An Improved Tabu Search Algorithm for Global Optimization of Engineering Design Problems

To find the global optimal solution of a multimodal function with both continuous and discrete variables, an improved tabu search algorithm is proposed. The improvements include new generating mechanisms for initial and neighborhood solutions, the exclusive use of the tabu list, the restarting methodology for different cycle of iterations as well as the shifting away from the worst solutions. The numerical results on two numerical examples are reported to demonstrate the feasibility and merit of the proposed algorithm.