Development of control strategies for regenerative automotive suspensions

The aim of the present master’s thesis is to develop a control logic for an existing regenerative damper to be used on commercial automotive suspensions. In a conventional suspension, the excitation due the unevenness of the road profile causes the kinetic energy to be stored in a spring. Then, the energy is released in the external environment by means of a damper, which dissipates it in the form of heat. In consideration of the current, growing concern for pollution and fuel consumption reduction even the kinetic energy released by a passive suspension is worth to be recovered. For this reason, car manufacturers are encouraging the development of regenerative suspension systems, which can convert the dynamic oscillation into electric power to be stored in the battery or to be used to actively control the suspension itself. This thesis deals with a suspension system equipped with a motor-pump group that exploits the oil flow caused by the piston moving in its tube to start a hydraulic pump. In turn, the pump movement drives an electric machine. Since the electric machine can be used both as a generator and as a motor, the system can both provide additional energy to the vehicle battery and use the electric power to force the piston movement. In particular, acting on the motor current it is possible to set the damping coefficient in order to adapt in real-time to the road roughness profile. First, the classical control strategies for active suspensions – Skyhook and Groundhook control – have been implemented on the regenerative system. The obtained results are consistent with the expected ones, although not optimal in terms of energy harvesting. For this reason a novel strategy based on the Maximum Power Point Tracking (MPPT) is developed. This control logic has to define the optimal damping to maximize the conversion efficiency, defined as the ratio between harvested power and dissipated one. To this end the algorithm introduced in this work exploits a hill-climb approach: It verifies the sign of the gradient of the optimized variable and acts on the command variable accordingly to find a local maximum. This logic is already employed in other energy harvesters such as photovoltaic panels and wind turbine generators. However, the implementation in automotive suspensions introduces additional challenges due to the impulsive nature of the road roughness and the relatively faster dynamics. The MPPT logic is first developed on a quarter-car model defined in the state space and then implemented and verified on a full-vehicle model. Numerical simulations through different road profiles demonstrate that the proposed technique is able to track the optimal efficiency of the regenerative suspension with adequate consistency.