IMU Data Noise Cleansing for Vehicular Applications

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/77552

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/77552

Title:	IMU Data Noise Cleansing for Vehicular Applications
Authors:	蔡玉楓;Tsai, Yu-Feng
Contributors:	資訊工程學系
Keywords:	慣性測量單元;IMU;calibration
Date:	2018-07-18
Issue Date:	2018-08-31 14:48:09 (UTC+8)
Publisher:	國立中央大學
Abstract:	隨著科技的進步，智能化設備被廣泛使用，例如：OBU(智能設備)備裝在車輛上，用來提高交通安全性，同時提供額外的信息助於行車，為了降低生產成本，這些感測器有所缺陷，通常需要經過校准才能夠備使用。在這篇論文中，我們提出一套系統，適用於整合成後照鏡的 OBU，所提出的系統，不需要人為進行校正，也不需要額外設備，此外，所提出的系統，結合 IMU 和 GPS 的資訊，提供更可靠的位置訊息。在實驗中，我們比較所提出的方法與傳統校準進行比較，結果顯示，提出的校正方法優於傳統方法，且限制更少;Along with the progress of hardware and software technologies, the smart devices have become ubiquitous. As an example, the smart device (i.e., OBU) is installed in the vehicle, which allows the intelligent transportation system to improve the transportation safety. While the GPS and IMU sensors built in OBU are able to provide the external information of the vehicle, the sensor readings generally require periodical calibration because these sensors are built with inexpensive and unreliable parts to reduce the cost. In this thesis, we propose a calibration system well suited for OBU built inside the rear mirror of the vehicle. The proposed system calibrates the sensor readings without human intervention and the need of external equipments. In addition, the proposed system fuses the IMU and GPS readings to provide more accurate vehicle position information. The experiments to compare the proposed calibration system and traditional calibration is conducted, and the results show that the proposed calibration system significantly outperforms the traditional counterpart.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	173	View/Open

社群 sharing

Loading...