资 源 简 介
% Run an optimization using the epsilon-moea algorithm. Assumes that the
% optimization problem is in canonical form - all target functions are to be
% minimized.
%
% Arguments:
% creature - a description of the problem as returned from construct creature.
% pop_size - number of individuals to use for the simulation. The more you use,
% The better is the pareto front, but it"s slower and has diminishing
% returns.
% conv_gens - after this many generations with no change in the archive
% population, the iteration stops.
% nun_gens - maximum total number of generations, with or without convergence.
% objectives - a function that given a population array returns the fitness
% array.
% grid - the size of the hypercubes in the epsilon-dominance tests.
%
% Returns:
% population - the population matrix after the latest iteration.
% fitness - the
