Odometry Error Reduction in Wheelchair Using More Than One Sensor
Autonomous wheelchair promises a safer and convenient mobility for disabled and senior citizens. Odometry is to estimate position change over time. It uses data from one or more sensors such as encoder attached to wheel and IMU. Odometry is important for navigation of wheelchair. Odometry via wheel...
| Main Author: | |
|---|---|
| Format: | Final Year Project / Dissertation / Thesis |
| Published: |
2019
|
| Subjects: | |
| Online Access: | http://eprints.utar.edu.my/3457/ http://eprints.utar.edu.my/3457/1/ME%2D2019%2D1301582%2D1.pdf |
| _version_ | 1848885906691325952 |
|---|---|
| author | Boey, Daniel Mun Weng |
| author_facet | Boey, Daniel Mun Weng |
| author_sort | Boey, Daniel Mun Weng |
| building | UTAR Institutional Repository |
| collection | Online Access |
| description | Autonomous wheelchair promises a safer and convenient mobility for disabled and senior citizens. Odometry is to estimate position change over time. It uses data from one or more sensors such as encoder attached to wheel and IMU. Odometry is important for navigation of wheelchair. Odometry via wheel rotary encoder is prone to random error such as wheel slip on slippery or uneven surface, and inaccurate measurement of wheelbase and wheel diameter use to calculate position. Meanwhile, IMU data are noisy and once integrated to obtain position and orientation, their values drift. The IMU comprises of 3 separate sensors: accelerometer which measures acceleration and gyroscope which measures angular velocity and magnetometer which measures direction of magnetic north. The IMU outputs acceleration, angular velocity and magnetic field values based on the orientation of the sensor which is referred to as sensor coordinate system. In order to compute meaningful position of the wheelchair, the sensor coordinate system has to be aligned with the wheelchair coordinate system. Rotation matrix is applied to the IMU data to transform the IMU data. IMU data that are transformed is then filtered to reduce noise. When the sensor is stationary, the output data after the exponential filter still fluctuates between ±0.01 degree/s. Over time, the integrated reading of the gyro sensor will drift due to the fluctuation. Since the fluctuation is very small, it can be assumed to be zero to reduce drift. Next, the data from encoder, accelerometer and gyroscope are combined together with Kalman filter. Test was performed to obtain position from encoder, IMU and sensor fusion output and the position results were compared to the truth. The resulting fused position reduced error by 76.5%. |
| first_indexed | 2025-11-15T19:30:03Z |
| format | Final Year Project / Dissertation / Thesis |
| id | utar-3457 |
| institution | Universiti Tunku Abdul Rahman |
| institution_category | Local University |
| last_indexed | 2025-11-15T19:30:03Z |
| publishDate | 2019 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | utar-34572019-08-01T15:24:25Z Odometry Error Reduction in Wheelchair Using More Than One Sensor Boey, Daniel Mun Weng TJ Mechanical engineering and machinery Autonomous wheelchair promises a safer and convenient mobility for disabled and senior citizens. Odometry is to estimate position change over time. It uses data from one or more sensors such as encoder attached to wheel and IMU. Odometry is important for navigation of wheelchair. Odometry via wheel rotary encoder is prone to random error such as wheel slip on slippery or uneven surface, and inaccurate measurement of wheelbase and wheel diameter use to calculate position. Meanwhile, IMU data are noisy and once integrated to obtain position and orientation, their values drift. The IMU comprises of 3 separate sensors: accelerometer which measures acceleration and gyroscope which measures angular velocity and magnetometer which measures direction of magnetic north. The IMU outputs acceleration, angular velocity and magnetic field values based on the orientation of the sensor which is referred to as sensor coordinate system. In order to compute meaningful position of the wheelchair, the sensor coordinate system has to be aligned with the wheelchair coordinate system. Rotation matrix is applied to the IMU data to transform the IMU data. IMU data that are transformed is then filtered to reduce noise. When the sensor is stationary, the output data after the exponential filter still fluctuates between ±0.01 degree/s. Over time, the integrated reading of the gyro sensor will drift due to the fluctuation. Since the fluctuation is very small, it can be assumed to be zero to reduce drift. Next, the data from encoder, accelerometer and gyroscope are combined together with Kalman filter. Test was performed to obtain position from encoder, IMU and sensor fusion output and the position results were compared to the truth. The resulting fused position reduced error by 76.5%. 2019-01 Final Year Project / Dissertation / Thesis NonPeerReviewed application/pdf http://eprints.utar.edu.my/3457/1/ME%2D2019%2D1301582%2D1.pdf Boey, Daniel Mun Weng (2019) Odometry Error Reduction in Wheelchair Using More Than One Sensor. Final Year Project, UTAR. http://eprints.utar.edu.my/3457/ |
| spellingShingle | TJ Mechanical engineering and machinery Boey, Daniel Mun Weng Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title | Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title_full | Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title_fullStr | Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title_full_unstemmed | Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title_short | Odometry Error Reduction in Wheelchair Using More Than One Sensor |
| title_sort | odometry error reduction in wheelchair using more than one sensor |
| topic | TJ Mechanical engineering and machinery |
| url | http://eprints.utar.edu.my/3457/ http://eprints.utar.edu.my/3457/1/ME%2D2019%2D1301582%2D1.pdf |