Some methods such as population isolation mechanism

optimum reserved strategy

arithmetic crossover

adaptive random mutation and heterogeneous strategy are used to improve genetic algorithm. Besides the advantage that the optimal solution can be found only by the value of objective function

the local searching capability is enhanced in this improve genetic algorithm. This algorithm is applied to solve nonlinear equations. Numerical examples demonstrated that this algorithm can solve the optimization problem which has nonlinear equality constraint. Furthermore

the heterogeneous strategy speeds up the process of convergence and raises the convergence probability of global optimal solution.