基于BP神经网络的模块化潮汐预报

发布时间：2018-04-15 17:55

本文选题：潮汐预报 + 调和分析　；参考：《大连海事大学》2015年硕士论文

【摘要】：潮汐预报在海上交通、港口建设和潮汐能利用等领域都具有重要意义,随着航运业的不断发展,以及对航行安全和航运效率的要求,对潮汐数值预报的精度也提出了更高的要求。将神经网络应用于潮汐预报领域是近年来出现的一种新的研究方向。反向传播学习(Back Propagation)申经网络在模式识别和系统预测领域应用广泛,本文将BP神经网络用于潮汐预报,对BP神经网络在潮汐预报领域的应用进行了探讨。传统调和分析法进行潮汐预报时,由于仅考虑了潮汐天文潮部分的影响,导致其在复杂环境因素影响下的海区预测精度明显下降。针对传统调和分析预报方法无法实现潮汐非天文潮部分准确预报的问题,本文建立了一种使用BP神经网络直接进行潮汐预报的模型。该模型基于实测潮汐数据进行实时的短期潮汐预测,提高了短期潮汐预测精度。模块化设计是解决复杂非线性问题的一种思路,通过分析潮汐的组成成分,提出了一种基于BP神经网络的模块化潮汐预报模型。模型包含了用于预测潮汐天文潮部分的调和分析预测模块以及用于预测非天文潮部分的BP神经网络预测模块。模块化模型有效实现了预测功能的区分,将调和分析预报法能够实现长期、稳定的天文潮预报的优点与BP神经网络能够实现潮汐非线性及未建模部分预报的优点相结合。在保证预测稳定性的前提下,进一步提高了预报的精度。将提出的模块化预测模型与传统调和分析法、BP神经网络直接预测法相比较,并进行了计算机仿真验证。实验证明,对于短期潮汐预报而言,模块化模型的预测性能要强于调和分析法和BP神经网络直接预报法。
[Abstract]:Tidal forecasting is of great significance in the fields of maritime traffic, port construction and tidal energy utilization. With the continuous development of the shipping industry, as well as the requirements of navigation safety and efficiency,A higher requirement for the accuracy of numerical tidal prediction is also put forward.The application of neural network to tidal prediction is a new research direction in recent years.Backpropagation Learning back Propagation (BP) network is widely used in pattern recognition and system prediction. In this paper, the application of BP neural network in tidal prediction is discussed.When the traditional harmonic analysis method is used to predict the tide, the accuracy of the sea area prediction under the influence of complex environmental factors is obviously decreased because the influence of the tidal astronomical tide part is only taken into account.Aiming at the problem that the traditional harmonic analysis and prediction method can not realize the accurate prediction of tidal non-astronomical tide, this paper presents a direct tidal prediction model using BP neural network.Based on the measured tidal data, the model can predict the short term tide in real time and improve the accuracy of the short term tide prediction.Modular design is a method to solve complex nonlinear problems. By analyzing the component of tide, a modular tidal prediction model based on BP neural network is proposed.The model includes harmonic analysis and prediction module for predicting tidal astronomical tide and BP neural network for predicting non-astronomical tide.The modular model can effectively distinguish the prediction function, combining the advantages of harmonic analysis forecasting method to achieve long-term and stable astronomical tide prediction, and BP neural network to achieve tidal nonlinear and unmodeled partial prediction.The prediction accuracy is further improved under the premise of ensuring the prediction stability.The proposed modular prediction model is compared with the BP neural network direct prediction method of traditional harmonic analysis method, and computer simulation is carried out to verify it.Experimental results show that the prediction performance of the modular model is better than that of harmonic analysis and BP neural network direct prediction for short term tidal forecasting.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P731.34

【参考文献】