Vehicle sideslip angle measurement based on sensor data fusion using an integrated ANFIS and an Unscented Kalman Filter algorithm Articles
Overview
published in
publication date
- May 2016
start page
- 832
end page
- 845
volume
- 72-73
Digital Object Identifier (DOI)
full text
International Standard Serial Number (ISSN)
- 0888-3270
Electronic International Standard Serial Number (EISSN)
- 1096-1216
abstract
- Most existing ESC (Electronic Stability Control) systems rely on the measurement of both yaw rate and sideslip angle. However, one of the main issues is that the sideslip angle cannot be measured directly because the sensors are too expensive. For this reason, sideslip angle estimation has been widely discussed in the relevant literature. The modeling of sideslip angle is complex due to the non-linear dynamics of the vehicle. In this paper, we propose a novel observer based on ANFIS, combined with Kalman Filters in order to estimate the sideslip angle, which in turn is used to control the vehicle dynamics and improve its behavior. For this reason, low-cost sensor measurements which are integrated into the actual vehicle and executed in real time have to be used. The ANFIS system estimates a "pseudo-sideslip angle" through parameters which are easily measured, using sensors equipped in actual vehicles (inertial sensors and steering wheel sensors); this value is introduced in UKF in order to filter noise and to minimize the variance of the estimation mean square error. The estimator has been validated by comparing the observed proposal with the values provided by the CARSIM model, which is a piece of experimentally validated software. The advantage of this estimation is the modeling of the non-linear dynamics of the vehicle, by means of signals which are directly measured from vehicle sensors. The results show the effectiveness of the proposed ANFIS+UKF-based sideslip angle estimator.
Classification
subjects
- Mechanical Engineering
keywords
- vehicle dynamics; sideslip angle; estimation; unscented kalman filter; adaptive neuro-fuzzy inference system