Optimization On Scheduling Multi-type Printed Circuit Boards In A SMT Assembly Line

Printed circuit board(PCB) is the circuit board on which a number of electronic are located at predefined positions and wired through electrical connections.The scheduling of PCB belongs to typical unisonous scheduling problem on production logistics in manufacturing field.Scheduling the production of printed circuit boards(PCBs) in electronics manufacturing has been a subject of intense research in recent years,primarily motivated by the need to develop newer and efficient methods to enhance productivity. PCB is assembled by surface mount manufacturing system.This system is an advanced manufacturing system,which has characteristics of surface mount technology(SMT), flexible manufacturing systems and computer integrated manufacturing system.PCB assembly is under high-mix and low-volume production environment,which must offer high product,variety and satisfy the dynamic product demand.Its objective is to maximize the output with the lowest cost and smallest defect rate in the shortest possible time.This trend emphasizes the importance of optimizing the SMT assembly processes.The assembly time to consider consists not only the placement time,but also the setup time and feeder exchange time between different types of PCB in the assembly line.Therefore,how to solve the problems of job grouping,components allocation on machines,feeder setup strategies and the PCB production sequence in the assembly line can affect the productivity of PCB.The dissertation mainly solved the production scheduling problem of multi-type and small-batch environment in an assembly line:the objective considered is to minimize the mean flow time.SMT production line was systematically introduced in this dissertation.A combination of group setup strategy and partial setup strategy is employed in this dissertation to optimize the scheduling of PCBs of various types and small quantities on the SMT assembly line composed by the high-speed placement machine and multi-function placement machine.A mathematical model of production scheduling problem was established in accordance with the actual problem of surface mount production line.And the production scheduling problem in the assembly line was decomposed into four subproblems:job grouping problem,components allocation on machines,feeder setup,and the sequence of PCB production.The relative similarity coefficient method was proposed to solve the job grouping problem,and the grouping result was compared with that of other group strategies.The new grouping strategy outperformed existing methods of grouping.The KTNS(keep tool needed soonest) strategy was applied to determine the feeder setup for a given sequence of groups to be processed after the boards were assigned to different groups.Finally,the reactive tabu search algorithm was used to sequence PCB groups at two levels:the outside level and inside level.At th outside level an outside tabu search algorithm was applied to move from one group sequence to another,and at the inside level an inside tabu search algorithm was invoked to move,within each group,from one board type sequence to another.Through simulation experiments,it showed that the proposed algorithm could find a better solution.