A differential evolutionary method for solving a class of differential equations numerically*

A method for solving differential equations using the differential evolutionary algorithm (DE) is presented in this contribution. In the case of differential equation is a polynomial function of its variables, the algorithm makes possible solution vectors evolve toward the best solution of the differential equation based on certain criteria, such as least squares or another error measure. In the case of non polynomial differential equations, the DE is applied in the context of variational methods. Numerical results for solving a segment of the Lorenz and van der Pol differential equations are provided.