基于LSTM的疏浚管道流速预测模型

疏浚管道输送系统是绞吸挖泥船的重要组成部分,在施工过程中对流速的控制至关重要。该文以现有疏浚管道输送实验台为对象,提出了一种疏浚管道输送系统的流速预测模型。首先针对反向传播神经网络(BP)无法处理序列间的关联信息以及传统循环神经网络(RNN)无法记忆久远关键信息的缺陷,提出了基于长短期记忆循环神经网络(LSTM)的流速预测模型;然后使用LSTM模型对疏浚管道输送实验台和绞吸挖泥船的数据集进行网络训练,并对管道流速进行预测。通过将流速的预测值与真实值进行对比,验证了该文提出的LSTM模型具有很强的适用性和很高的准确性。

Prediction Model of Dredging Pipeline Velocity Based on LSTM

Dredging pipeline transportation system is an important part of cutter suction dredger,and it is very important to control the flow rate in the construction process. This paper presents a velocity prediction model for dredging pipeline transportation system based on the existing dredging pipeline transportation test bed. Firstly,a flow rate prediction model based on long short term memory(LSTM) was proposed to solve the problem that the back-propagation neural network(BP) could not process the correlation information between sequences and the traditional recurrent neural network(RNN) could not remember the long-term key information. Then LSTM model is used to train the data set of dredged pipeline transport test bed and cutter suction dredger,and the pipeline velocity is predicted. By comparing the predicted value of flow velocity with the real value,the LSTM model proposed in this paper is proved to have strong applicability and high accuracy.