Vehicle Dynamics Control System Actuating an Active Differential 2007-01-0928

In the last years automotive industry has shown a growing interest in exploring the field of actively controlled differentials due to the promising expectations of their application for the improvement of handling performances. The differential here considered has two clutches, which connect each driveshaft with a second part that, through a gear, always rotates faster. This device allows the control system to transfer torque from one wheel to the other one, almost independently from their relative velocity.

The control algorithms development has been carried out using a vehicle model that can precisely simulate the handling response, the powertrain dynamics and the actuation system behaviour. To obtain a representative powertrain response, the development of a customized library in Simulink has been required. The results of this preliminary modelling phase are shown in the first part of the paper.

Applying a methodology commonly shared among car manufacturers, for this research a two-steps approach has been adopted: firstly an intensive simulation campaign has been executed to design and tune the controller in the virtual environment; secondly the logic has been tested using HIL techniques.

The second part of the paper introduces the general structure of control system developed and its main algorithms. Vehicle performance improvements, evaluated through simulation and HIL tests, are underlined.

The aim of the control system presented in this paper is to improve vehicle dynamics, exploiting the potentialities of an active differential, of a 4WD sport vehicle. This powertrain layout is characterized by the presence of three differentials and, in particular, by one centre self-locking differential. As better explained in [7] e [15], this kind of device can physically lock, reducing the number of degrees of freedom. In this case the two output shafts rotate at the same speed and the differential does not transfer torque in terms of its nominal performance. As also underlined by [9], in the condition of locked axle the driving torque distribution is dominated by tires behaviour. Due to this fact the modelling phase has been carried on with great care to all these phenomena. In this work it has been necessary to adopt a reliable simulation tool for the development of the presented control system. In the next section the critical aspects of the interaction among vehicle, tires and powertrain are discussed, focusing on this specific application. Then the general structure and the principles behind the developed algorithms are presented, together with the encouraging simulation results. Due to industrial confidential agreements, the scales of the most significant plots will not be reported. Anyhow the relative improvement, with respect to passive vehicle behaviour, will be presented, underlined and discussed.