遗传算法改进的BP神经网络对汛期三门峡水库泥沙冲淤量的计算_刘媛媛.pdfVIP

24 4 Vol. 24 No.4 2005 8 JOURNAL OF HYDROELECTRIC ENGINEERING Aug. , 2005 BP 刘媛媛, 练继建 ( , 300072) : BP , BP , BP BP , BP , BP : ; ; BP ; ; + : TV145 11 : A Calculation of sediment and scour of Sanmenxia reservoir in flood season by the BP neural network improved by GA LIU Yuanyuan, LIAN Jijian ( School fo Civil Engineering, Tianj in University, Tianj in 300072) Abstract: In t is paper, t e problems of slow convergence speed and being prone to converge to minimum are solved by combining t e c aracteristics of global optimization of GA ( Genetic Algorit m) wit local optimization of BP neural network, t e calculation accuracy and convergence rate of BP neural network are improved.T e sediment and scour of Sanmenxia reservoir in flood season are calculated and predicted by t is model and BP neural network. Compared wit normal BP neural network, t e numerical results of BP neural network improved by GA s ow t at t e average errors of calculation and prediction are diminis ed and t e iterative numbers are muc less t an t at of t e normal BP neural network. T us t e BP neural network improved by GA is suitable to solve nonlinear problems suc as calculation of sediment and scour of reservoirs. Keywords: river sediment engineering; GA ( Genetic Algorit m) ; BP neural network; sediment and scour of reservoir; Sanmenxia reservoir 1 , , , , , , , , , , , , BP( Back Propagation) , , , , ; ( Genetic : 2004-04-20 : : , 1978, , 4 : BP 111 Algorit m, GA) , , , , : BP , , ; , , 2 BP BP