|
Other FMEA
Sources
| |
Title: FMEA-based expert system for electrolysis diagnosis
Author: Berezin, AI; Polyakov, PV; Rodnov, OO; Yasinski, VL; Stont, PD
Abstract: Reduction of nonproductive expenditures at the cost of decreasing of total amount of sick cells is one of the actual targets. Expert diagnostics system is worked out for exposure of sick cells. Diagnostics process consists of the following procedures: Identification of cell "noises" and their classification as symptoms of a sickness with the help of specially trained neuron net; Type definition and calculation of seriousness of sickness according to expert rules with the help of fuzzy logics; Calculation of priority number of risks - PNR - according to FMEA technique (Potential Failure Mode and Effects Analysis) for each symptom, sickness, cell, potroom. Visualization and documentation of diagnostics results using FMEA protocols. Cell parameters and SPC data (Statistic process control) are the input information for the expert system of diagnostics. Diagnostics system is able to define PNR of 20 types of potroom "sickness", which allows: to reduce the number of sick cells; to reduce the frequency of AE; to define the maximum permissible condition of the cell for relining.
Source: LIGHT METALS 2005
Publication Year: 2005
Volume:
Issue nr:
Pages: 429 - 434
Science Code: Metallurgy & Metallurgical Engineering
Tags:
|