论文部分内容阅读
针对标准BP神经网络存在收敛速度慢和易陷入局部最小值的问题,提出用附加动量法和自适应学习速率法来优化BP神经网络,提高其收敛速度;引入具有全局搜索能力的模拟退火算法,克服其容易陷入局部最小值问题。应用综合改进后的BP神经网络对已知的实际边坡进行了预测,并将其预测结果与标准BP神经网络和实际值进行对比分析。结果表明:综合改进后的BP神经网络在边坡稳定性预测具有较好的预测效果,与标准的BP神经网络相比,不仅提高了计算速度,而且较大地提高了预测精度,具有较好的应用前景。
Aiming at the problem that the standard BP neural network has slow convergence rate and easy to fall into the local minimum, this paper proposes an additional momentum method and an adaptive learning rate method to optimize the BP neural network to improve the convergence speed. A simulated annealing algorithm with global search capability is introduced, Overcome its easy to fall into the local minimum problem. The well-known BP neural network is used to predict the actual slope, and the predicted results are compared with the standard BP neural network and the actual value. The results show that the comprehensively improved BP neural network has good prediction results in slope stability prediction. Compared with the standard BP neural network, the improved BP neural network not only improves the calculation speed, but also greatly improves the prediction precision, which has a good Application prospects.