A carregar...
Publicação
Crescimento de raízes no solo baseado em modelos de enxame
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.17 KB | Unknown | |||
| 31.96 KB | Unknown | |||
| 23.65 KB | Unknown | |||
| 9.74 KB | Unknown | |||
| 16.25 KB | Unknown |
Autores
Resumo(s)
Recentemente, surgiram novas propostas de explicação para o crescimento de raízes que sugerem que estas apresentam uma forma de inteligência colectiva, proveniente dos comportamentos simples dos ápices que se baseiam em informações locais. A esta propriedade, é dado o nome de Swarm Intelligence. Com o objectivo de verificar a viabilidade de explicar o crescimento das raízes utilizando modelos de inteligência colectiva, no que se refere à exploração que efectuam do solo, foi criado um modelo do crescimento de raízes no solo discreto, programado em Java e acessível através de uma página de Internet. Este incorpora processos de escolha distribuídos que são realizados pelos ápices, que decidem, baseados em informação local, que nutriente explorar, se irão ramificar-se e em que direcção do solo crescer. O solo faz também parte do modelo, sendo composto por vários elementos cúbicos, dispostos num paralelepípedo. Cada um pode conter uma quantidade variável de água, azoto e fósforo, apresentando entre si processos de difusão simples. A raiz interage com o solo através do processo de uptake, que consiste no processo de aquisição de nutrientes do solo pela raiz. O modelo é controlado através de uma interface gráfica que apresenta um variado leque de sliders e checkboxes que permitem controlar os vários parâmetros envolvidos no crescimento dos ápices e apresenta como resultado a arquitectura tridimensional da raiz, passível de ser visualizada de várias formas. Com a observação das simulações realizadas através do modelo, verificou-se que as características gerais da exploração realizada por raízes eram replicadas por este. Concluiu-se, portanto, que, apesar de algumas limitações do modelo, a explicação das estratégias de exploração do solo pela raiz à luz de uma forma primitiva de inteligência é, efectivamente, viável.
Recently, new proposals have arisen regarding the explanation of root growth. These suggest that the roots present a type of collective intelligence derived from the simple behaviours of the apexes, which are based on local information. This property is named Swarm Intelligence. With the objective of verifying the viability of explaining root growth using models of collective intelligence, namely regarding the process of soil exploration, a discrete model of root growth in soil, programmed in Java and accessible via a website, was constructed. The model incorporates processes of distributed decisions that are made by the apexes, which decide, based on local information, what nutrient to explore, if they’ll branch and the direction of growth. The soil is also integral to the model, being composed of several cubic elements, arranged in a parallelepiped. Each cube contains a variable quantity of water, nitrogen and phosphorus, and interacts with its neighbours by processes of diffusion. The root interacts with the soil by the process of uptake, which consists in the acquisition of nutrients from the soil by the root. A graphical user interface, utilized to control the model, presents a vast array of sliders and checkboxes that allow the user to control the various parameters involved in the growth of the apexes. The result of the model is presented by a display of the simulated three-dimensional architecture of the root that can be visualized in a variety of ways. With the observation of the resulting simulations, it was verified that the model replicated the general characteristics of the exploration of the soil by the root. It was thus concluded that, despite some limitations of the model, the explanation of the soil exploration strategies by the root, in light of this type of intelligence, is indeed viable.
Recently, new proposals have arisen regarding the explanation of root growth. These suggest that the roots present a type of collective intelligence derived from the simple behaviours of the apexes, which are based on local information. This property is named Swarm Intelligence. With the objective of verifying the viability of explaining root growth using models of collective intelligence, namely regarding the process of soil exploration, a discrete model of root growth in soil, programmed in Java and accessible via a website, was constructed. The model incorporates processes of distributed decisions that are made by the apexes, which decide, based on local information, what nutrient to explore, if they’ll branch and the direction of growth. The soil is also integral to the model, being composed of several cubic elements, arranged in a parallelepiped. Each cube contains a variable quantity of water, nitrogen and phosphorus, and interacts with its neighbours by processes of diffusion. The root interacts with the soil by the process of uptake, which consists in the acquisition of nutrients from the soil by the root. A graphical user interface, utilized to control the model, presents a vast array of sliders and checkboxes that allow the user to control the various parameters involved in the growth of the apexes. The result of the model is presented by a display of the simulated three-dimensional architecture of the root that can be visualized in a variety of ways. With the observation of the resulting simulations, it was verified that the model replicated the general characteristics of the exploration of the soil by the root. It was thus concluded that, despite some limitations of the model, the explanation of the soil exploration strategies by the root, in light of this type of intelligence, is indeed viable.
Descrição
Tese de mestrado. Biologia (Bioinformática e Biologia Computacional). Universidade de Lisboa, Faculdade de Ciências, 2011
Palavras-chave
Raiz Solo Inteligência Colectiva Simulação Visualização Teses de Mestrado, 2011