航空发动机表面管路是航空发动机的“心血管”，与发动机的安全和可靠性直接相关。管路振动是管路故障的重要原因之一。接触式的电测法是当前航空发动机表面管路振动测试的主要手段，而航空发动机表面管路有大量的观测点，这带来了一系列诸如测试效率低，测试精度不高等问题。因此针对航空发动机表面管路振动问题和不确定激励的特点，本课题组基于多相机机器视觉立体成像技术，提出了一种非接触式振动测试技术。本项目将系统地研究非接触式振动测试技术的误差作用机理，探索周期性与随机振动相叠加的复杂振动的高精度三维重建方法。由于非接触机器视觉测量难以获得激励信息，传统的传递函数模态分析法已不适用，因此在不确定激励下进行模态分析的方法，是本技术实现工程应用的关键。本项目还将研究基于时间序列振动位移信息的模态分析模型，为这种测试技术的发展及其在工程实际中的应用提供理论和实验依据。

The surface tubes like the cardiovascular system of the turbine, which is important for the safety and reliability of the turbine. The vibration is one of major cause of the failure of surface tubes. The contact electro measurement methods are the major technology for the vibration test of surface tubes now. But there are numerous test points for the turbine, which results in the low efficiency and low accuracy. Thus this project proposed a noncontact vibration test technology based on the binocular stereo vision in order to solve the problems for the vibration tests of turbine with unknown excitation properties. This project will systematically investigate the mechanisms of the error sources of the proposed technology and establish the 3D reconstruction method for the complicated vibration with periodic vibration plus random vibration. Because the excitation information is very difficult to detect by noncontact technology, the modal analysis based on transfer function is not available at this situation, which is key for the industrial application. The project will investigate the modal analysis method for the subjects under unknown excitation based on time series displacement information, which will provide the theoretical and experimental support for the development of the proposed test technology.