Beschreibung:
This book discusses white- and black-box approaches to fault diagnosis in condition monitoring, delivering a thorough evaluation of the latest artificial intelligence tools. It addresses nearest-neighbor-based, clustering-based, statistical, and information theory-based techniques, considering the merits of each technique as well as the issues associated with real-life application. It covers classification methods, from neural networks to Bayesian and support vector machines. It proposes fuzzy logic to explain the uncertainties associated with diagnostic processes. It also provides data sets, sample signals, and MATLAB code for algorithm testing.
Massive Field Data Collection: Issues and Challenges. Condition Monitoring: Available Techniques. Challenges of Condition Monitoring Using AI Techniques. Input and Output Data. Two-Stage Response Surface Approaches to Modeling Drug Interaction. Nearest-Neighbor-Based Techniques. Clustering-Based Techniques. Statistical Techniques. Information Theory-Based Techniques. Uncertainty Management.