Supervisão de sistemas de produção baseada em redes de petri.

Abstract

This work cleals with the supervision of production systcms using Petri Nets. In this dissertation productions systcms refer to both, manufacturing systcms and batch production systcms. The main problcms tacklcd are rclatcd to the integration and adaptability of such systcms to changes in the operational requirements. This changes may be motivated by rnarket demand, planning requirements, availability of raw materiais, among others. Furthcrmore, we take into account fault tolcrance aspeets as wcll as the verification of production orders for manufacturing systcms and recipes for batch production systcms. Duc to the adaptations, the original structure of the model and the supervisor may nced to be changed. On the other hand, the preservatioh of such structures is important so that it promotes the adaptation in a agile way. As wcll as other kind of systcms, production systcms have to be fault tolerant in order to guarantec productivify. Thcrcforc, fault tolerancc aspeets as well as agility are taken into account in this research. During the design phasc, onc has to prove that the production systems are bounded. In the case of manufacturing systcms boundness properties can be used, for example, to adequate buffers and storages in the shop floor, that may lead, for example, to cost reduction. In the case of batch production systems, ovcrflows must be avoided, because they may result in production loss. In order to avoid such undesired situations, it is important to prove that the recipe is bounded. In the research wc introducc a systcmatic and structured modeling procedure to obtain both, the model of the system and the supervisor for production systcms. The introduced solutions allows the generalization of the structure of the supervisor, therefore promoting higher levei of adaptability in case of changes in the operational requirements.