Solving ordinary differential equations using genetic algorithms and the Taylor series matrix method
Daniel Gutierrez-Navarro, Servando López-Aguayo · Journal of Physics Communications · 2018
A method for solving ordinary differential equations based in evolutionary algorithms is introduced. The concept of Taylor series matrix is defined, allowing to transform a differential equation into an optimization problem, in which the objective function is constituted by the coefficients of a series expansion. An ad-hoc genetic algorithm is used to find such coefficients that satisfy particular conditions. The efficiency of the algorithm is demonstrated by solving several ordinary differential equations.