A Hybrid Optimized Algorithm Based on Simplex Method and Genetic Algorithm

Based on the simplex method and real-code genetic algorithm, a hybrid computational algorithm has been presented in this paper. In this hybrid genetic algorithm some improved genetic mechanisms, for example non-linear ranking selection, improved crossover operation combining the differential computation with arithmetic crossover and non-uniform mutation operation, are also adopted to overcome the slow convergence and premature problem in the simple genetic algorithm. The experimental results show that the new algorithm not only improves the global optimization performance, but also quickens the convergence speed and obtains robust results with good quality, which indicates this new algorithm is a promising approach for solving global optimization problems.