Contribució en la modelització de la rugositat superficial obtinguda en els processos de honing utilitzant xarxes neuronals artificials
- Autores: Mauricio Sivatte Adroer
- Directores de la Tesis: Joan Vivancos Calvet (dir. tes.), Francesc Xavier Parra Llanas (codir. tes.)
- Lectura: En la Universitat Politècnica de Catalunya (UPC) ( España ) en 2013
- Idioma: español
- Tribunal Calificador de la Tesis: Fernando Romero Subirón (presid.), Andreu Català Mallofre (secret.), Guillermo Reyes Pozo (voc.)
- Calificación: Sobresaliente
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
From experimental data obtained from a test machine built for the purpose, the objective of this thesis is to model the process of burnished steel inner cylinder (honing process) using artificial neural networks.
We selected the five process variables that are considered relevant and defined the multilayer perceptron with back propagation learning algorithm and cross-validation as validation system, as a neural network base.
In a first step a neural network model is designed so that data from the selected process, given the parameter roughness "Ra" that correspond.
In a second stage models the inverse problem. As the value of surface roughness "Ra" is desired, the neural network model explicitly the values of the process variables that would make it possible.
Finally, we designed an iterative model using selected networks in the previous phases, to improve inverse modelling.
