遗传算法优化BP神经网络在大坝扬压力预测中的应用

Application of BP Neural Network Based on GA Optimization in Uplift Pressure Forecasting of Dam

针对BP神经网络的局部极小和收敛慢等问题,提出了利用遗传算法的选择、交叉和变异操作优化BP神经网络的权值和阈值,将优化后的BP神经网络用于预测大坝扬压力。通过实例应用,将遗传算法优化的BP神经网络与逐步回归、BP神经网络预测相对比,结果表明遗传算法优化的BP神经网络收敛快且预测结果精度高。

Aiming at the problems of local minimum and slow convergence in BP neural networks,the selection,crossover and mutation operators in genetic algorithm is proposed to optimize the weights and thresholds of BP neural network.And then the BP neural network optimized by genetic algorithm is applied to forecast the dam uplift pressure.Comparison of stepwise regression and BP neural network,example results show that the proposed BP neural network has characteristics of quick speed convergence and high prediction accuracy.