视觉位感条纹同步测量转子多维动态位移的研究

The synchronous measurement for the multiple dynamic parameters of rotating rotor, including the three-dimensional vibration, rotational angle and swing angle, is a significant challenge for single-sensor-based measurement occasion. A novel solution for synchronously sensing all of the parameters is proposed by using a displacement-sensitive fringe pattern (DSFP) as an encoder and a camera as a detector. Based on the united simulation model of rotor system and monocular vision-based imaging system, the DSFP with good coding efficiency is designed and optimized firstly. Subsequently, the one DSFP image-based multi-displacement synchronous coding mechanism of rotor is established. The relationships between the image dynamic coding parameter and the real displacement in each dimension is investigated, from which both the decoding strategy and mathematical model for the separation of each displacement is established. The error sources that has influence on the sensing performance are analyzed. The affecting and transfer mechanisms of the key error sources to the measurement performance parameters and its error suppression principle and compensation strategy in each dimension will be revealed and established. Finally, based on the fundamental theories and achievements, a complete theory and advanced measurement system for simultaneous measurement of rotor multi-dimensional displacements will be proposed. The achievements of this project has great engineering and scientific significance to promote the development of the dynamic measurement technology, to improve the engineering measuring ability and efficiency, and to improve the accuracy and intelligence level of online monitoring, diagnosis and control of rotating machinery.

本项目针对单传感器测量模式下转子三维振动、转动和摆动等多动态参数难以同步测量问题，以位感条纹为编码介质，开展单目视觉下转子多维动态位移同步测量的研究。利用转子多维运动和视觉成像的联合仿真模型，寻求实现转子多维动态位移高效同步编码位感条纹的创新设计与优化，建立单目视觉成像位感条纹下转子多维动态位移的同步编码机制；研究成像位感条纹各维编码参数随转子真实多维动态位移的变化规律和映射关系，建立转子各维动态位移的解码分离策略和数学模型；在此基础上，分析影响系统测量性能的误差源，揭示关键性误差源对各维测量性能的影响规律和传递机制，研究相应的误差抑制原理和补偿策略，最终建立完善的基于单个位感条纹的转子多维动态单目视觉测量理论和系统。项目研究对推进我国动态测量技术的发展，提高工程测量能力和效率，推进旋转机械的在线监测、诊断和控制朝精确化、智能化发展具有重要的工程和科学意义。

旋转机械在制造、能源、航空等重要领域应用广泛，转子部件健康状态监测和在线故障诊断，对保障旋转机械的安全运行、提高设备的利用率、避免重大事故的发生有着重大的经济效益和社会价值。本项目针对单传感器测量模式下转子三维振动、转动等多动态参数难以同步测量问题，以位感条纹(对位置变化敏感的条纹)为编码介质，开展单目视觉下转子多维动态位移同步测量方法及系统的研究。提出了多种适用于转子多维动态位移同步编码的投影和非投影位感条纹，建立了转轴视觉位感条纹多维动态同步编码机制，并提出了成像位感条纹中转子各维动态位移的解码分离策略。探讨了转子异维运动与成像位感条纹各维动态编码参数之间的相互影响机制，制定了消除影响的抑制原理和补偿方法；研究了系统参数标定原理和方法，建立图像中各维编码参数与实际位移的映射关系和数学模型。分析了影响系统测量性能的影响因素及误差源，揭示关键性误差源对各维测量性能的影响规律和传递机制，研究相应的误差抑制原理和补偿策略，最终建立完善的基于单个位感条纹的转子多维动态单目视觉测量理论和系统。搭建了实验测量系统，并对测量系统的测量性能进行理论和实验分析，实现了转子三维振动及转动位移便捷、高效、精确的测量。结果表明：提出的测量方法及系统可实现转轴三维振动的高精度同步测量，其中径向位移测量精度可达亚像素级，轴向位移测量精度可达亚微米量级。研究进一步将视觉位感条纹多维测量系统应用于多种结构的振动监测、阶次分析及其转子的故障诊断应用中，实现了结构的多维动态监测及转子缺陷类型及损伤程度的检测。项目研究对推进我国动态测量技术的发展，提高工程测量能力和效率，推进旋转机械的在线监测、诊断和控制朝精确化、智能化发展具有一定的工程和科学意义。