A Method for Prediction of Waterlogging Economic Losses in a Subway Station Project

In order to effectively solve the problems of low prediction accuracy and calculation efficiency existing methods for estimating economic loss in a subway station engineering project due rainstorm flooding, new intelligent model is developed using sparrow search algorithm (SSA), least-squares support vector machine (LSSVM) mean impact value (MIV) method. First, this study, 11 input variables are determined from disaster rate asset value, complete method provided acquiring processing data all variables. Then, SSA method, with strong optimization ability, fast convergence few parameters, used optimize kernel function penalty factor parameters LSSVM. Finally, MIV identify important variables, so as reduce predicted achieve higher accuracy. addition, 45 projects China were selected empirical analysis. The results revealed that linear correlation between output was weak, which demonstrated necessity adopting nonlinear analysis such Compared other forecasting methods, multiple regression analysis, backpropagation neural network (BPNN), BPNN optimized by particle swarm optimization, SSA, LSSVM, LSSVM genetic algorithm, PSO-LSSVM Grey Wolf Optimizer, proposed paper had stability rainstorms.