在逆傅里叶变换法生成平稳高斯桥面不平顺随机激励的基础上，建立随机激励过程相位角与高阶中心矩之间的关系，通过不断调整相位角大小来改变桥面不平顺的非高斯特征，多次迭代生成给定功率谱密度、偏斜度和峭度的平稳非高斯桥面不平顺随机激励；同时，根据现场实测的桥面不平度数据，采用相位调制法重构桥面不平顺随机过程。数值算例和实测桥面不平顺的重构结果表明，该研究基于相位调制生成的平稳非高斯桥面不平顺能够满足给定目标功率谱密度、峭度和偏斜度的要求且模拟精度较高。桥面不平顺的幅值分布区域和幅值最大值随着峭度的增大而增大；偏斜度越大，桥面不平顺正偏或负偏的幅值分布区域就越大。工程应用结果表明，超高斯桥面不平顺激励的车桥振动响应最大，高斯桥面不平顺次之，亚高斯桥面不平顺激励的车桥振动响应相对较小；车辆振动对非高斯桥面不平顺激励的敏感程度大于桥梁振动。

On the basis of generating the stationary Gaussian bridge deck irregularity by the inverse fourier transform method, the relationship between the phase angle and the high-order central moment of the random excitation process was established, and then the non-Gaussian characteristics of the bridge deck irregularity were changed by continuously adjusting the phase angle, the random excitation of non-Gaussian bridge deck irregularity with given power spectral density, skewness and kurtosis was generated by iteration.At the same time, based on the measured data of deck roughness, the stochastic process of deck roughness was reconstructed by phase modulation method.Numerical examples and reconstruction results of measured bridge deck irregularities show that the non-Gaussian bridge deck irregularities generated by phase modulation can meet the requirements of given target power spectral density, kurtosis and skewness, and the simulation accuracy is high.The area of amplitude distribution and the maximum value of amplitude increase with the increase of kurtosis, and the amplitude distribution area of the positive or negative deviation of the deck irregularity increases with the increase of the skewness.The engineering application results show that the vehicle-bridge vibration response excited by super-Gaussian bridge deck irregularity is the largest, followed by Gaussian bridge deck irregularity, and the vehicle-bridge vibration response excited by sub-Gaussian bridge deck irregularity is relatively small.The sensitivity of vehicle vibration to non-Gaussian bridge deck irregularity excitation is greater than that of bridge vibration.