Vision-based Motion Tracking And Servo Control Of The Precision Positioning Stages

The rapid development of precision positioning platforms with large strokes,high precision,and multiple degrees of freedom has brought great challenges to current motion detection methods and control technologies.Due to the characteristics of non-contact,high flexibility,visualization,etc.,machine vision technology has the potential to solve these problems.However,the shortcomings inherent in the vision technology,such as the large amount of calculation and the vulnerability to environmental disturbances,make it limited in the application of precision positioning.This dissertation aims to study vision-based motion tracking and servo control methods with high-performance for precision positioing stages(PPS)with multi-degreeof-freedom(DOF).The main research content of this article is as follows:(1)Two vision based motion measurement systems are designed for planar multi-DOF PPS with different working spaces in scale.In order to solve the problem of hard to calibration caused by reasons such as small field of view and low depth of field of the micro-vision measurement system(MVMS),an automatic calibration method based on line element was proposed.The nonlinear imaging model of the MVMS is introduced and the main principles and steps involved in the proposed line-based calibration method are deduced.Finally,calibration patterns which contain different types of line elements were used.The experimental results show that the proposed calibration method can effectively calibrate the complete geometric model parameters of the MVMS and improve the measurement accuracy of the MVMS.(2)To realize motion tracking of the PPS with large stroke,a degenerate perspective-n-point(DPnP)method is proposed.This method integrates the camera imaging model and projective transformation to transform the planar 3-DOF motion tracking problem into a DPnP problem,and then constructs a regularization equations group to obtain the solutions.In order to further improve the accuracy of this method,the factors that influence measurement accuracy of the system including lens distortion,non-perpendicular error and iput parameters uncertianties are modeled and analyzed.Finally,the effectiveness and performance of the proposed method are verified by simulations and experiments,the results show that the proposed method is robust and accurate when used for motion measurement of 3DOF PPS.(3)For the pose measurement of PPS with small working space(micrometer level),an optimized template matching(OTM)tracking method is proposed.In the OTM method,to overcome the shortcomings of large amount of calculation and less freedom of measurement in the conventional TM technique,the TM method and the projective transformation group are combined,so that measurement of motions with different DOF is converted into the problems of parametric TM.To solve the above-mentioned multi-parameter nonlinear optimization problems,effective numerical optimization methods and strategies are studied.Finally,two tracking algorithms for PPS with 2DOF and 3DOF based on tranlational transfomration and Euclidean transformation are developed,respectively,and different experiments are performed to verify effectiveness and performance of the developed algorithms.(4)Visual servoing of the PPS is studied.Firstly,a method based on the OTM method is proposed,the key steps including the error function in the image space and the image Jacobian matrix are deduced.Then,an experimental workstation which mainly invloves an PPS developed in our laboratory and the MVMS developed in this dissertation is constructed.Finally,different experiments are performed to show the effectiveness of the proposed method.Finally,the research contents of the full text are summarized and the future research is prospected.