Page 114 - Handbook of Biomechatronics

110 Naser Mehrabi and John McPhee

it is unlikely to find a unique steering setting that provides optimum steering
feel for the general population. This work does not directly deal with setting
of the preferred steering feel for a specific vehicle type. However, when a
preferred steering feel is set, the proposed EPS system can provide equal
steering feel across different predefined driver populations.
Realistic driver models can play a major role in accelerating the devel-
opment of driver-assistance technologies by reducing the cost and time asso-
ciated with physical experiments. The driver models are usually developed
to assess the vehicle performance and not the driver preference (e.g., path-
following driver model). Few studies have developed driver-centered
models that consider the driver’s physiology (i.e., neuromusculoskeletal sys-
tem) (Mehrabi et al., 2015a; Cole, 2012). These models can be used to give
insight about how our body interacts with the steering system. Understand-
ing and quantifying these interactions facilitates the development of the next
generation of driver-assistance technologies. A forward dynamic simulation
can simulate the interaction between driver and vehicle, and also provide a
platform to ask “what if” questions such as “what if a stronger driver steers
the same vehicle.” These predictive simulations can support the design of
individualized EPS controllers for different driver populations. Accordingly,
the following work presents a systematic approach to standardize EPS sys-
tems (e.g., steering feel) for various driver populations by considering the
human physiology.

3.2 Dynamic Model of Biomechatronic System

In this section, we present the models and methods used to develop and ver-
ify an individualized EPS system. We have two integrated models of driver
and vehicle that we will refer to as (1) high-fidelity and (2) simplified models.
The simplified model was used to design the EPS system, and the high-
fidelity model was used in MIL simulations to verify the performance of
the EPS controller. Finally, the characteristic curves and the EPS controllers
used in these models are presented.

3.2.1 High-Fidelity Driver-Vehicle Model
The high-fidelity integrated driver-vehicle model described in Mehrabi
et al. (2015a) and shown in Fig. 6A was used to simulate real-world driving
conditions. This model consists of a multibody dynamic model of a vehicle
and a three-dimensional (3D) neuromusculoskeletal model of a driver. The
muscle activities predicted by the neuromusculoskeletal driver model were
verified against the electromyographic activities of a driver’s arm muscles