大跨径PC连续刚构桥神经网络控制系统的研究

【作者】 任更锋；

【导师】 徐岳；

【作者基本信息】 长安大学 ， 桥梁与隧道工程， 2003， 硕士

【摘要】 本文以国内外大跨径预应力混凝土连续刚构桥施工控制资料收集分析为基础，借鉴现有的施工控制研究成果，对常用的施工控制方法进行分析，旨在将现代预测控制理论的最新研究成果神经网络引入大跨度PC连续刚构桥的施工控制。 通过对PC连续刚构桥现有施工控制资料的收集，借鉴已有研究成果，着重探究各种常用控制方法进行预测控制的本质，提出了基于神经网络为基本理论的施工控制方法。 从神经网络基本理论出发，分析了神经元结构、神经网络组成，重点进行神经元传递函数的分析，确定其在桥梁施工控制应用中的可行性，提出了BP网络为实桥施工的基本预测控制模型。 根据神经网络自身的特点，以及桥梁立模标高影响因素的复杂性，对神经网络影响因素进行分析，在满足桥梁施工精度的前提下，提出了在实桥施工过程中神经网络输入矢量的确定原则和方法。 借鉴现有的研究成果，确定以立模标高为神经网络的目标矢量，提出了目标矢量的组成、计算和调控方法。 借助MATALAB图形用户界面，构建神经网络，完成对样本矢量的输入，以及对桥梁施工监控网络的训练，进而实现了对实桥施工过程的预测功能。更多还原

【Abstract】 On the basis of the construction control data collection and the analysis of long span prestressed concrete continuous rigid-frame bridge at home and abroad, the paper draws the existing research achievement in this field, and studies the common methods, aims to introduce the theory of Artificial Neural Network(ANN),the mordenest prediction theory, into long span prestressed concrete continuous rigid-frame bridge construction control.Through the collection of long span prestressed concrete continuous rigid-frame bridge construction control data, with the help of the existing research achievement, the paper pursues for quality of the common forecasting method, and raises the method of construction control ,based on the theory of ANN.From basic construction of the ANN, the paper analyses the construction and the unit of ANN, pays more attention to transfer function of the unit. The paper gets the feasibility of using ANN in bridge construction control ,and raises BP network as control model.According to the characteristic of ANN itself and the complexity of factors which influence the elevation, the paper analyses the influence aspects of ANN. On the promise of bridge construction precision, the paper raise determining principle and method for neural network’s import vector in bridge construction process.Based on the existing research achievement, the paper take the elevation of formwork as the objective vector for ANN, and raises the construction calculation and control method of objective vector.With the help of Graphical User Interfaces (GUI) of Matlab software, the paper built ANN, imports the sample vector ,and trains the bridge construction ANN, thus makes the forecast function in the bridge construction into reality.更多还原