A genetic optimization algorithm is presented for interval finite element analysis method（FEM） of structure randomness problem.In this method

the Least Square Regression Response Surface（LSRRS） of control equations is taken as objective function and the process of biotic genetics and evolution is regarded as the process of function optimization

then interval solutions of equations are obtained.The algorithm has the advantages of global convergence and not needing derivative information

and is easy to implement on the conventional finite element software without writing complicated programs and so on.A numerical example shows that the method has a high calculating efficiency and accuracy.