Arquitetura de controle para gestão de recursos na agricultura de precisão
Data
2019-02-22
Autores
Pena, Ismael da Silva
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Esta tese consiste na investigação de um paradigma de controle para a gestão coordenada de recursos na agricultura em um contexto em que diversas unidades de produção, que buscam o máximo retorno econômico em cada ciclo de produção, são articuladas de forma a garantir a sustentabilidade ambiental e econômico-social de longo prazo. Para isso, utiliza-se um esquema de controle com dois níveis, em que no nível inferior são resolvidos problemas de controle ótimo com restrições para cada uma das unidades de produção e, no nível superior utiliza-se um novo paradigma de Controle Preditivo para promover a coordenação global no sentido de assegurar a sustentabilidade ambiental no longo prazo. A novidade deste esquema de Controle Preditivo com relação aos convencionais consiste no fato de que o modelo do sistema dinâmico associado não está definido a priori, mas surge das decisões de gestão e controle de cada uma das unidades de produção. Além do estudo de algumas propriedades importantes, demonstra-se a existência de solução global que resulta da convergência do esquema global, assim como a estabilidade assintótica do mesmo.
This Thesis consists of investigating a control paradigm for coordinated management of agricultural resources in a context where a number of production units, which seek the maximum economic return in each production cycle, are articulated to guarantee environmental, social and economic development. For this, a two-level control scheme is proposed, where at the lower level, optimal control problems with constraints are solved for each of the production units, and on the upper level, a new Model Predictive Control (MPC) paradigm is used to promote global coordination to ensure long-term environmental sustainability. The novelty of this MPC scheme over conventional ones is that the associated dynamic system model is not defined a priori, but arises from the management and control decisions made by each of the production units. Besides the study of some important properties for this problem formulation, it is demonstrated the existence of a global solution that results from the convergence of the global scheme, as well as of its asymptotic stability.
This Thesis consists of investigating a control paradigm for coordinated management of agricultural resources in a context where a number of production units, which seek the maximum economic return in each production cycle, are articulated to guarantee environmental, social and economic development. For this, a two-level control scheme is proposed, where at the lower level, optimal control problems with constraints are solved for each of the production units, and on the upper level, a new Model Predictive Control (MPC) paradigm is used to promote global coordination to ensure long-term environmental sustainability. The novelty of this MPC scheme over conventional ones is that the associated dynamic system model is not defined a priori, but arises from the management and control decisions made by each of the production units. Besides the study of some important properties for this problem formulation, it is demonstrated the existence of a global solution that results from the convergence of the global scheme, as well as of its asymptotic stability.
Descrição
Palavras-chave
Controle ótimo, Controle preditivo, Arquitetura de controle, Estabilidade assintótica, Agricultura de precisão, Optimal control, Model predictive control, Control architecture, Asymptotic stability, Precision agriculture