摘要
轨道变形是影响地铁安全运营的重要因素,对地铁轨道变形进行监测和预测具有重要意义。论文基于极限学习机(ELM)、粒子群优化(PSO)算法,构建了小波去噪的PSO-ELM组合模型,有效解决了监测数据波动性干扰预测结果和模型参数选取随机性影响预测精度的问题。以某市地铁2号线轨道监测数据为例进行实验分析,表明该组合模型比单一ELM模型、小波去噪ELM模型的预测精度更高,预测误差发散性更小,具有更高的稳健性与适应性。
Track deformation is an important factor affecting the safe operation of subway.It is of great significance to monitor and forecast the track deformation.In this paper,based on extreme learning Machine(ELM)and particle swarm optimization(PSO)algorithms,a PSO-ELM combined model with wavelet de-noising was constructed,which effectively solved the problems that the volatility of monitoring data interfered with the prediction results and the randomness of model parameter selection affected the prediction accuracy.Taking the track monitoring data of Metro Line 2 in a city as an example,the experimental analysis shows that the combined model has higher prediction accuracy,less divergence of prediction error,and higher robustness and adaptability than the single ELM model and the wavelet de-noising ELM model.
作者
王智慧
蒋文
WANG Zhi-hui;JIANG Wen(China Institute of Building Standard Design and Research,Beijing 100048,China;Chongqing-Guizhou Railway Co.Ltd.,Chongqing 400037,China)
出处
《工程建设与设计》
2024年第19期89-92,共4页
Construction & Design for Engineering
关键词
地铁轨道
沉降
小波去噪
粒子群优化
极限学习机
subway tracks
settlement
wavelet denoising
particle swarm optimization
extreme learning machine
