Control Of Weed Threshold Using Artificial Neural Networks

Cereal production is a function of several parameters. Weeds play an important role according to their density and the period in which they proliferate. Their spatial distribution in terms of wheat competition is also a factor that should not be overlooked. The use of pesticides the adaptability of crops the agricultural mode the climatic impact the nature of the soil the weed species and the crop species are combined parameters. What characterizes this environment is complexity and imprecision. This study proposes an intelligent analysis system with artificial neural networks. The proposed network has three layers. An input layer comprising the variables that enter the process a hidden layer and an output layer express the wheat yield. It is a question of making the correspondence between the two spaces of inputs and output starting from the real values measured. This phase represents the learning of the network. Once the transfer function is established it will be possible to predict the yield of a plot from the conditions and the growing environment.