为分析地铁隧道振动源强的影响因素，以某地铁线路直线段为研究对象，同步采集车致振动信号和速度 信号，统计分析全天测试样本的时域、频域特征。结果表明：早晚高峰时段，隧道内振动响应不一定完全最大，振 动源强的离散程度与车辆状态直接相关，V1、V2 两车所引起的隧道壁测点Z 振级的均值相差6.7 dB；取40 组以 上的测试样本计算振动源强，其结果更接近全天样本均值；由轮轨系统引起的隧道振动主频随车速的变化而偏移， 车速与道床、隧道壁的振动加速度级呈现较强的线性正相关，车速每提高10 km/h，振动分别增加约1.7、2.6 dB； 速度差在10 km/h 以内时，振动源强在1 dB 以内变化，速度差在10～20 km/h 时，振动增幅为1.5～2.2 dB；计算 得到车速修正CV 的系数为18.5，与环评振动预测公式中列车速度修正所使用系数接近；隧道壁振动在40 Hz 以下 的低频段离散较大，但该频段对振动源强的贡献度较小，占15.29%，优势频段为50～63 Hz，该频段对振动源强 的贡献度达59.55%。研究成果可为环评振动预测公式的进一步精确化提供参考，为提高地铁隧道源强测试结果的 准确性提供理论依据。

This study focuses on a straight-line section of a subway line to analyze the factors influencing the vibration source strength in subway tunnels. Vehicle-induced vibration signals and velocity signals were collected synchronously, and the timedomain and frequency-domain characteristics of all-day test samples were statistically analyzed. The results showed that during peak hours in the morning and evening, the vibration response inside the tunnel may not be completely maximized, and the degree of dispersion of the vibration source intensity is directly related to the vehicle state. The study found an average difference of 6.7 dB in the VLZ of the tunnel wall measurement points caused by V1 and V2 vehicles. Using more than 40 sets of test samples to calculate the vibration source intensity yielded results closer to the daily sample mean. The main frequency of tunnel vibration caused by the wheel-rail system shifts with changes in vehicle speed. A strong linear positive correlation was observed between vehicle speed and the vibration acceleration levels of the track bed and tunnel wall. Specifically, for every 10 km/h increase in vehicle speed, the vibration increases by about 1.7 dB and 2.6 dB, respectively. When the speed difference is within 10 km/h, the vibration source intensity changes within 1 dB; when the speed difference is between 10~20 km/h, the increase is about 1.5~2.2 dB. The coefficient of speed-corrected CV was calculated to be 18.5, aligning closely with the coefficient used for train speed correction in the vibration prediction formula. Tunnel wall vibration was more discrete in the low-frequency band below 40 Hz, but this band contributed minimally to the source strength of vibration, accounting for 15.29%. The dominant frequency band was 50-63 Hz, contributing 59.55% to the source strength of vibration. These research findings can provide a reference for further refining the vibration prediction formula and offer a theoretical basis for improving the accuracy of metro tunnel source strength test results.

TH113；U231

国家自然科学基金(52178423，52068029)