Hydrid And Improvement Of Ant Algorithm

Scheduling problem, which means making operation plans during the manufacturingprocess, seeks to make full use of the human recourses, material and machines. Theobjective is to minimize the production costs. Effective scheduling can improve themanufacturing efficiency and make better use of the equipments, thus help thecorporation to be more competitive.Ant Algorithm is a novel parallel algorithm brought forth by an Italian scientistM.Dorigo, who was inspired by ants and other social insects' behaviors. AntAlgorithm mimics the cooperation of the natural ants. It can be easily applied and ofhighly robustness. Ant algorithm has been widely applied in various areas such asoptimization and industrial design.However, Ant Algorithm has one defect that when applied in large-scale optimizationproblems it is apt to be sucked in local optima. So this thesis chooses annealingalgorithm and ant algorithm to hydride in order to improve its performance.First the thesis introduces the scheduling problems, including the brief introduction,characters and some operation indexes. Then the thesis introduces the biologicalmechanism of the Ant Algorithm, the procedure of the algorithm, and expatiates theMax Min Ant System in detail. Next the thesis emphasizes on the detail of theoperation and how to choose parameters of the hydride Ant Algorithm. At last thealgorithms are tested against by 21 Rec FlowShop problems. Also, the topologycharacteristics and the performance are analyzed as well.The hybrid Ant Algorithm combines the traits of the Ant Algorithm and the annealingalgorithm, thus enhancing its searching capability. The hybrid Ant Algorithm issuperior in optimization ability, efficiency and liability. It also can be applied in theindustry.