Intelligent Algorithms For 3D Printing Job-shop Scheduling Problems In Personalization Production

Personalization production not only satisfies the functional needs of customer products,but also creates higher added value.Thus,personalization production has become a way for manufacturing companies to respond to the fierce competition in the global market.Under this circumstance,3D printing is becoming more and more significant,and has become an important part of advanced manufacturing systems.The industry scale of 3D printing is growing rapidly,as it is widely used in industrial production,medical care,education and other fields.With the growing use of 3D printing technologies and the maturity of 3D printing services,the development of production planning,scheduling and control of 3D printing combining with smart manufacturing has become a critical issue that needs to be solved to achieve personalization production,which aims to reduce production costs,improve production efficiency of 3D printing,meet individualized needs,and promote the application of 3D printing by providing an optimized scheduling plan.To this end,this research first fully studies and analyzes the scheduling problem of 3D printing workshop in literature,then builds its mathematical model.Thereafter,based on the process characteristics,the optimization methods for the scheduling problems in parallel machine environment,complex simulation environment and dynamic environment are studied.The specific contents are shown as follows:(1)The development trend of manufacturing paradigms towards personalization production is addressed.The relationship between personalization production,current advanced manufacturing systems and advanced manufacturing technologies is analyzed.And the significance of 3D printing in personalization production is clarified.Further,the research progress and importance of the 3D printing workshop scheduling problems is discussed.And research direction and contents of this thesis are given.(2)The production process of 3D printing is analyzed in deep.And a brief overview on the characteristics of commonly used 3D printing technologies is conducted.The key characteristics of 3D printing job-shop are summarized.The workflow of 3D printing jobshop is analyzed.Based on the scheduling theory,a mathematical model of 3D printing jobshop scheduling problem is presented.(3)For the scheduling of 3D printing job-shop with parallel identical 3D printers environment,an improved genetic algorithm based on a two layer coding is designed to tack into account the jobs allocation and parts packing problems integrally.The no-fit raster method is introduced to deal with the irregular shapes packing problem and ensure a proper spacing among parts in a batch.Combining the theory of genetic algorithm with the characteristics of production process,a new initialization strategy and a local search strategy are introduced to improve the efficiency of the scheduling algorithm.Finally,the effectiveness of the algorithm is verified by numerical simulation experiments.(4)As the objective function is complicated and time-consuming in simulation-based scheduling problems in 3D printing job-shop,a Bayesian discrete optimization algorithm is proposed.A discrete distance measure method is presented to adopt Bayesian optimization in discrete domain,and a new kernel function is developed based on the characteristics of scheduling problems.To improve the efficiency and scalability of the algorithm,a sparse GP model is further developed with a simulated annealing to select its inducing points.Numerical simulation experiments show that the proposed algorithm has better optimization ability in complex problems.(5)To solve the dynamic,complex and diverse problem in 3D printing job-shop due to personalized,diversified,single one and small batch production,this research introduces a RFID-based workpiece identification and tracking method to achieve a real-time status perception of both jobs and workshop.Then,a complex event process method is presented to define the original events,simple events and complex events,and to excavate the casual,temporal or spatial correlation between events based on business processes.A dynamic tracking and response job-shop is built,and the dynamic scheduling is achieved based on the complex event processing technology.(6)A 3D printing job-shop scheduling prototype system is set up to verify the effectiveness of the proposed method and theory.Specifically,the study includes the application background,design principles,and hierarchical architecture of the system,and in particular the core module functions of the system are described to verify the feasibility of the prototype system.