Study on the Convergence of Hybrid Ant Colony Algorithm for Job Shop Scheduling Problems

To improve the performance of intelligence optimization algorithm for solving Job Shop Scheduling Problem, a hybrid ant colony algorithm called tabu search and ant (TSANT) algorithm with global convergence was proposed. In the hybrid ant colony algorithm, the MMAS algorithm was applied to search in the global solution space, and the tabu search algorithm was utilized as the local algorithm. The global convergence of TSANT algorithm proved to be true by analyzing the convergence of MMAS algorithm and TS algorithm by Markov chain theory. Under the guidance of the above convergence theory, we applied the hybrid algorithm to some typical benchmarks problems and found out the optimums of problems FT10, LA25and LA39 in a short period, which improved the quality of the solutions of Job Shop Scheduling Problem and demonstrated the effectiveness of the hybrid ant colony algorithm both in theory and practice.