Zero, dynamics is an important feature in system analysis and controller de
sign. Its behavior plays a major role in determining, the performance limit
s of certain feedback systems. Since the intrinsic zero dynamics can not be
influenced by feedback compensation , it is important to design physical s
ystems so that they possess desired zero dynamics. However; the calculation
of the zero dynamics is usually complicated especially if a form which is
closely related to the physical system and suitable for design is required.
In this paper, a method is proposed to derive the zero dynamics of physica
l systems from bond graph models. This method incorporates the definition o
f zero dynamics in the differential geometric approach and the causality ma
nipulation in the bond graph, representation. By doing so, the state equati
ons of the zero dynamics can be easily! obtained. The system elements which
are responsible for the zero dynamics can be identified. In addition, if i
solated subsystems which exhibit the zero dynamics exists that can be found
. Thus, the design of physical systems including the consideration of the z
ero dynamics become straightforward. This approach is generalized for MIMO
systems In the Part II paper.