Previous efforts to automate the optimisation of International America’s Cup Class (IACC) yachts have typically used an objective function that evaluates the performance of an individual boat using direct Computational Fluid Dynamic (CFD) analysis of the hull design. This approach suffers from the use of an inappropriate measure of merit as well as having extremely long execution times. A superior method is the use of an objective function incorporating a match racing tournament amongst a population of candidate designs. The resulting need to maintain a population of designs makes the problem well suited to population based optimisation methods such as Genetic Algorithms (GA). Performance issues are addressed through the use of a neural network based metamodel, trained using parameters sampled from the design space and calculated using the SPLASH potential flow code. This has resulted in an optimisation system that gives good results while retaining reasonable execution times.
[1]
Thomas W. Sederberg,et al.
Free-form deformation of solid geometric models
,
1986,
SIGGRAPH.
[2]
Peter N. Joubert,et al.
The development of the winged keel for twelve-metre yachts
,
1986,
Journal of Fluid Mechanics.
[3]
I. Sobol.
On the distribution of points in a cube and the approximate evaluation of integrals
,
1967
.
[4]
Nils Salvesen,et al.
Stars & stripes
,
1987
.
[5]
Shane G. Henderson,et al.
A Simulation Model for Predicting Yacht Match Race Outcomes
,
2004,
Oper. Res..
[6]
D'arcy W. Thompson.
On growth and form i
,
1943
.