PCB assembly scheduling with alternative nozzle types for one component type

Abstract

The placement machine is the bottleneck of a printed circuit board (PCB) assembly line. The type of machine considered in this paper is the beam-type placement machine that can simultaneously pick up several components from feeders. It is assumed that the number of nozzle types (NTs) is less than the number of heads on the beam. The objective of the PCB assembly scheduling for a single placement machine is to minimize the cycle time based on the average machine operation time instead of the travelling distance. To minimize the cycle time, the number of turns and the number of pickups should be minimized. The PCB assembly scheduling is hierarchically decomposed into four problems: the nozzle assignment problem, the head allocation problem, the component type (CT) grouping problem and the pickup clustering problem, which are optimized successively and iteratively. First, the nozzle assignment problem considering alternative NTs for one CT is dealt with by the proposed genetic algorithm. For a given nozzle assignment solution, the head allocation problem is solved by a previously greedy heuristic to minimize the number of turns.Then, the CT grouping problem and the pickup clustering problem are solved by a proposed greedy heuristic and a modified agglomerative hierarchical clustering approach, respectively, to minimize the number of pickups. Numerical experiments are carried out to examine the performances of these proposed heuristic approaches. The importance of considering alternative NTs for one CT for the cycle time is also confirmed.

An MILP model and a hybrid evolutionary algorithm for integrated operation optimisation of multi-head surface mounting machines in PCB assembly

“…Among the five sub-problems, the assignments of nozzles and components to heads are usually considered separately to balance the workloads over heads or to minimize the number of nozzle changes (Guo et al, 2011, Ashayeri et al 2011, which can indirectly reduce the total placement time. However, their solutions only provide the number of components (workload) and the set of nozzle types assigned to each head, not the details of nozzle types and components in each assembly cycle.…”

“…Nozzle assignment is to assign nozzle types to heads to grasp different types of components, with the aim of minimizing the number of nozzle changes and balancing the workloads over heads. For the problem, constructive heuristics can be found in previous researches (Raduly-Baka et al, 2008;Guo et al, 2011;Knuutila et al, 2007;Ashayeri et al, 2011). Its mathematical models have been developed by Torabi et al (2013) and Ashayeri et al (2011) under different machine constraints.…”