基于归一化最小均方算法的钢轨波磨识别方法

打开文本图片集

中图分类号：U213.42 文献标志码：A doi：10.3969/j.issn.1006-0316.2025.11.005

文章编号：1006-0316（2025）11-0030-07

Abstract ： Rail corrugation is a common form of rail surface degradation in railway systems，characterized by regular undulations on the rail surface.Such wearand tear may not only reduce the smoothness oftrain operation and ride comfort for passengers，but also increase safety risks due to aggravated dynamic wheel-rail interactions. To address the challenges in rail corrugation detection， this paper proposes a wavelength identification method for metro rails based on the Normalized Least Mean Squares （NLMS） algorithm.The core innovation lies in employing the NLMS algorithm to construct anadaptive filter that dynamically adjusts its parameters to suppress interference from wheel wear，thereby improving wavelength detection accuracy. To verify the feasibilityof thealgorithm，simulation testswere conductedbyestablishing arigid-flexible coupled vehicle-track dynamics model using existing metro vehicle parameters，and field tests were performed on metro vehiclesbefore wheel reprofiling on severely corrugated track sections.Simulation results demonstrate a wavelength identificationaccuracyof 99.2% ，while field testsconfirmanaccuracy of 96% Key words ∵ rail corrugation ; axle-box vibration ; normalized least mean square algorithm

伴随着经济水平的不断提升，诸多城市正面临着交通运行效率低下、事故风险上升和生态环境恶化等多重挑战。（剩余7546字）