Virtual testing is an essential tool in the analysis
of many automotive control concepts and in many case accurate
models of the vehicle dynamics are important. Traditional models,
as normally used in multi-body dynamics, are usually too
complex for this use and too difficult to derive. A solution that
is often much faster is to infer estimates of the parameter values
from measurements obtained by performing different driving
maneuvers with the car. However, most methodologies described
in the literature so far are focused on the identification of
single vehicle parameters, assuming most other parameters
to be known a priori, and often require a sophisticated and
expensive test setup. In this paper we show how methods from
stochastic subspace identification (SSI), model updating (MU)
and direct continuous time system identification (CTSI) can be
combined to obtain a fully parametrized model of the vehicle
suspension system from scratch, using only data from simple
dynamical tests and inexpensive measurement equipment. The
newly proposed method is evaluated on a real test car and
compared to the performance of a model obtained from static
tests. It was found that the model identified using the new
method matches the dynamics of both the real car and the
model obtained in static tests sufficiently well.