Kalman滤波在高铁桥梁变形预测分析中的应用

针对高速铁路桥梁架梁后许多沉降变形点沉降量级较小,变形曲线呈现＂小量级,大波动＂特点,观测数据中可能存在大量的随机噪声,对沉降变形分析产生干扰,影响预测结果的可信度,本文将Kalman滤波引入到高速铁路桥梁变形分析数据预处理中,建立基于Kalman滤波的动态模糊神经网络模型。通过应用实例分析表明,基于Kalman滤波的动态模糊神经网络模型的预测精度有所改善,具有一定的优势。

Application of Kalman Filtering in High-speed Railway Bridge Deformation Prediction Analysis

For settlement magnitude smaller of settlement de- formation point after high-speed railway bridge, girder deforma- tion curve presents the characteristics of small magnitude, large fluctuations. There may be a lot of random noise in the ob- served data, which affects the settlement deformation analysis and the reliability of the predicted results. This paper introduces the Kalmau filter to the high-speed railway bridge deforma- tion analysis of data preprocessing, establishes dynamic fuzzy neural network model based on Kalman filtering. The application of case study shows that the of the prediction accuracy dynamic fuzzy neural network based on Kalman filtering model is improved, which has a certain advantage.