Prediction of the Yield Performance and Failure Mode of RC Columns under Cyclic-Load by PSO-BP Neural Network

The yield performances and failure mode of reinforced concrete (RC) columns, which are critical structural to the design research engineering structures, have a significant impact on dynamic response, performance level, seismic ductility. traditional empirical theoretical method used predict leads large dispersions in most cases. To better estimate RC this paper developed novel neural network method. Empirical models determine input parameters by analyzing factors that affect rationality these is verified sensitivity analysis. back-propagation (BP) was adopted. influence number hidden neurons studied improve model accuracy. Comparative analysis revealed prediction results good agreement with experimental more accurate than other models. initial were optimized using particle swarm optimization (PSO), has been proven be superior genetic algorithm (GA) sparrow search (SSA) methods terms effectiveness computation time. high generalization ability calibrated test validation sets another eight additional data. proposed provides new way under actions when data insufficient.