Kaplan, Sargent and Goodman 
Ship trajectory data was also generated on an analog computer, 
using only linear equations in v and r while assuming a constant 
forward speed. The equations in that case were given arbitrary initial 
conditions and then allowed to seek equilibrium. The v and r analog 
signals were sampled every two seconds by A/D convertors to pro- 
vide the input to the system identification program. The results using 
the data supplied from the analog experiment are also included in 
Table 1 and in that case, even though the data can at best be consider- 
ed to be 1% accurate, better values for the four coefficients consider- 
ed as unknowns are obtained than that predicted using the noisy di- 
gitally generated data. This is ascribed to the fact that more accurate 
values of unknown coefficients can be predicted for a simpler system 
than for a system with a larger number of unknown coefficients. It 
is also possible that real ''noise'' from the analog computer output, 
which is closer to true processed experimental data, may not be as 
severe as the artifically generated digital noisy data. 
For the case of a hydrofoil craft, the nonlinear longitudinal 
equations of motion for a typical hydrofoil craft under autopilot con- 
trol are : 
Normal Force Equation 
L L 
= F R ° . 
hi, S_ Sle 17 eases theecs Cc, We C, We (14) 
Pitching Moment Equation 
L L 
Bi . R 
Gian ey a - ao 
; ~ Cc, ee | Fs + C, We (15) 
The lift on the forward foil is 
lp Wy Bits Caer whee 
= ae f Se Ge era * SHLERs Oe 
m 3 V 5 e + YY aT F 
6 6 
+ 2 
F 
+ — 
Cy v (16) 
1642 
