Multi-Objective Integration Of Collaborative Planning And Scheduling And Its Evolutionary Algorithm

Planning and Scheduling are the most important subsystems of manufacturing system. In traditional manufacturing system, planning and scheduling are divided as two serial stages, making planning first then generating scheduling schema. This approach causes several problems such as infeasible planning in the actual scheduling, planning that difficult to update, and unbalanced load, the scheduling with collaborative plan under the circumstance of supply chain is also not considered. What’s more, researches focus on single objective, however, in actual production several objectives need to be considered. So it’s significant to study on multi-objective integrated collaborative planning and scheduling.There are two approaches to implements integrated planning and scheduling, one is approach based on Agent, the other is approach based on algorithm. Because of high feasibility, approach based on algorithm becomes the main way to implement integrated planning and scheduling. In many algorithms, evolutionary algorithm is the first choice to implement integrated planning and scheduling for its advantages of easy integration with models, good convergence, and excellent efficiency for solving complex problems.In this paper, multi-objective integrated collaborative planning and scheduling problem is the research object. First, a novel model of multi-objective integrated collaborative planning and scheduling is proposed. Second, algorithms based on multi-objective evolutionary algorithm to implements model are designed, and their performances are evaluated overall.1. Propose multi-objective integrated collaborative planning and scheduling model. Since that collaborative plan, manufacturing plan and the scheduling solutions are hard to be paralleled and synchronized, and considering the scheduling with collaborative plan under the circumstance of supply chain, a novel model of integrated collaborative plan and scheduling is proposed, taking four criterions of dead schedule time, total cost, total delay time, average flow time which are mostly used in supply chain coordination as standards.2. Algorithm based on population-based evolutionary algorithm to implement model. For the disadvantage that traditional aggregating function approaches are difficult to find non convex solutions when they deal with multi-objective optimization problems, a population-based multi-objective evolutionary algorithm is designed to implement multi-objective integrated collaborative planning and scheduling. Making collaborative planning and scheduling schema synchronously and optimization of the whole model are realized by using this algorithm. 3. Evolutionary algorithm based on Pareto Optimal to implements the model. Because that the best solutions searched by population-based multi-objective evolutionary algorithm are not global optimal and speciation problem, a multi-objective evolutionary algorithm based on Pareto Optimal to implement multi-objective integrated collaborative planning and scheduling model is designed. As well as implements the model, the algorithm is evaluated from the two aspects of optimize performance and convergence. It is proved that this algorithm is effective and excellent.