Affiliations: Departamento de Expresión Gráfica y Cartografia, Universidad de Alicante, Alicante, Spain | Departamento de Ciencia de la Computación e Inteligencia Artificial, Universidad de Alicante, Alicante, Spain
Note: [] Corresponding author. Leandro Tortosa, Departamento de Ciencia de la Computación e Inteligencia Artificial, Universidad de Alicante, E–03080 Alicante, Spain. Tel.: +34 966903900; E-mail: [email protected] (L. Tortosa), [email protected] (J. L. Oliver), [email protected] (J. F. Vicent).
Abstract: A 2D triangle mesh simplification model is described in this paper, with the main objective of preserving the shape of the original mesh. The proposed model consists of a self-organizing algorithm whose objective is to generate the positions of the nodes of the simplified mesh; afterwards, a triangulation algorithm is performed to reconstruct the triangles of the new simplified mesh. The self-organizing algorithm is an unsupervised learning algorithm that provides a set of nodes representing the best approximation of the original mesh. An adaptation of the neural network algorithm is proposed with the primary objective to work in the context of urban transport networks. We verify the effectiveness of this model through the design and development of some urban network problems. Specifically, the algorithm is applied to two real problems, the first one is the design of a tramway network in a town, and the second one is the design of an information point network within a real bus transport network.
