CHAPTER 7 
EXAMPLES OF MODEL FITTING TECHNIQUE APPLIED TO THE 
ANALYSIS OF SEAKEEPING DATA 
In this chapter, examples of application of the model fitting technique to seakeeping 
data will be presented to demonstrate the applicability of this technique to the analysis of 
these data. 
7.1 EXAMPLES OF AR(n) MODEL FITTING FOR THE PREDICTION 
OF SEAKEEPING DATA 
Figure 7.14° shows examples of fitting of AR(n) models to observed seakeeping 
data. After the appropriate order number n was determined by the MAIC method, as 
described in Section 5.5.4, the parameters a ... a, were calculated from 800 observed 
data points for the respective processes. Here for rolling n=7 and for swaying n=19 were 
found to be the optimum. Figure 7.1 shows the values predicted for each process by 
Xp = — 4 Xj-] — A2K_-2 —- —— ApXi-n (7.1) 
PREDICTION START 1 - STEP AHEAD FOR ROLLING 
a 
a 
DATA LEVEL 
} 
0.00 1050 21.00 |3150/ 4200 S250 63.00 7350 84.00 4.50 105.00 115.50 126.00 136.50 147.00 157.50 168.00 178.50 189.00 199.50 210.00 
~a.04 Se 7 
TIME ~ 
PREDICTION START 10 - STEP AHEAD FOR SWAYING 
DATA LEVEL 
‘ 2 
8 
Se 
‘0.00 1050 21.00 3150 ,4200 5250} 63.00 A 94.50 105.00' 115.50 126.00 413650 147.00, 
~ TIME \ 
PREDICTION START 1 - STEP AHEAD FOR SWAYING 
} t 
0.00 1050 21.00 31.50 p4200 S250} 63.00 73.50 | 64.00 }94.50 105.00_ 115.50 126.00 £136.50 147.00 
\ Si if \ 7\ TIME Vy! V 
DATA LEVEL 
157.50, 168.00 178.50 189.00 199.50 210.00 
y 
\ 
Fig. 7.1. Comparison between the predicted values of seakeeping data 
and the observed values. 
(From Yamanouchi, et al.46) 
227 
