For numerical optimization constraints appear in the large-scale, multi-function polymorphism, on global opti-mization with discrete variables under such circumstances, General optimization method, convergence is slow to seek global extremum of low probability. Made with genetic algorithm of numerical optimization constraints problem solution, by numerical simulation experimental results indicates that, the algorithm performance better than existing other algorithm, it not only can processing linear equation constraints, and also can processing nonlinear equation constraints, while improve has convergence speed reconciliation of precision, is efficient sound of intelligent algorithm, has is high of global found excellent ability and soon of convergence speed, on solution complex more peak more State function of optimization constraints problem has feasibility and effectiveness.