【作者】 沈典栋；

【导师】 沈成武； 杜国东；

【作者基本信息】 武汉理工大学 ， 船舶与海洋工程， 2003， 硕士

【摘要】 随着科学技术和交通事业的发展，预应力混凝土连续梁桥以其施工简便、造价经济、受力合理、行车舒适等独特优势在近年来得到了迅速发展，在主跨100～300m范围内几乎成为首选桥型。但由于它出现较晚，其理论和经验还不十分完善，在修建过程中也存在一些技术上的问题。在现代桥梁施工中，施工监制技术是工程施工安全和内在质量的保证，正确计算和设置施工过程参数已成为大跨径桥梁施工领域重要研究课题之一。 近年来，神经网络误差调整方法成为国内外一个热门的研究课题。人们普遍认为这些领域的发展将会带来重大的研究成果和应用前景。神经网络是研究复杂系统的一种有效的信息处理方法，它不需要任何数学模型，但可以处理非线性问题。常用于预测、分类等各种数据处理的场合。近年来，人工神经网络技术在工程领域得到了广泛应用。BP网络是应用最广泛的一种网络，其学习方法是δ学习算法的推广和发展，是一种有导师的一种学习。BP算法经过多年的发展，现已得到许多改进。 襄樊汉江四桥是汉十-襄荆高速公路连接线上跨越汉江的一座特大桥，主跨为70+5*120+70米，其结构型式为预应力混凝土连续梁桥。该桥采用节段式悬拼施工法，其施工预拱度的控制是关键指标，目前没有一个成熟预测方法。本文以襄樊汉江四桥为依托，将神经网络误差调整方法用于大跨度混凝土桥梁的施工参数的识别。建立施工预拱度的数学模型，根据施工时的预应力张拉后等工况桥面实测标高与设计值的差异，用BP神经网络识别预拱度模型的参数，确定有关各截面的预拱值。分析表明该方法是有效的，与实测预拱度结果较吻合。 本文以大桥成桥线形满足设计要求和成桥内力控制在设计容许范围内为目的，运用现代控制理论的思想，引用神经网络基本原理，详细介绍BP神经网络的研究过程，从而建立预应力混凝土连续梁桥施工预拱度的神经网络模型，并详细讨论了模型的神经网络结点信息的提取，在工程实例的基础上进行了数值计 武汉理工大学硕士学位论文算，数值表明，将神经网络应用于预应力混凝土连续梁桥施工控制，获得了满意的结果，在预应力混凝土连续梁桥施工控制技术研究方面，具有方便有效、精度高的优点，具有良好的应用前景。进行了一种新的尝试。更多还原

【Abstract】 With the development of science and technology and traffic enterprise, the prestressed concrete continuous bridge has been growing up rapidly in recent 20 years with it’s unique predominance of handy construction, economical cost, reasonable internal force and comfortable traveling. It almost has become the first selective bridge style When the main span is in the range from 100m to 300m. But the theories and experience in it are not very perfect because of it’ s short appearance. In modern bridge building, the the construction controlling technique is a warranty of safe and quality of the bridge in the construction process. This paper introduced the importance of the camber of the longer span bridge in the construction.In recent years, Artificial Neural Network(ANN) has become a striking research topic in china and abroad. It is generally believed that development in this field will bring significant outcomes and great promise for application. ANN is one of the efficient information processing’method in studying the sophisticated systems. It need not mathematical model, but can deal with non-liner fuzzy problems. It is often used in predicting, classification and so on. Inrecent years, ANN has been used in engineering fields. Back Propagation neural Network(BPN) is one of the most widely used network, The practice algorithm is similar to deta rule. With long period developing, BPN has been improving.The No. 4 bridge of Xiangfan bridge over-crossing Hanriver is a specially long span river bridge on freeway by tieing freeway form HanShi to Xiangjing. The main span is 70+5*120+70m. It is a prestressed concrete continuous beam bridge. The bridge has builded by balanced cantileverconstruction technique. The beam camber of construction is a controlling index of construction quality control.It is nonlinear, and now is not mature forecasting method. Relying on The No. 4 bridge of Xiangfan bridge over-crossing Hanriver bridge, in this paper a ANN is applied to the parameter of the long span bridge in the construction beam. A model with some parameters is used to forecasting the beam camber in the construction. Based on the differences between the design value and the survey of the bridge deck elevation in constructing the pre-stressed and so on working condition, the model parameters are identified by ANN, and the beam camber of different sections are given. The results demonstrated that the ANN is efficient and the theoretical results fit well with that of forecasting the beam camber in construction.In this thesis, aim to the beam axis shape be exactitude for design, the internal force of the beam be appeased by design request, the concept of modem cybernetics is applied, At the same time, the effecting parameter has collected. The paper build the ANN model in predicting the beam camber an construction and discuss indetail the extraction of inputting nodes information when quoting the fundamental principle of ANN model and introducing research process of the improved BNN. According to them, ANN is applied in the construction control of prestressed concrete continuous bridge. It is successful for used in the No. 4 bridge of Xiangfan bridge over-crossing Hanriver brideg. To predict the construction camber of by ANN method is useful and effective. On the bridge construction control technique, attempted a new method.更多还原