The hybrid genetic algorithm for solving nonlinear programming
Wang Hong-gang, Zeng Jianchao · 2002
Genetic algorithms have been shown to be robust optimization algorithms for real value functions defined over domains of the form R/sup n/ (R denotes the real number). But there exist some obstacles in genetic algorithms such as premature convergence and slow convergence speed. A new approach called Hybrid Genetic Algorithms (HGA) is presented to overcome these obstacles for nonlinear programming by combining genetic algorithms with the feasible path method after introducing a learning operator. Finally, the validity of the approach is illustrated by providing HGA for nonlinear programming.