改进灰色因果时序组合模型在大坝位移监测中的应用研究

发布时间：2018-04-14 17:39

本文选题：大坝监测位移 + 灰色因果模型　；参考：《合肥工业大学》2014年硕士论文

【摘要】：位移是大坝安全状态的重要指标，大坝位移监测资料的分析工作关系到大坝和坝区人们的生命财产安全，因此，加强对大坝的位移监测是保障大坝安全的一项意义重大的工作。大坝监测位移通常受到水压、温度、时效等因素的影响，而且各因素之间相互制约、相互影响，使得大坝监测位移系统成为非线性系统。为了准确的分析大坝位移监测资料，进行实时预报，本文针对大坝位移监测的特点并对位移监测资料和监测方法分析的基础上，提出以改进灰色因果时序组合模型对大坝位移监测资料进行分析和研究。考虑到大坝受到多因素的影响，结合灰色因果模型的应用特点，本文通过建立灰色因果模型对大坝位移监测资料进行建模分析。灰色因果模型建模的主要工作是对因子的选择和模型程序的编译，通过选取合理的因子，对大坝位移监测资料建立灰色因果模型。模型精度是评判模型优劣的一个标准，在运用灰色因果模型时，，为了提高拟合和预测精度，本文将基于Simpson公式改进GM(1，N)模型，并将改进的灰色因果模型应用在大坝监测位移中，通过分析比较传统灰色因果模型和改进后的模型拟合及预测效果。对于建立的改进GM(1，N)模型所产生的拟合残差，由于残差序列具有随机性和动态性，而时间序列分析是处理这类数据的有效方法，所以文中对残差序列建立ARMA模型。基于上述分析，本文以改进后的灰色因果模型和ARMA模型建立改进灰色因果时序组合模型对大坝位移监测资料进行建模分析。本文是以某拱坝位移监测资料为基础建立实例，并通过监测资料的分析对模型的效果进行比较分析。研究表明，本文所采用的Simpson改进方法，具有一定的适用性，改进后的方法在工程实例中取得了良好的效果。其次，建立的组合模型预测结果相对于改进后的灰色因果模型，又有进一步的提高，因此组合模型能更准确的反映大坝的位移情况，为大坝位移监测资料的应用研究和保障大坝安全运行提供了有效工具。
[Abstract]:Displacement is an important indicator of dam safety. The analysis of dam displacement monitoring data is related to the safety of people's lives and property in dam and dam area. Therefore, it is of great significance to strengthen the monitoring of dam displacement.Dam monitoring displacement is usually affected by water pressure, temperature, aging and other factors, and each factor restricts and affects each other, which makes the dam monitoring displacement system become a nonlinear system.In order to accurately analyze the dam displacement monitoring data and carry out real-time prediction, this paper analyzes the characteristics of dam displacement monitoring and the analysis of displacement monitoring data and monitoring methods.An improved grey causal time series combination model is proposed to analyze and study the dam displacement monitoring data.Considering that the dam is affected by many factors and combined with the application characteristics of the grey causality model, this paper establishes a grey causal model to model and analyze the dam displacement monitoring data.The main work of the grey causal model modeling is to select the factors and compile the model program. By selecting the reasonable factors, the grey causality model is established for the dam displacement monitoring data.The model precision is a criterion for evaluating the model. In order to improve the fitting and prediction accuracy of the grey causality model, this paper improves the GM1N) model based on the Simpson formula, and applies the improved grey causality model to the dam monitoring displacement.By analyzing and comparing the traditional grey causality model with the improved model fitting and forecasting effect.For the fitting residuals produced by the improved GM1N) model, because the residual sequence is random and dynamic, and time series analysis is an effective method to deal with this kind of data, the ARMA model of residual sequence is established in this paper.Based on the above analysis, the improved grey causality model and ARMA model are used to model and analyze the dam displacement monitoring data.Based on the monitoring data of the displacement of an arch dam, an example is established in this paper, and the effect of the model is compared and analyzed through the analysis of the monitoring data.The research shows that the improved Simpson method adopted in this paper has certain applicability, and the improved method has achieved good results in engineering examples.Secondly, compared with the improved grey causality model, the prediction results of the combined model are further improved, so the combined model can reflect the displacement of the dam more accurately.It provides an effective tool for the application research of dam displacement monitoring data and the guarantee of dam safe operation.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TV698.11

【参考文献】