Valterlin da Silva Santos

Um Modelo de Otimização Multiobjetivo para Análise de Sistema de Recursos Hídricos: II Aplicação

This work involves the application of a multiobjective optimization model based on sequential linear programming to a two-reservoir system connected in sequence and a 500 ha irrigated perimeter, located in the semiarid region of Paraiba state. The water demands are for urban water supply, irrigation and the regulation of river flow downstream from the reservoirs. The following objectives were considered to establish an optimal operation: to meet the water demands for urban supply, regulated flow downstream from the reservoirs and the reservoir target volumes, as well as maximizing net profit and manpower from agricultural activity. Scenarios that vary the priorities of different objectives were created in order to evaluate the model response regarding the consistency of results and required computation time. The results have shown that the objectives were fulfilled in accordance with the assigned priorities and the problem constraints were met. The computation time and the number of iterations can be considered satisfactory (19 seconds and 5 iterations, on average, for a problem having 3802 decision variables and 874 constraints when considering 9 years of system operation). The results attained for system operation can support the generation of operational rules for water allocation in order to meet the required demands, objectives and physical constraints. Key-words: Water Resources, optimization, linear programming, linearization.