Fuzzy logic Kalman filter estimation for 2-wheel steerable vehicles

This article addresses the multisensor data fusion problem in the position estimation of a two-wheel steerable vehicle converted from a golf buggy. The fusion is based mainly on the extended Kalman filter approach. This paper describes how the estimator integrates sensory data from the differential global positioning system (DGPS), gyroscope and odometry to provide recursively an optimal estimate of the position and orientation of the vehicle. In addition, a modified kinematic process model is proposed that accounts and estimates the side-slip angles at the wheels. A technique that incorporates fuzzy logic to maintain the estimation consistency of the filter is also described Finally, the filter's performance is evaluated with simulations conducted using true data obtained from field trials.