Estimation of absolute vehicle speed using fuzzy logic rule-based Kalman filter

Accurate knowledge on the absolute or true speed of a vehicle, if and when available, can be used to enhance advanced vehicle dynamics control systems such as anti-lock brake systems (ABS) and auto-traction systems (ATS) control schemes. Current conventional method uses wheel speed measurements to estimate the speed of the vehicle. As a result, indication of the vehicle speed becomes erroneous and, thus, unreliable when large slips occur between the wheels and terrain. This paper describes a fuzzy rule-based Kalman filtering technique which employs an additional accelerometer to complement the wheel-based speed sensor, and produce an accurate estimation of the true speed of a vehicle. We use the Kalman filters to deal with the noise and uncertainties in the speed and acceleration models, and fuzzy logic to tune the covariances and reset the initialization of the filter according to slip conditions detected and measurement-estimation condition. Experiments were conducted using an actual vehicle to verify the proposed strategy.