Page 103 - Handbook of Biomechatronics

Model-Based Control of Biomechatronic Systems 99

1.2.2 Predictive Simulation
A predictive simulation is a forward dynamic simulation that can predict the
kinematics and kinetics of a task of interest based on the underlying physi-
ological phenomena governing its dynamics. In these simulations, a math-
ematical controller representing the human CNS coordinates the
movements of the biomechanical model for that task. However, to develop
such a controller, we should first understand how our CNS controls our
body. As first formulated by Bernstein (1967), the CNS simultaneously
coordinates the kinematics and kinetics of body motions, despite uncertain
(future) trajectories and the redundancy in muscle actuators. As an example,
during reaching and pointing tasks, where only the final position of the hand
is specified, an infinite number of hand trajectories (and muscle activation
patterns) can be expected to reach the target. However, despite the possible
variations, individuals usually choose a similar trajectory. The early observa-
tions of reaching and pointing tasks led to the well-known “Minimum-X”
models (e.g., minimum-jerk model (Flash and Hogan, 1985; Wada et al.,
2001), minimum-torque-change model (Uno et al., 1989), minimum-
variance model (Harris and Wolpert, 1998), and minimum-work model
(Soechting et al., 1995)) to predict the hand trajectory. These models
hypothesize that the CNS coordinates the body movement such that an
exertion (X) is minimized. Later, this hypothesis was extended to consider
physiologically motivated exertions such as muscle activation effort
(Crowninshield and Brand, 1981; Ackermann and van den Bogert, 2010;
Happee and Van der Helm, 1995), metabolic energy expenditure
(Anderson and Pandy, 2001; Peasgood et al., 2006), and muscle fatigue
(Sharif Razavian et al., 2015).
In computer simulations, the Minimum-X model has been successfully
implemented using dynamic optimization (DO) to predict the normative
human motion for a given task. A common DO approach parameterizes
the muscle activation profiles for the period of motion and searches the fea-
sible space to find the profiles that minimize X (Anderson and Pandy, 2001;
Davy and Audu, 1987; Yamaguchi and Zajac, 1990; Neptune and Hull,
1998; Kaplan and Heegaard, 2001; Sha and Thomas, 2013). This approach
provides an open-loop (feedforward) command of muscle activations to
control the given task. This command can represent the descending com-
mand of a well-repeated/well-learned task (e.g., platform diving
(Koschorreck and Mombaur, 2011)). In this approach, the CNS only recalls
the learned information, and does not intelligently adjust the commands in
real time.