Vibrodiagnostics of Pre-Stressed Structures : a Pattern Recognition Based Approach

The following paper presents a pattern recognition approach to the condition monitoring of technical objects. The Noise-Assisted Feature Subset Evaluation (NAFSE) method addressed for the extraction of diagnostic parameters was used. The NBV-based classifier conducted the final recognition of the object’s condition on the basis of the diagnostic parameters obtained from the NAFSE method. The effectiveness of the proposed pattern recognition approach is illustrated by effective vibrodiagnostics of pre-stressed concrete structures. The approach shows potential for the damage detection of pre-stressed concrete structures at an early stage of damage development.