Abstract: The search algorithm of the steepest direction is proposed to solve the optimization problem that the objective function is complex and the derivative is hard or impossible to obtain, The algorithm doesn’t need to obtain the derivative in its searching process, overcoming the weakness of needing to acquire the derivative during their searching process, which some other local search algorithms have. A hybrid genetic algorithm, based on the search algorithm of the steepest direction, is proposed. The algorithm combines the search algorithm of the steepest direction with the genetic algorithm effectively, enhancing the searching ability of the genetic algorithm. The experimental results show that the improved genetic algorithm is more effective than some current optimization algorithms.

Key words: genetic algorithm; steepest direction; local search; fitness sharing