Abstract: Numerous strategies have been created and concentrated to distinguish harm through the difference in powerfulreaction of a construction. Because of its ability to perceive design and to correspond non-straight and non-one of a kindissue, Artificial Neural Networks (ANN) have gotten expanding consideration for use in distinguishing harm in structuresdependent on statically modular boundaries. Best works detailed in the utilization of ANN for harm location are restrictedto statically models and little controlled exploratory models as it were. This is a direct result of the two primary limitationsfor its reasonable application in identifying harm in genuine designs. They are: 1) the unavoidable presence ofvulnerabilities in vibration estimation information and limited component demonstrating of the construction, which mayprompt wrong forecast of primary conditions; and 2) colossal computational exertion needed to dependably prepare anANN model when it includes structures with numerous levels of opportunity. Along these lines, most uses of ANN in harmlocation are restricted to structure frameworks with few levels of opportunity and very critical harm levels. The unwaveringquality and productivity of this strategy is shown utilizing both mathematical and test models. Likewise, a parametric reportis completed to research the affectability of the proposed strategy to various harm levels and to various vulnerability levels.As an ANN model requires colossal computational exertion in preparing the ANN model when the quantity of levels ofopportunity is generally huge, a sub structuring approach utilizing multi-stage ANN is proposed to handle the issue.Through this technique, a construc