Publications

Detailed Information

Tilt Estimation Using Low Cost Inertial Sensors in Dynamic Situations : 동적 상황에서 저가 관성 센서를 이용한 틸트 추정

DC Field Value Language
dc.contributor.advisor박찬국-
dc.contributor.author박정민-
dc.date.accessioned2019-05-07T03:09:37Z-
dc.date.available2019-05-07T03:09:37Z-
dc.date.issued2019-02-
dc.identifier.other000000154168-
dc.identifier.urihttps://hdl.handle.net/10371/150663-
dc.description학위논문 (석사)-- 서울대학교 대학원 : 공과대학 기계항공공학부, 2019. 2. 박찬국.-
dc.description.abstractIn this paper, we propose a method to improve the performance of attitude reference system (ARS) in dynamic situations. In order to do so, the method estimates acceleration using gyros simultaneously while estimating tilt using accelerometers. In general, ARS methods cannot be used in dynamic situations-
dc.description.abstracthowever, it is possible when the magnitude and direction of acceleration which the sensor is subject to is known. In many smart device usage scenarios, acceleration is produced by rotation about a fixed body joint. In such cases, it is possible to determine the acceleration from gyro measurements. By correcting for acceleration, the accuracy the availability of tilt estimation can be made nearly equal to that in static mode, and gyro drift can be compensated even in dynamic situations. We tested the proposed method in various smart device usage scenarios with VICON motion capture system as reference, and have confirmed that the proposed method improves the attitude estimation performance.-
dc.description.abstract본 논문에서는 동적 상황에서 저가 관성 센서 기반 기기의 틸트 추정 성능을 향상시키기 위해 가속도계 측정치를 활용해 자세를 계산함과 동시에 자이로 측정치를 활용해 가속도를 추정한다. 일반적인 Attitude Heading and Reference System (ARS)에서는 정적 상황(static mode)을 가정하고 가속도계의 측정치로부터 틸트를 계산하며, 물체의 가속도를 알지 못하면 동적 상황(dynamic mode)에서는 가속도계 측정치를 활용할 수 없다. 그러나 스마트 기기의 경우 신체 관절 중심의 회전운동에 의한 가속이 일어나고, 이러한 경우 자이로 정보를 이용하여 스마트 기기의 가속도를 추정할 수 있다. 물체의 가속도를 알면 동적 상황에서도 정적 상황에서와 마찬가지로 자세를 추정할 수 있다. 본 논문에서는 VICON 모션 캡처 시스템을 이용한 실험을 통해 제안한 기법을 적용하였을 때 다양한 동작에 대해 스마트 기기의 틸트 추정 성능이 향상됨을 검증하였다.-
dc.description.tableofcontentsChapter 1 Introduction 1
1.1 Motivation and Background 1
1.2 Objectives and Contributions 2
1.3 Organization of the Thesis 4
Chapter 2 Tilt Estimation Using Low Cost Sensors 5
2.1 Gyro Integration for Tilt Estimation 5
2.1.1 Performance by Gyro White Nosie 5
2.1.2 Performance by Gyro Bias 8
2.2 Accelerometer Based Tilt Estimation 13
2.2.1 Performance by Accelerometer White Noise 14
2.2.2 Performance by Disturbance 15
2.3 Gyro and Accelerometer Fusion using the Kalman Filter 19
2.3.1 Steady-State Error Covariance 19
2.3.2 Choosing Compatible Sensors 23
2.3.3 Performance by Disturbance 24
Chapter 3 ARS in Dynamic Situations 25
3.1 Covariance Adaptation Methods 26
3.1.1 Accelerometer Meausrement Based Adaptation 26
3.1.2 Gyro Measurement Based Adaptation 27
3.1.3 Innovation Based Adaptation 28
3.1.4 Probability Model Based Adaptation 29
3.2 Kinematic Modelling 29
3.2.1 Multiple-Joint Rotation Model for Human Motion 30
3.2.2 Applications of the Multiple Rigid Body Model 30
Chapter 4 ARS with Single-Joint Kinematic Constraint 32
4.1 Method 32
4.1.1 Single Rigid Body Rotation Model for Human Motion 33
4.1.2 Online Estimation of Center of Rotation 37
4.1.3 Verification of Estimation Performance 39
4.2 Experiment 41
4.2.1 Walking (Smartphone) 42
4.2.2 Walking (Smartwatch) 44
4.2.3 Jogging (Smartwatch) 47
4.2.4 Table Tennis (Smartwatch) 49
4.2.5 Virtual Reality (HMD) 52
Chapter 5 Conclusion 59
-
dc.language.isoeng-
dc.publisher서울대학교 대학원-
dc.subject.ddc621-
dc.titleTilt Estimation Using Low Cost Inertial Sensors in Dynamic Situations-
dc.title.alternative동적 상황에서 저가 관성 센서를 이용한 틸트 추정-
dc.typeThesis-
dc.typeDissertation-
dc.description.degreeMaster-
dc.contributor.affiliation공과대학 기계항공공학부-
dc.date.awarded2019-02-
dc.identifier.uciI804:11032-000000154168-
dc.identifier.holdings000000000026▲000000000039▲000000154168▲-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share