This paper describes a method that how a relatively compensate the position errors in the using of low cost Inertial Measurement Unit (IMU) has been evaluated and compared with the well established method based on a Kalman Filter (KF). The compensation algorithm for IMU has been applied to the problem of integrating information in Inertial Navigation System (INS). The KF is used to estimate and compensate the errors of an INS by using the integrated INS velocity and position, respectively.

First by using Kalman Filter, we try to reduce noise of acceleration data, where two of an acceleration, constant drift and period drift, are considered. With the constant drift, it depends on sensor and it always keeps on constant error. When using double integration for calculates distance and velocity, these kinds of drifts can make increasing velocity and position errors. So, we tried to find these errors and used constant compensation algorithm for compensation of errors in data.

Second, external environment circumstance is changed ordinarily. Almost of them can be changed on periodic time. The average drift can be obtained during constant periodic time. And use this value, we consider with a factor as a periodic external disturbance which affects to the exact position. We used a repetitive method to reduce the external environment change. We verified the proposed algorithm by simulation results.
Appears in Collections:
동북아물류시스템학과 > Thesis
Files in This Item:
000002174188.pdf Download

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse