Full text
| |
| Author(s): |
Mariana Antonia Aguiar
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Escola Politécnica (EP/BC) |
| Defense date: | 2001-06-18 |
| Examining board members: |
Zsolt Laszlo Kovacs;
Vitor Heloiz Nascimento;
José Sotelo Junior
|
| Advisor: | Zsolt Laszlo Kovacs |
| Abstract | |
This scientific research presents the study and development of Automatic Diagnostic System (ADS) of failures for complex electromechanical equipments based on technology of Artificial Neural Networks. After the research and revision of available literature on detection and diagnosis of failures, the architecture of ADS is proposed. This architecture is constituted by available database of a generic equipament and by the Articial Neural Networks apllied in order to getting the best diagnosis. As study of case for this research a dispenser machine of ballots found in bank agencies was used. The choice of this machine occurred due to the complaints of its users on wastefulnesses of time and parts changed without necessity during the maintenance of the equipment. The difficulty in getting a reliable diagnosis without a tool of appropriate analysis is the main cause of the deceits committed for the technicians of support and maintenance. The Automatic Diagnosis System helps the technicians providing a guided diagnosis of failures of the dispenser machine of ballots base on the historical analysis of sensors and on the knowledge of functioning of the machine which is analysed and solved by Artificial Neural Networks. The effectiveness of the Diagnostic Automatic System is shown through the Curves of Performance in the end of this work. (AU) | |