Indirect Feedback Kalman Filter Based Sensor Fusion for Reducing Navigation Errors of an Autonomous Wheelchair
Patients with severe motor disabilities have difficulty maneuvering a wheelchair. An autonomous wheelchair with facility for destination selection via a brain-computer interface or eye tracker would be a possible solution. Accurate localization is important for such an autonomous wheelchair. Normal...
保存先:
| 第一著者: | |
|---|---|
| フォーマット: | Final Year Project / Dissertation / Thesis |
| 出版事項: |
2018
|
| 主題: | |
| オンライン･アクセス: | http://eprints.utar.edu.my/3622/1/ESA%2D2018%2D1406778%2D1.pdf http://eprints.utar.edu.my/3622/ |
| タグ: |
タグ追加
タグなし, このレコードへの初めてのタグを付けませんか!
|
| 要約: | Patients with severe motor disabilities have difficulty maneuvering a wheelchair. An autonomous wheelchair with facility for destination selection via a brain-computer interface or eye tracker would be a possible solution.
Accurate localization is important for such an autonomous wheelchair. Normally relative localization of the wheelchair is carried out using an odometry method based on data from the wheel encoders. The current study
aims to reduce wheelchair navigation errors in an indoor environment by the introduction of an additional sensor - a gyroscope. Fusion of the wheel encoders and gyroscope was effected using indirect feedback Kalman filter
algorithm. The algorithm was programmed in a small memory microcontroller to increase the portability of the wheelchair. The results of the study showed that fusion of encoders and gyroscope using indirect feedback Kalman filter significantly improved the wheelchair navigation accuracy (as high as 7.8 folds) in terms of mean distance errors compared to using odometry. |
|---|