中文摘要：

轨面短波不平顺对轮轨噪声、车辆轨道各部件的动力学破坏作用巨大，为得到其具有代表性的时域波形，并进一步研究轮轨振动噪声和轨面状态仿真分析提供基础，提出一种基于轨面粗糙度水平谱反演短波不平顺时域波形的算法。由于反演过程中相位的随机性，使该算法在保证代表性的同时又具备一定随机性。在此基础上，用该算法实现了ISO3095标准谱与上海轨道交通实测谱的反演，结果表明反演算法可行且正确，反演得到的轨面短波不平顺时域波形可作为轮轨动力学仿真模型的激励输入，也可为预测轨面不平顺发展提供初始数据。

英文摘要：

Rail surface shortwave irregularities not only have enormous impact on wheel/rail noise,but also bring huge dynamic damage to vehicle and track components. In order to obtain the representative time domain waveform for further study of wheel/rail noise and wheel/rail interface simulation,a time domain inversion al-gorithm of rail surface irregularities based on rail roughness level spectrum has been put forward. The algorithm provided at the same time has a certain randomness due to its phase randomness in the inversion process. The al-gorithm is then used to implement the ISO3095 standard spectrum and measured spectrum on Shanghai Metro. Results show that the inversion algorithm is feasible and correct,which can be used as excitation input of wheel/rail dynamics simulation model and initial datum of rail surface irregularities prediction model.