PhD Alumni

Delvecchio Diego

Present position:

Thesis title:  Design and analysis of semi-active systems for vehicles
Advisor:  Sergio Savaresi
Research area:  Control, automation and measurement
Thesis abstract:  
This Thesis deals with the analysis and the design of control systems of semi-active suspensions for ground vehicles. In particular the Thesis discusses the following topics: • Analysis and modeling of semi-active actuators. A protocol for the characterization and non-linear models of semi-active dampers are proposed. • A new comfort-oriented control strategy based on the half car model is presented to compensate the limitations of the state-of-art algorithm designed for quarter-car representation of the vehicle. • A new handling-oriented control strategy has been developed to maximize the efficiency of the transmission of force between the tire and the ground. • The application of the model reference control technique to semi-active dampers is discussed. In particular, the description of an industrial application of such control approach is illustrated. • A simulation study of the influence of suspension systems on the out-of-plane dynamics of a motorcycle is deeply analyzed. Furthermore, a novel control strategy which aims at increasing the stability of the motorcycle during critical maneuvers is designed. • For industrial application of semi-active suspensions system, information regarding the vertical dynamics of the vehicle is required to be as much robust and cost effective as possible. Observers based on Kalman-filtering technique for estimation of suspension stroke speed and tire deflection with different sensors layouts are designed and analyzed. Finally, a sensor reduction in a 4-wheels vehicle equipped with semi-active suspensions is presented. • In the last Chapter an innovative architecture of suspension, the dual stage suspension system, is presented and analyzed. Such system is able to approximate the behavior of a reactive system in which both the damping and the stiffness coefficients can be modulated.