/img alt="Imagem da capa" class="recordcover" src="""/>
Trabalho de Conclusão de Curso - Graduação
Predição do consumo de água por meio de redes neurais artificiais: um estudo de caso em Belém-PA
Water scarcity is a serious and constant problem in many cities in Pará and Brazil. In this scenario, the various techniques of Artificial Intelligence, especially Artificial Neural Networks, appear as alternatives capable of assisting in the planning and management of Water Supply Systems. Among...
Autor principal: | ROCHA, Jessé da Costa |
---|---|
Grau: | Trabalho de Conclusão de Curso - Graduação |
Publicado em: |
2019
|
Assuntos: | |
Acesso em linha: |
http://bdm.ufpa.br/jspui/handle/prefix/1335 |
Resumo: |
---|
Water scarcity is a serious and constant problem in many cities in Pará and Brazil.
In this scenario, the various techniques of Artificial Intelligence, especially Artificial Neural
Networks, appear as alternatives capable of assisting in the planning and management
of Water Supply Systems. Among the many possible approaches, we opted to analyze
variations in water consumption as a function of climate (temperature, humidity and
rainfall). This choice is justified because the climate is among the factors that most
influence the consumption of water and also because there are still few studies published in
this line of research. In this perspective, the main objective of this work was to elaborate
an Artificial Neural Network to predict the water consumption in a Supply System located
in the city of Belém-PA, and the main motivation was to stimulate the collection and
storage of water consumption, as well as the use of this data in intelligent systems capable
of supporting the decisions of the water resources managers in the city of Belém and in
the State of Pará. After four experiments it was concluded that the best architecture and
configuration of RNA to solve the proposed problem is a network with a single hidden
layer with 5 neurons and sigmoidal activation function, and an output layer with a neuron
with linear activation function. |