On hierarchical routing control in discrete manufacturing plants

The interest in automating demanufacturing processes has constantly grown inthe recent years, especially for electronic products. The main feature required bythis application is high flexibility due to the uncertain configuration of the productsand the variability of the operations. Specifically, the problem of routingpallets across a plant is considered here. The purpose of this thesis is to improveand implement a new approach to address this problem: the HierarchicalRouting Control (HRC). It is based on Model Predictive Control strategy actingat supervisory level, combined with a low-level path following approach. A newsolution for the path generation problem and path assignment redundancy problemis proposed and additional constraints are introduced in the plant model vianew specific sub-routines. The conditions for lockout detection are identified andtheir implementation allows an Intelligent Move Blocking. Then the optimizationproblem is reformulated and sub-optimal solution methods are defined to reducethe computational time. A data exchange protocol is added to control algorithmin order to communicate with the integrated low-level logic that directly controlsthe pallet handling system. Finally the developed model is verified to be consistentwith a real plant and the HRC strategy reaches high performances with verylow computational cost.