改进梯度下降BP算法在地下水位预测中的应用

本文给出了BP神经网络预测模型的原理,分析了标准BP算法缺陷,通过改变学习率和增加动量项改进BP算法。用改进的算法预测某地地下水位,并对训练过程进行优化,实验结果表明,改进的BP神经网络能有效地提高地下水位预测的速度和精度,比标准BP算法预测性能有较大改善。

Application of Improved Gradient Descent by BP Algorithm to Groundwater Level Forecast

This paper gives the principles of BP neural network forecast model.After analyzing the limitations of the standard BP algorithm,improves BP algorithm by modifying the learning rate and adding a momentum item.To forcast the groundwater level in a certain place by the improved BP algorithm,and optimize the training process.The experimental results show that the improved BP neural network can effectively increase speed and accuracy of the groundwater level forecast, has a great improvement in performance compared with the standard BP algorithm.