The time evolution of a simple model for crossover is discussed. A variant ofthis model with an improved exploration behavior in phase space is derived as asubset of standard one- and multi-point crossover operations. This model issolved analytically in the flat fitness case. Numerical simulations compare theway of phase space exploration of different genetic operators. In the case of anon-flat fitness landscape, numerical solutions of the evolution equationspoint out ways to estimate premature convergence.
展开▼