Abstract: The vibration signal required by the extended transfer path analysis (OPAX) method is the working condition data of the vehicle, but the noise signal is mixed in the experiment, which has a certain impact on the analysis and identification accuracy. To solve this problem, the OPAX method is improved by using empirical mode decomposition threshold noise reduction. Taking an electric vehicle as the research object, the load parametric identification is carried out by using the frequency response function and working condition data, and then the vibration source in the electric vehicle is identified. The results show that under the uniform acceleration condition, the energy absorption effect of the rear mount z direction on the vibration of motor, reducer and other excitation sources is poor. The fitting response of the improved OPAX method is compared with that of the OPAX method, It is found that the fitting response of the improved OPAX method is more consistent with the measured response, which shows that the improved OPAX method can more accurately identify the vibration sources affecting the vibration in the vehicle.

Key words: OPAX method; Empirical mode decomposition; Threshold noise reduction; Identification of vibration source; Improved OPAX method; Electric vehicle

摘要： 扩展传递路径分析（OPAX）方法所需的振动信号为汽车的工况数据，但实验时掺杂有 噪声信号，对其分析识别精度有一定的影响。针对该问题，应用经验模态分解阈值降噪对 OPAX 方法进行改进。以某电动汽车为研究对象，利用频响函数和工况数据进行载荷参数化 识别，进而识别电动汽车车内振动源。结果表明，在匀加速工况下，后悬置 z 方向对电机、 减速器等激励源振动的吸能效果较差，并对比改进 OPAX 方法的拟合响应与 OPAX 方法的拟 合响应，发现改进 OPAX 方法的拟合响应与实测响应更加吻合，表明改进 OPAX 方法能够更 准确地识别出影响车内振动的振源。

关键词: OPAX 方法；经验模态分解；阈值降噪；振动源识别；改进 OPAX 方法；电动汽车