Abstract:

 As the current stochastic optimization algorithms almost simulate evolutional process to solve the real optimization problems and  the searching result can’t reach the optimum solution, and it is difficult to use them in the real-time application,  a novel, fast and direct random optimization algorithm was proposed. The random search was directly used to find the optimum to cut off the additional time, and the random search process was divided into two different phases. In the first one a global optimizer was created through connecting three sub-optimizers including increasing parameters in serials and a global optimization module was formed with the paralleling optimizers to get a global solution; in the second one local optimization module was created to obtain  more precise an optimum. The tests for quite a few complicated functions indicate that the proposed optimization algorithm is rapid and effective and outperforms the current global optimization algorithms.

Key words: optimization method, direct random optimization algorithm (DROA), global search, local search, function optimization