【摘要】 现有轨道不平顺检测方法中，惯性基准法难以检测短波，弦测法存在波形畸变问题，因此本文从波形复原与数据融合方面进行研究，提出一种基于卡尔曼滤波融合轴箱加速度与中点弦测模型的检测方法，以提高轨道短波不平顺的检测精度。通过建立基于高频采样的10 m中点弦测模型，采用批量梯度下降算法复原检测波形。将中点弦测模型检测值复原后的轨道不平顺作为观测量，轴箱加速度作为卡尔曼滤波算法的输入，建立状态方程，对轨道不平顺进行求解。仿真验证的结果表明，卡尔曼滤波融合算法检测短波不平顺的最大平均绝对误差为0.0438 mm，相关系数都在0.95以上，满足轨道短波不平顺检测需要，验证了数据融合算法的有效性。更多还原

【Abstract】 To address the issue that the inertial reference method is difficult to detect short waves and solve the the problem of waveform distortion in the chord measurement method in the existing track detection, a detection method based on Kalman filter fusion of axle box acceleration and mid-point chord measurement model is proposed from the aspects of waveform restoration and data fusionto improve the detection accuracy of the track short-wave irregularity. A 10 m mid-point chord model based on high-frequency sampling is established, and a batch gradient descent algorithm is used to recover the detected waveform. The recovered track irregularity of the mid-point chord model is taken as the observation quantity, and the acceleration of the axle box is taken as the input of the Kalman filter algorithm, so as to establish the state equations and solve for the track irregularity. The results of simulation verification show that the maximum mean absolute error of the Kalman filter fusion algorithm for detecting short-wave irregularity is 0.0438 mm, and the correlation coefficient is above 0.95, which meets the need for detecting short-wave irregularity of the track, and verifies the effectiveness of the data fusion algorithm.更多还原