Abstract: Free trim hydrostatic ship characteristics are calculated applying the genetic algorithm (GA). According to the equilibrium condition of ship free floating, the calculation of free floatation comes down to a multi-objective constrained optimization problem, solved by a GA with real number coding. Necessary improvement of the GA is made in practice in order to speed up the evolution. As the foundation, the method of the hull surface representation, plane–surface intersection, tanks division and capacity calculation are described briefly. The hull surface and deck are represented with a single NURBS surface interpolating the waterlines, and deck side line(s) and the camber arcs, respectively. On the basis of an efficient intersection algorithm and the geometric properties calculation method, the virtual subdivision is realized and the capacity at any inclination surface calculated. The floatation calculation results and their comparison with the conventional results of full-scale ships verify the feasibility of the solution with a GA. The above 3D hull surface and tanks representation and the related calculation method determine the increased precision and convenience of the floatation calculation than that based on 2D ones. Compared with the other iteration algorithm, it is unnecessary to specify the initial iteration point and nothing is needed other than the ship hull offset, the total weight of ship and the center of it.
Keywords: Free trim, multi-objective constrained optimization, genetic algorithm, real number coding, NURBS, hull representation, virtual subdivision, plane–surface intersection
