Job Shop Scheduling Based On Improved Ant Colony Algorithm

Scheduling is of great significance during the manufacturing process. Using reasonable dispatching decision, the utilization rate could be improved and the cost could be reduced. Consequently, the operational efficiency and economic benefits could be improved._Simple scheduling problem could be solved by mathematical programming. However, with the increase of the scheduling problems’scale, the traditional optimization methods couldn’t fit for such situations. Scheduling problem has been proved as NP-hard problem. With the development of computer technology, various of intelligent algorithms have been proposed and used in the scheduling problem, which has made the scheduling problem great development. This paper was using ant colony algorithm to solve the scheduling problems, the main contributions are listed as follows:This paper analyzed the coding problem generated when using intelligent algorithm in JSP problem. The redundancy solution and the infeasible solution and be generated during the coding process. It’s found that different the influence could be caused by the redundant solutions and the infeasible solutions when face different complexity. Such analysis could be very helpful in the choice of code.This paper improved the ant colony algorithm, and put the improved algorithm into JSP problem. New state transition rule and parameter adaptive rule was developed for improved ant colony algorithm. Such rules improved the performance of ant colony algorithm.This paper has made the choice of code—machine-based representation. Such selection is on the basis of analysis above. And infeasible solutions would be generated. To solve the problem of infeasible solutions, rules are put forward to transform infeasible solution into feasible solution. The diversity of population is improved, and the decline of algorithm’s efficiency caused by infeasible solution would be avoided.