Sensor fusion and noise modeling for improved vehicle localization

Thumbnail Image
Khot, Lav
Major Professor
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Organizational Unit
Agricultural and Biosystems Engineering

Since 1905, the Department of Agricultural Engineering, now the Department of Agricultural and Biosystems Engineering (ABE), has been a leader in providing engineering solutions to agricultural problems in the United States and the world. The department’s original mission was to mechanize agriculture. That mission has evolved to encompass a global view of the entire food production system–the wise management of natural resources in the production, processing, storage, handling, and use of food fiber and other biological products.

In 1905 Agricultural Engineering was recognized as a subdivision of the Department of Agronomy, and in 1907 it was recognized as a unique department. It was renamed the Department of Agricultural and Biosystems Engineering in 1990. The department merged with the Department of Industrial Education and Technology in 2004.

Dates of Existence

Historical Names

  • Department of Agricultural Engineering (1907–1990)

Related Units

Journal Issue
Is Version Of

This thesis examines the application of sensor fusion technique for the real-time vehicle localization improvement. The Kalman filtering technique has been used for the fusion process in this research. Previous researchers found Kalman filtering creditable in removing the white noises in sensor measurements but very few of them explained the fine tuning of the system and measurement noise matrices which are the foremost performance decisive parameters in the recursive Kalman filter algorithm. In the first application, the Kalman filtering was used to improve the navigational context recognization ability of an autonomous robot when the robot is navigating in simulated tree plantation nursery. An Extended Kalman Filtering (EKF) algorithm was developed and implemented to improve the accuracy of posture estimation of a skid-steered autonomous robot. A kinematic system model consisting of seven states was developed for implementing an EKF algorithm. The GPS error along with external vibration noise and Dynamic Measurement Unit's static bias drift was found to be the main sources of error affecting the position and heading of the robot vehicle. In addition to the EKF, a second order autoregression error model was developed to model the real-time kinematic-GPS (RTK-GPS) errors. The EKF with Autoregressive error model enhanced robot's localization accuracy over the EKF without incorporating an error model. Furthermore, the developed filtering and K-means clustering algorithms were successful in recognizing and reconstructing the navigational context of an autonomous weeding robot in a simulated tree plantation nursery. The second application of EKF was for improving the attitude angle estimates of the self-propelled sprayer by fusing the roll and pitch estimates from a digital elevation model (DEM) with the roll measurements from terrain compensation module sensor (TCM) and pitch estimates from a single GPS sensor. The EKF algorithm was capable of estimating the sprayer attitude angles even when the DEMs attitude estimates were not available for a certain period due to the out of bound circumstance of the DEM. The EKF and AR error model algorithms were also capable of removing the high frequency noise associated with the TCM and GPS sensor measurements.

Sun Jan 01 00:00:00 UTC 2006