Estimation based on acceleration measures of an active suspension plant

The vehicle suspension system is responsible for comfort, safety and vehicle control. In order to positively manipulate these properties, control and estimation theory are used to adapt the system to different road conditions. This paper considers three estimation methods, which are designed to retrieve the system states using only acceleration measures: the Kalman Filter, Particle Filter and Artificial Neuronal Network. Also it considers two control methods: LQR, pole location, which it minimizes, the chassis acceleration (a variable used to improve the vehicle comfort). Finally the controllers and estimators are implemented in simulation, using the model of the Quanser active suspension plant. This plant corresponds to a linear dynamic system and is considered in this paper because it resembles a quarter-car suspension.