存在拓扑结构切换的物像运动补偿有限时间多边控制问题研究

In imaging system, the active scanning and the motion of foundation support lead to the image movement, which seriously degrades the imaging quality. It is significantly valuable in science to investigate the control of multiple optical actuators for compensating the image movement. Compared with traditional control of several independent actuators, this research extends the idea of multilateral control from the field of teleoperation robots to the field of optical imaging. Based on considering the overall system as one controlled plant, the strategy of multilateral control is proposed for compensating the image movement, which improves the accuracy. Based on early foundation, this research focuses on the problems of unavoidable topology switching and the limitation of time performance, such as the frame rate and exposure time, due to the convergence time. The strategy of finite time multilateral control is to be mainly investigated. Based on the idea of modal transmission, a medium can be found between multilateral control with topology switching and the extant investigation on finite-time control. By studying the finite-time stability criterion and controllers synthesis of multilateral control, the theories on multilateral control can be enriched. The application effect can be also improved. Moreover, quantifiable verification on the space and time performances of the multilateral control is to be implemented using the optical indicator of modulation transfer function. This research gives the stability criterion, the controller synsthesis and the performance evaluation, which establishes the foundation of multilateral control theory for imaging system with high accuracy comepnsation of image movement. It also supports essential theories in the development of next generation high performance imaging system.

成像系统的主动扫描和基座运动会导致物像运动，严重降低成像质量。以成像系统中的多个光学执行原件为被控对象的物像运动补偿控制研究具有重要的科学意义。本课题抛开传统成像系统中对各执行器分别控制的方法，将遥操作机器人领域的多边控制思想引入光学成像领域，提出物像运动的多边控制补偿策略，从整体上提高物像运动补偿控制的精度。课题在前期研究的基础上，针对实际成像系统多边控制中存在不可避免的拓扑切换、切换后系统收敛时间限制成像帧频和曝光时间等问题，以有限时间多边控制策略研究为主线，以模态变换思想为桥梁，找到有拓扑结构改变的多边控制系统与当前有限时间控制研究的契合点。通过研究多边控制有限时间稳定性判据和有限时间控制器设计等问题，完善多边控制的理论构架并提升其应用效果。以光学的调制传递函数为依据定量衡量控制性能。为解决成像系统高精度、快收敛物像运动补偿问题和新一代成像系统的研制提供必要的理论支撑。

光学成像系统的主动扫描和基座运动严重降低成像质量，需要进行主动的运动补偿。现代成像系统中往往有多个执行器，其动力学特性、制动机理、工作环境、受干扰特性均有较大差异。如何协调多个执行器，使其在光学系统中合理地协同从而高精度地补偿成像过程中像的运动是一项待解决的难题。本项目首先针对机械照准架、压电陶瓷促动器等典型执行器的特性，研究了其干扰抑制和有限时间控制问题，奠定了多执行器协同控制的基础。再将遥操作机器人领域的多边控制思想引入光学成像领域，提出物像运动的多边控制补偿策略，从整体上提高物像运动补偿控制的精度。考虑了曝光时间、帧频等时间指标，建立了存在控制拓扑结构切换的多执行器有限时间收敛协同控制策略。基于终端滑模控制，给出了具体的有限时间控制器设计和稳定性分析等结果，完善了多边控制的理论构架，扩展了其应用外延。基于新方法实现的执行器对指令跟踪误差和协同运动误差与传统方法相比分别下降73.38%和38.02%。在项目支持下，开展了大量实验，并进行了光学成像验证，以光学成像性能为直接依据衡量控制性能，为解决成像系统高精度、快收敛物像运动补偿问题和新一代成像系统的研制提供必要的理论支撑。