Please use this identifier to cite or link to this item:
|Title:||Predicting mean time between failures of a maintained euipment using artificial neural network|
|Authors:||Oladokun, V. O|
Shodimu, M. O.
|Keywords:||Artificial Neural Network (ANN),|
Mean Time between Failures Multi-Layer Peceptrons (MLPs)
|Abstract:||In this study an attempt is made to use the Artificial Neural Network (ANN) model predict the Mean Time between Failures of manufacturing equipment. The equipment failure pattern was carefully studied and some key factors affecting the Mean Time between Failures were identified. An Artificial Neural Network model, the Multi-Layer perceptron, with two hidden layers and seven processing elements was built. Twelve months of maintenance data of the machine was collected. The data items were divided into 3 sets Training, Validation and Testing sets for analysis. Using the method of Back-Propagation, the Artificial Neural Network model was trained and tested. Eight input factors were identified; the output was classified into three Low, Medium and High Mean Time Between Failures. The analysis of the model confusion matrix indicates an overall model performance accuracy of 82% with a Normalised Mean Square Error of 0.24. It can be concluded that with the availability of adequate records, the Artificial Neural Network can serve as a useful maintenance management tool.|
|Appears in Collections:||scholarly works|
Files in This Item:
|(12)ui_art_oladokun_predicting_2010.pdf||486.01 kB||Adobe PDF|
Items in UISpace are protected by copyright, with all rights reserved, unless otherwise indicated.