轨道车辆运行平稳性评价算法一致性分析

Analysis of Consistency of Ride Quality Evaluation Algorithm for Railway Vehicles

摘要:

国内外轨道车辆运行时振动相关舒适性评价基于ISO2631-1、UIC513以及GB/T 5599标准，其中我国广泛采用的GB/T 5599-2019规范中的Sperling指标算法，将采样时间由1985版本的18~20 s缩短为与UIC513标准相同的5 s，引起相同测试数据的平稳性指标结果增大的问题。分析上述三种运行平稳性及舒适性指标算法，通过车辆线路实测数据与仿真数据相结合，研究不同算法关于采样时间长度的指标一致性及其机理，采用频率分辨率量化了非平方加权谱由于时频能量转换不对应导致的指标结果不稳定现象，基于此提出相应的平稳性指标一致性修正算法。结果表明：ISO2631-1与UIC513标准算法基于振动数据时频域转换的能量对应性，其计算结果与采样时间长度不相关，而GB/T 5599中采用的频域立方值算法因不满足能量一致性原则，在栅栏效应和能量泄露的影响下，缩短采样时间使得加速度频谱幅值增大，进而引起运行平稳性指标增大。提出的基于统一频率分辨率的平稳性指标一致性修正算法，并经动力学仿真数据验证有效，这为完善轨道车辆动力学分析与测试中振动舒适性的相关算法与评估限值制定提供理论依据。

关键词:轨道车辆;运行平稳性;频率分辨率;Sperling指标;指标一致性

Abstract:

The evaluation of ride comfort of railway vehicles related to vibration is mainly based on the standard of ISO 2631-1， UIC 513 and GB/T 5599. Among them， the Sperling index algorithm in the GB/T 5599—2019 specification widely used in China shortens the sampling time from 18–20 s in the 1985 version to 5 s， the same as the UIC513 standard， which results in an increase in the index result from the same test data. The algorithms of the three standards for ride quality or ride comfort were analyzed with the measured data and the simulation data to study the index consistency about the sampling time and the mechanism. Then the frequency resolution was used to quantify the instability of the index results caused by the weighted spectrum with non-square value due to the mismatch of time-frequency energy conversion. Based on this， the corresponding correction algorithm of ride quality consistency is proposed. The results show that the algorithms in the standards ISO 2631-1 and UIC 513 are based on the energy correspondence of time-frequency domain conversion of vibration data， whose ride comfort results are not related to the sampling time length. However， the cube value algorithm in the frequency domain adopted in GB/T 5599 does not meet the energy consistency principle. With the influence of the fence effect and energy leakage， shortening the sampling time increases the amplitude of the acceleration spectrum， which leads to an increase in the ride quality index in turn. The consistency correction algorithm of ride quality index based on the unified frequency resolution proposed in this paper was validated by the dynamic simulation data， providing a theoretical basis for improving the vibration comfort related algorithms and evaluation limits in vehicle dynamics analysis and testing.

Key words:railway vehicle;ride quality;frequency resolution;Sperling index;index consistence