Grouping & Optimization of Structural Design Using FCM Algorithm

Abstract The basis for grouping structural elements is a tradeoff between optimization in design and construction. The grouped elements should have common optimal design specifications that are structurally safe and convenient for construction. Fuzzy clustering provides flexibility to regroup the elements using membership values indicating their affinity to a cluster for the considered parameter. The proposed Fuzzy C-Means algorithm demonstrates grouping efficacy based on a dominant design parameter and also provides leverage for regrouping of an element considering other design or construction aspects. The method is evaluated by applying it to footings of four buildings with varying complexity, taking load on footing as basis for grouping. Distinct and homogeneous clusters are obtained with small computational effort, with further scope for regrouping members based on other parameters using their membership values for the parameter considered. The paper also proposes optimization methodology for rebar design of Reinforced Cement Concrete structures with fixed topology and cross-sectional dimensions.