On solution of nonlinear models: A hybrid algorithm of artificial bee colony algorithm and artificial showering algorithm

Solving nonlinear systems of equations is a typically difficult computation problem. Classical solvers like Newton-type methods suffer from convergence to a local optimum due to their high sensitivity to the initialization conditions. This study combines the global exploration capabilities of Artificial Bee Colony Algorithm and the high exploitation characteristics of Artificial Showering Algorithm to model a hybrid mechanism for solving systems of nonlinear models. The proposed method is applied to solve three nonlinear systems including the “Chemical Equilibrium” and “Geometry Size Thin Wall Girder Section” physical models. Statistical analysis of the results shows that the hybrid approach possesses a high success rate and a quick convergence rate.