Analysis of an Intelligent Hybrid System for Fault Diagnosis in Cracked Structure

To develop a robust fault diagnostic tool based on genetic algorithm and fuzzy logic, the current research explores the use of dynamic responses of cracked and intact cantilever beam structure. Theoretical, finite element and experimental analyses were carried out to find the combined impact of crack locations and crack depths on the vibrational characteristics (natural frequencies, mode shapes) of the cantilever beam. The calculated vibration signatures are used to design and train the GA-fuzzy controller. The inputs to the GA layer of the hybrid system are the first three relative natural frequencies and the first three relative mode shape differences. The interim outputs from the GA controller are the first three relative natural frequencies and the first three relative mode shape differences that are used as inputs to the fuzzy segment of the proposed hybrid technique. The final outputs from the developed hybrid system are rcl1_final, rcd1_final, rcl2_final and rcd2_final. The viability of the proposed technique has been investigated both analytically and experimentally for the cantilever beam containing multiple cracks. From the analysis of the results obtained from theoretical, finite element, GA-fuzzy controller and experimental analysis, the applicability of the proposed method for multiple cracks diagnosis is discussed.