Adaptive Neuro-Fuzzy Inference System in Structural Damage Assessment

Eleni Vrochidou, Petros Fotios Alvanitopoulos, Ioannis Andreadis, and Anaxagoras Elenas


ANFIS, earthquake damage classification, seismic parameters, damage indices


The non-linear dynamic analysis of constructions is a time consuming process. This paper proposes a method based on an Adaptive Neuro-Fuzzy Inference System (ANFIS) for automatic seismic structural damage classification. A set of 200 artificial accelerograms is utilized. The severity of these seismic excitations is described by 20 widely used seismic parameters. The global structural damage index Maximum Inter-Storey Drift Ratio (MISDR) is used as metric to quantify and classify the structural damage caused to a certain structure, into 4 categories (negligible, weak, moderate, severe). Results indicate that the proposed model is suitable for adaptive predictive control of the behavior of the construction for seismic signals and its accuracy rates up to 89.5%. The proposed method is compared to previous reported fuzzy techniques. Further study attempts to decrease the number of parameters used and maintain the same level of classification performance.

Important Links:

Go Back