单站多模式跟踪测量机器人动态误差研究

This project focuses on the critical measurement problems of robotic dynamic errors, originally introduces a cascade-prism multi-mode tracking method into the robotic dynamic measurement, and combines with the single-station dual-field 3D reconstruction method, which can not only realize the dynamic measurement of coarse-fine sequence tracking, time-varying tracking, continuous tracking and so on, but also produce a variety of tracking patterns such as a straight-line shape and an arc one, and meet the dynamic multi-DOF measuring requirements of large field of view, high resolution imaging, large-range high-accuracy orientation as well. Research contents include: build the measuring project and mathematical model of joint control for both the cascade-prism coarse-fine coupling tracking and the dual-field imaging; study on the parameter matching, the mode conversion, the extraction of measuring information and the image processing of coarse-fine tracking and dual-field imaging; build the theoretical model, error model and experiment project of the robotic dynamic error measurement according to the measuring requirements, and realize the precise calibration of the measuring system; provide a scientific basis for the robotic error measurement and evaluate the measuring precision through the measuring experiment of robotic dynamic errors. The single-station multi-mode tracking measuring method proposed by the project is original and feasible in order to establish an effective measuring method of robotic dynamic errors, and provide a technical development basis of the practical measuring tools, which has an important theoretical significance and application value.

本项目瞄准机器人动态误差测量中的关键问题，原创性的将级联棱镜多模式跟踪方法引入机器人动态测量中，结合单站双视场三维重建方案，不仅可以实现粗精顺序跟踪、时变跟踪和连续跟踪等动态测量要求，而且能够产生直线形和圆弧形等多种跟踪样式，同时满足大视场、高分辨率成像和大范围、高精度定向的动态多自由度测量要求。研究内容包括：建立级联棱镜粗精耦合跟踪和双视场成像联控的测量方案和数学模型；研究粗精跟踪和双视场成像的参数匹配、模式转换、测量信息提取与图像处理方法；根据测量要求，建立机器人动态误差测量的理论模型、误差模型和实验方案，并实现测量系统的精确标定；通过机器人动态误差的测量实验，为机器人误差测量提供科学依据，同时对测量精度进行评定。本项目提出的单站多模式跟踪测量方法具有独创性和可行性，旨在建立有效的机器人动态误差测量方法，为发展实用的测量工具奠定技术基础，具有重要的理论意义和应用价值。

针对机器人动态误差测量中的大范围、高精度、多模式等要求，本项目原创性地将级联棱镜多模式跟踪方法引入机器人动态测量中，结合单站双视场三维重建方案，实现了对于机器人运动轨迹的粗精耦合顺序跟踪、时变跟踪以及连续动态跟踪，满足了大视场、高精度空间定向以及实时动态跟踪等测量要求。本项目建立了级联棱镜跟踪系统理论模型，完成了跟踪机构与成像系统的参数匹配；分析了单站多模式跟踪测量系统的不同误差来源，确定了各误差源对跟踪测量误差的贡献，提出了相应的误差修正方法；采用双视场成像方法提取目标位置信息，研究了光斑位置的亚像素提取方法与棱镜的跟踪控制问题，提出了将非线性问题转移至传动机构轮廓设计的方法简化非线性控制难题；建立了自标定法、自准直法和干涉测量法，结合相机标定算法，实现了对于单站多模式跟踪测量系统的精确标定，粗跟踪系统跟踪精度优于30μad，精跟踪系统跟踪精度优于0.5μad；建立了单站多模式测量机器人动态误差测量方法，实现了机器人动态误差的精确评定。本项目提出的单站多模式跟踪测量方法具有独创性和可行性，创新了机器人动态误差测量方法，为发展实用的测量工具奠定技术基础，具有重要的理论意义和应用价值。