In this research, we consider a quasiperiodic system of nonlinear differential difference equations. We prove a new theorem to solve these equations. It says that one can always assure the existence of quasiperiodic solution by checking several conditions on an obtained approximate solution and further give a method to obtain an error bound of the approximate solution. The approximate solutions are constructed by the method of Galerkin's procedure based on trigonometric polynomials. We carry out some numerical experiments using the method of Galerkin's procedure.
We consider mixed integer programming (MIP) problems with uncertain data. Robust solutions to such problems are formulated as solutions of second-order cone programming problems with integer constraints, which can be solved by an adaptation of the Benders decomposition technique towards MIP with conic constraints. Preliminary numerical computation against robust 0-1 knapsack problems indicates that robustness can be achieved without substantial deterioration in optimal values.
Rapid reanalysis of eigensolutions after modifications is a problem of considerable practical importance. Several methods have been developed to compute eigenvalues by using modified parts. This paper proposes a numerical method for calculating eigenvalues of modified system using the numerical solutions of unmodified system with the Inverse Power method. The advantages of the proposed method are examined comparing with the solutions of inverse power method by several numerical examples. The results obtained agree well with the exact solutions in short CPU time and this indicates that the proposed method provides efficient convergency.