Human Modeling: Kinematics - Human Motion Simulation: Predictive Dynamics - page 9

(e.g., the foot). This chapter addresses exactly this issue. It will provide a rigorous
method for formulating a set of equations that have the joint variables as their
parameters. If the final hand position is required, variations in the joint variables
are substituted into the equation and the final position is readily obtained.
We shall also introduce a method for modeling human joints, as simple or as
complex as necessary, that represents the resulting interaction between any two
segmental links. The simplest form of these joints is the rotational joint (such as
the elbow joint). The combination of a number of simple joints can become com-
plex in nature but still be represented using this straightforward approach.
The Denavit

Hartenberg (DH) method was created in the 1950s to systemati-
cally represent the relation between two coordinate systems, but was only exten-
sively used in the early 1980s with the appearance of computational methods and
hardware that enable the necessary calculations. The method is currently used to
a great extent in the analysis and control of robotic manipulators. This method
has also been successful in addressing human motion, in particular towards gain-
ing a better understanding of the mechanics of human motion.
It is important to distinguish the difference between a rigid body and a flexible
body. A rigid body is one that cannot deform (we typically consider bone as non-
deforming, at least for the moment). However, a flexible body (or deformable
object) is one that undergoes relatively large strains when subjected to a load
(e.g., soft tissue). For the approach presented in this section, only rigid body
motion is assumed at all times. Indeed, for ergonomic design considerations, rigid
body motion is adequate to address most problems. Muscle interaction and defor-
mation will be addressed in later chapters.
The DH modeling method is suitable for addressing the motion of kinematic
structures that are arranged in series. The DH method will be used to perform
analysis on the human body in this chapter, and will be used to predict postures
and perform ergonomic analysis in later chapters. A posture is defined in this text
as the configuration of a series of segmental links in the human body .
The human body is indeed arranged in series, where each independent anatom-
ical structure is connected to another via a joint. Consider, for example, that there
exists a main coordinate system located at the waist. From that coordinate system,
one may be able to draw a branch by identifying a rigid link, connected through a
joint to another rigid link, connected to another link, until you reach the hand.
Each finger also comprises a number of segmental links connected via joints.
Similarly, also starting from the waist, one may follow the connection to reach
the head, the other hand, the left foot, and the right foot. We shall refer to one
such chain as a branch. For example, Figure 2.3 depicts the modeling of a human
into a number of kinematic branches. A hand can be represented by five branches,
one for each finger.
Because we consider gross human motion, detailed modeling of the joint con-
nection is less important at this stage. Nevertheless, because in many cases accu-
rate biomechanical modeling of a joint is needed, in Chapter 4 we will also
present a more elaborate method for representing the kinematic interactivity