首页|基于协同降噪与IGWO-SVR的高填方路基沉降预测

基于协同降噪与IGWO-SVR的高填方路基沉降预测

扫码查看
原文链接
  • 万方数据
  • 维普
高填方路基沉降影响山岭重丘区重载铁路运营安全.为克服实测沉降数据掺杂随机噪声、现有预测模型适用性差的不足,提出基于协同降噪算法与IGWO-SVR模型的沉降预测方法.运用互补集合经验模态分解法(CEEMD)与小波包变换法(WPT)对含噪沉降数据进行协同降噪处理;提出基于佳点集初始化均布、非线性收敛控制与自身历史最优记忆位置更新的改进灰狼优化(IGWO)算法,并结合支持向量回归模型(SVR),构建IGWO-SVR沉降预测模型.进一步地,利用大准铁路工点及现有文献研究成果,验证IGWO-SVR模型的优越性.结果表明:协同降噪法可有效消除原数据中噪声项的干扰波动;在小样本数据集上,IGWO-SVR模型较传统沉降预测模型与现有文献所述预测模型,具有更高的预测精度与稳定性.研究成果为重载铁路高填方路基沉降预测提供了新途径.
Settlement Prediction of High-filled Embankment Based on Collaborative Denoising and IGWO-SVR
The settlement of high-filled embankment affects the safety of heavy-haul railway in mountainous and hilly ar-eas.This paper proposed a prediction method based on collaborative denoising algorithm and IGWO-SVR model for the settlement of the high-filled embankment on heavy-haul railway to address the problems of measured settlement data con-taminated with random noise,and poor applicability of existing prediction models.The complementary ensemble empirical mode decomposition(CEEMD)and wavelet packet transform(WPT)were used to collaboratively denoise the noisy settlement data.With an improved grey wolf optimization algorithm(IGWO)proposed based on good point set ini-tialization,nonlinear convergence control,and individual position update with its best value in history,the IGWO-SVR settlement prediction model was built by combining the IGWO and the support vector regression model(SVR).Further-more,the superiority of the IGWO-SVR model was validated using the high embankments of the Dazhun Railway and the results of previous literature research.The results indicate that the collaborative denoising method effectively eliminates the interference fluctuations of noise in the original data.On small sample datasets,the IGWO-SVR model has higher accuracy and stability than traditional settlement prediction models and the prediction model described in the literature.The findings of this research offer a new approach for predicting the settlement of high-filled embankments on heavy-haul railways.

heavy-haul railwayhigh-filled embankmentsettlement predictioncollaborative denoisingimproved grey wolf optimizationsupport vector regression

苏谦、张棋、张宗宇、牛云彬、陈德

展开 >

西南交通大学土木工程学院,四川成都 610031

西南交通大学高速铁路线路工程教育部重点实验室,四川成都 610031

重载铁路 高填方路基 沉降预测 协同降噪 改进灰狼优化 支持向量回归

国家自然科学基金国家自然科学基金国家自然科学基金四川省自然科学基金国家重点研发计划

5197858851808462U22682132023NSFSC03462016YFC0802203

2024

10.3969/j.issn.1001-8360.2024.03.010

铁道学报
中国铁道学会

铁道学报

CSTPCD北大核心
影响因子：0.9
ISSN：1001-8360
年,卷(期)：2024.46(3)
  • 38