Accelerometer-based estimation of the elongation speed in a motorcycle suspension via Kalman-filter techniques

This paper is about the estimation of the elongation speed in a motorcycle electro-hydraulic suspension using a couple of accelerometers or a single sensor at wheel side. A Kalman filtering approach allows to successfully solve the estimation problem in both the cases, overcoming the vibrational disturbances which heavily affect the accelerometers on a sport motorbike. To be implemented on an off-the-shelf ECU, the Kalman observers need a low computational cost estimator of force. Therefore the secondary aim of the paper is to present a simplified model of the semi-active damper and compare it with a Neural Network based benchmark. Experimental results show that both velocity and force can be correctly estimated, also in the case of a single accelerometer with just a slight loss of performances.