Dynamtes of Naval Craft - System Identtftcatton 
make the intelligent engineering decisions that he is calling for. It 
would be interesting to have his comments on that. 
REPLY TO DISCUSSION 
Paul Kaplan 
Oceantes Ine. 
New York, U.S.A. 
A number of points are raised by Mr Blacklock. With regard 
to the particular case of a hydrofoil craft, the design of a control 
system for such a craft is a function of the basic dynamics of that 
craft, which differs from that for a surface ship, a submarine or an 
SES craft. This difference is manifested by finding what are the basic 
dynamics of the vehicle and then using that information to guide the 
application of particular sensor outputs to provide commands to the 
control. This is most often done nowadays for very complex systems 
by means of computer simulation, and certainly well before actual 
experiments with the vehicle at sea. 
The procedure in this paper is not in any way to be consider- 
‘ed as statistical. What we are doing here is applying something that 
is independent of actual statistics but functions within the domain of 
randomness, which represents the real situation. We are finding 
values of the coefficients that are supposed to be invariant for a ve- 
hicle throughout the whole range of maneuvers or motions that it will 
experience. This is a way of dealing with full scale dynamics, which 
is really another method of replacing getting data from a towing tank 
under controlled tests. When you go to sea how is one t> find out what 
the numerical values of the coefficients are ? All you have is a result- 
ing trajectory for which there are a lot of possible ways that one can 
"fit'' some representation to. However there is something that is 
unique and invariant which must be close to what is being measured in 
the tank in order to allow an analyst to structure a representation of 
the original system. That is a feature of this method, since it brings 
you to that point of comparison with model test data as well as the ef- 
fective values of the coefficients representing the full scale craft dyn- 
amics. 
The next point is in regard to Dr Fink's comments. There are 
a number of features of the work of Lt Hayes that are somewhat si- 
1693 
