Genetic Accelerated Trust Region Algorithm without Constraints

In this paper, a genetic accelerated trust region algorithm without constraints presented, which can deviate saddle point, and quickly reach more excellent point by using probabilistic optimization approach. While trust region algorithm can only converge to partial optimum, the new algorithm can converge to global optimum with an example proved the validity.