Multi-objective JSP Scheduling Based On OCTPN And Hybrid Genetic Algorithms

The present dissertation uses object-oriented colored timed Petri net (OCTPN) and hybrid genetic algorithms as tool to study the scheduling optimization of job shop with multiple process plants. The principal research results and contents can be outlined as follows:Firstly, combining colored Petri net and object-oriented technology, a new OCTPN method is proposed. Single-resource, dual-resource and multi-resource job shop scheduling models are studied using OCTPN. The OCTPN model is based on object-oriented technology, takes the parts of describing system essence and ability as different objects and takes the processes of describing system dynamic behavior as instances of objects, different instances have different colors. Model reusability is improved and model size is reduced greatly.Secondly, a new hybrid genetic algorithm by embedding Pareto ranking strategy into Pareto competition method is proposed for solving mubltiobjective scheduling problems. In order to promote solution diversity and to avoid falling in local optimal solution, the niche technology is used here. Many kinds of crossover operations and mutation operations are adopted to improve convergence rate. Pareto filter saves the optimum individual occurring in the course of evolution, which avoids losing the optimum solutions. The result has shown that the proposed algorithm is correct and excellent. Biobjective and multiobjective scheduling problems for dual-resource job shop are researched based on the algorithms.To the biobjective scheduling problem for dual-resource job shop, the objects of makespan and production costs are studied firstly; then the effect of worker staffing level and worker assignment on scheduling are studied, and finally dual-resource job shop dynamic scheduling is studied based on static scheduling:classify for machine breaks down, worker leaves, order cancels based on time and task, then decide whether carry out re-scheduling; different strategies are chosed according to different situations, and effective scheduling strategies are chosed in order to maintain throughput of job shop furthest.To the multiobjective scheduling problem for dual-resource job shop, the objects of makespan, production costs, machine maximal load, machine total loads and client satisfaction are studied according to the benefit of different departments of enterprise and anticipant goals of every department, and the optimal solution satisfying decision-maker is obtained from the optimal solution sets using the analytical hierarchy process.Finally, a job shop scheduling software is developed using Delphi and Flexsim.