Contribuições à análise de desempenho de células de manufatura baseada na teoria de controle supervisório

• Ano de defesa: 2012
• Autor(a) principal: Kawamura, Celso Matsumi
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Tecnológica Federal do Paraná; Cornelio Procopio; Programa de Pós-Graduação em Engenharia Elétrica
• Palavras-chave em Português: Controle de processo; Sistemas de tempo discreto; Sistemas de controle supervisório; Process control; Discrete-time systems; Supervisory control systems
• Link de acesso: http://repositorio.utfpr.edu.br/jspui/handle/1/784

The growing competition among manufacturing industries causes companies to worry more and more with issues related to quality, productivity and cost reduction. International competition, quite incisive among raw material processing manufacturers, is actually a very concerning situation. To become more competitive, enterprises have to produce more with the same amount of raw materials, or in a shorter time, making the constraints of time and resources increasingly severe. A better performance and more efficient processes can be obtained by means of industrial automation. However, such projects are normally designed without any formal approach and the achieved results depend, more than anything else, on engineers’ expertise. It cannot be stated that a machine set is working without any deadlock situation, or is producing in the most efficient way possible, considering operations´ execution time as efficiency criteria. By means of formal approaches such as Supervisory Control Theory (SCT), it is possible to synthesize supervisors automatically, that allow a plant to operate, in a way that, not only safety requirements are fulfilled, but the system is also prevented to reach a deadlock situation. This work presents this theory and its extensions, so the optimal supervisor satisfies task scheduling requirements and job completion deadlines, by using timed automata frameworks. The job shop scheduling problem, which directly relates to manufacturers´ production planning, is an optimization problem, where all the processing steps are organized in way that the total processing time is minimized. A methodology is presented for the job shop scheduling, by means of timed automata and supervisory control theory. An algorithm for acquiring an optimal operation sequence is proposed, considering the total time needed to finalize a process as the optimization criteria. The method is applied in an automated cell in a elevator manufacturing industry, where it was possible to obtain the minimal time needed to process a certain part. Through algorithms and methods for job shop scheduling in industrial environments, it is intended to meet the manufacturers´ strategic needs, in their search for manufacturing systems that are increasingly more efficient and competitive.