Tabu Search Algorithm Based On Harmony Strategy

Tabu Search (TS) algorithm is a kind of novel intelligence optimization algorithm to solve the optimization problems. It simulates the process of human's intelligence and expands the Local Neighborhood Search. It has been widely applied in the combinatorial optimization problems, particularly in some complex combinatorial optimization problems, and it has achieved many notable results.Initial solution is an important factor to impact the performance of Tabu Search algorithm. The good initial solution can help to get the best solution fast.Harmony is a kind of tone combination constituted by instruments by certain rules. Harmony search strategy is givern on the basis of the principle of harmony. It has a strong capacity to generate new solutions and each search will generate several better solutions in place of the inferior solutions to assure the quality of the next search. Harmony strategy will generate an initial solution set, so it has obvious parallel feature and it can assure the intensification and diversification of search and the global optimum of the algorithm.A new tabu search algorithm based on harmony search strategy (HTS) is proposed, combining the characteristics of Tabu Search algorithm and Harmony strategy. The new algorithm constructs several better solutions based on the Harmony strategy and conducts the search with several initial solutions avoiding getting into the local optimum. Furthermore, the convergence of the new algorithm is proved.The simulation result of new algorithm was very good, and it could achieve or be close to the currently best result. Comparing with the GTTS algorithm and TIS algorithm, it is better than the former and not inferior to the latter.