8 Ekim 2015 Perşembe

The Use of Artificial Neural Networks to Assess the Capacity of Transport Measures


In the area of logistics management both managers and engineers rely primarily on proven computational
algorithms, for this reason, it is often difficult to convince them to the use of artificial neural networks in solving decision problems. The paper presents the possibilities of using the FANN library in building of a computer application applied in the area of logistics. The possibilities of the component are presented on the example of applications of artificial neural networks to estimate the capacity of transport vehicles based on their dimensions. The example presented in the work was solved with the use of a multi-network Layered Perceptron. The example depicted not only the possibility of using artificial neural networks for solving poorly structured tasks but also practical application of the TFannNetwork component.

Fast Artificial Neural Network library (FANN) is an open-source project that implements a
multi-layer one-way neural network networks with support for both full and weakly
connected networks [14]. FANN is easy to use, comprehensive, well documented and fast in
acting [15], [16]. There are links to over 15 programming languages, including Delphi 7
TFannNetwork is a component of Delphi (created by Pereira Maia) that connects the
application with FANN library. Undoubtedly it is not necessary to install TFannNetwork to
use FANN in Delphi but the component makes the library more friendly for Delphi
environment [14].

The Use of Artificial Neural Networks to Assess the Capacity of Transport Measures

Artur Duchaczek / Dariusz Skorupka

General Tadeusz Kościuszko Military Academy of Land Forces Faculty of Management


0 yorum:

Yorum Gönder