Bayesian approach and time series dimensionality reduction to LSTM-based model-building for fault diagnosis of a reciprocating compressor.
Diego CabreraAdriana GuamánShaohui ZhangMariela CerradaRené-Vinicio SánchezJuan CevallosJianyu LongChuan LiPublished in: Neurocomputing (2020)
Keyphrases
- fault diagnosis
- dimensionality reduction
- expert systems
- neural network
- principal component analysis
- operating conditions
- bp neural network
- high dimensional
- fuzzy logic
- condition monitoring
- power transformers
- feature extraction
- fault detection
- non stationary
- pattern recognition
- fault detection and diagnosis
- gas turbine
- rbf neural network
- recurrent neural networks
- industrial systems
- multiple faults
- electronic equipment
- power plant
- data points
- monitoring and fault diagnosis
- failure diagnosis
- fault identification
- rotating machinery
- fault diagnostic
- chemical process
- feature selection
- fault detection and isolation
- multi sensor information fusion
- analog circuits
- feature space
- back propagation
- fault tree
- steam turbine
- tennessee eastman
- artificial neural networks
- genetic algorithm