A New Hybrid of Evolutionary and Conventional Optimization Algorithms

Combination of the Evolutionary and Conventional algorithms has opened a new horizon to Optimization algorithms. In the previous hybrid algorithm, firstly an evolutionary algorithm has been applied and then the obtained result has been employed as an initial guess for a conventional algorithm. In this method the advantages of each algorithm cannot be realized while they are running at the same time. In addition, it is not determined in which iteration the evolutionary algorithm should stop and the conventional one should start. Certainly an improper iteration selection leads to an inefficient hybrid algorithm. To overcome above shortages, this paper proposes a novel hybrid algorithm that uses the abilities of evolutionary and conventional algorithm simultaneously. In each iteration of the proposed algorithm both evolutionary and conventional algorithms have been applied. Simulation results on some benchmark problems show that the proposed hybrid 816 A. Khosravi, A. Lari and J. Addeh algorithm has faster convergence and a more reliable solution than the conventional hybrid algorithm.