Detection Of Biomass In Lake Areas Using Artificial Intelligence: Applying a study case-Brates Lake, Galati, Romania

Abstract

Biomass continues to be the main source of renewable energy in the EU, accounting for almost 60%, with the largest end-user of biomass being the heating and cooling sector, which uses about 75% of all biomasses. The cartographic activity of the area of the lake areas supposes the elaboration of the maps, of the topographic plans and the identification of the surfaces that represent biomass sources. At the same time, it is necessary to inventory the vegetation present in the mapped areas, to follow the productivity cycle of the crops that will be used as biomass, to identify possible disturbing factors of crops (temperature, humidity, existence of pollution sources, etc.). To detect biomass, we can use convolutional neural networks (CNN), coupled with a special type of feedforward neural networks that consist of several convolution layers and grouping layers. A feedforward neural network is an artificial neural network in which connections between nodes do not form a loop. These networks, which perform patterns like the activity of neurons in the brain, are generally presented as systems of interconnected processing units (artificial neurons) that can calculate values from inputs, resulting in an output that can be used in later units. Artificial neurons are basically processing units that compute some operations on several input variables and usually have an output calculated by the activation function. For this research, an experimental design using artificial intelligence was made to valorize it energetically.