Research On Multi-shop Multi-Objective Integrated Scheduling Algorithm

The components or parts of a product may be distributed and produced at various workshops in a collaborative way for the purpose of expanding capacity and reducing costs. Therefore, the processing mode of multi-shop collaborative manufacturing has become absolutely essential for some manufacturing enterprises. If the scheduling plan is rational, the processing mode of multi-shop collaborative manufacturing will reduce idle rate of equipment, balance the processing load of multiple workshops, minimize the completion time, and improve the profits of enterprises.Integrated scheduling problem is diffrent from JSP and FSP. It is a new third catogory of scheduling problems.There is very few research on multi objective in existing documents about integrated scheduling problem. The multi-shop scheduling problem has been researched by a few documents, but there is little research on the transportation distance of the workshops.Aiming at traditional methods for integrated scheduling optimized only one objective, a multi-objective flexible scheduling algorithm with four objectives has been proposed. The proposed algorithm based on priority rule sequence. One priority rule sequence correspond to one scheduling solution. In order to provide for decision maker choice from among the different pareto solutions, the analytic hierarchy process approach is used to select the optimal compromise solution. Analysis and instance show that the implementation of algorithm achieve the product in the shorter processing time and the less costs.Aiming at existing methods for multi-shop integrated scheduling used transportation frequency in lieu of transportation time, a dual-objective particle swarm optimization algorithm for multi-shop integrated scheduling problem has been proposed. In order to avoid the problem of direct encoding method generates a large number of illegal particles, a method based on the priority of the processes has been designed, this method can ensure that each particle is a legitimate individual; In order to improve the efficiency of the non-dominated sorting, an improved Pareto non-dominated ranking method has been proposed; To simplify the particle's location update operation, the update equation is expressed without speed. To simplify the update operation of archive population, an archive management method based on retention interval and a global guide particle selection method based on the frequency of retention interval have been proposed.Instance data in the existing integrated scheduling algorithm is expressed with tree graph. It is not very convenient. The existing integrated scheduling algorithm is lack of analysis of benchmark instances.