深海环境局部自主焊接机器人任务空间三维视觉重建与路径规划技术研究

It is promising to employ underwater welding robot instead of diving welder to fulfill underwater welding tasks. From the point of resolving those problems of three dimensional reconstruction for the task space and the path plan for underwater welding robot system, this project researches the macroscopic teleoperation, microscopic autonomous control method to complete the underwater welding task, This research puts forward a novel stereo vision approach based on the combination of the active vision method and the passive vision method, develops the image enhancement algorithm and three dimensional matching algorithm, and establishes a operator-machine shared three dimensional reconstruction model of the robotic task space. This novel approach supplies a reliable information foundation for the task scheduling and the path plan of the deep-sea welding robot. In oder to avoid the problem of depending too much on human operator during welding process in the traditional remote welding system, this research proposes a path plan strategy that the human operator real time remotely control the underwater welding robot on the key position, and the underwater welding robot autonomously control the welding process. The present remote-local autonomous path plan strategy serves a theoretical foundation for the development and breakthrough of the intelligent level of deep-sea welding robot. The welding robot path plan revealed the effect mechanism of the welding torch angle, the welding torch swinging and the travel speed on the welding joint quality, build the path plan parameter model based on regression analysis, which offer the technique support for improving the accuracy of the automation welding path plan.This project will serve as a solid foundation for the development of the underwater robotic welding technology, weighing signaficantly both theoretically and in practically.

由水下焊接机器人代替潜水焊工执行深海环境焊接任务是水下焊接技术发展和应用的一个重要方向.本项目主要围绕水下焊接机器人任务空间三维重建和运动路径规划问题，研究宏观遥控与局部自主完成焊接任务的方法.针对深海环境光照条件差、可见度低的特点，提出主动视觉与被动视觉相结合的机器视觉方法，建立人机共享的水下焊接机器人任务空间场景信息，为水下机器人焊接任务分配和规划提供可靠信息依据.针对传统遥控焊接对操作者过度依赖的问题，提出关键位置实时遥控、焊接任务自主控制的深海环境机器人焊接分区域路径规划策略，为水下焊接机器人的智能化水平的提高和突破奠定理论基础.揭示深海环境下焊枪角度、焊枪摆动和焊接速度等焊接路径参数对焊接接头质量的影响规律，建立基于回归分析的路径参数模型，为提高焊接机器人自主规划路径的准确性提供技术支撑.本项目将奠定深海环境机器人焊接技术发展和应用的坚实基础，具有重要理论研究和实际应用价值.

采用较高自动化水平的水下焊接机器人系统代替潜水焊工执行深海环境下的焊接任务，是水下焊接技术走向工程应用的必经之路。本项目建立了一个主动视觉与被动视觉相结合的深海环境立体视觉三维重建系统，深入研究了宏观遥控局部自主的水下焊接实施方法，开发一种人工遥控与局部自主相结合的深海环境机器人焊接路径规划算法。本课题的成功实施，为提高水下机器人智能水平、推动深海环境机器人焊接技术的发展奠定了理论基础。项目执行期间，共发表学术论文6 篇，其中EI 检索3 篇，中文核心期刊3篇。申请发明专利3 项。.根据水下图像降质的物理过程，定量分析了水下光线传播和物体成像过程中的主要影响因素，研究新型的图像增强算法，对水下摄相机采集获得的图片进行清晰化处理。在此基础上，针对深海任务空间场景的特点，开发了海底环境立体视觉中的特征提取、立体匹配算法，为深海环境下焊接机器人的任务层路径规划提供了可靠的三维数字模型。.焊接机器人工作空间场景信息包括周边环境，工件和焊缝三部分。对于上述三种场景信息对象，立体视觉三维重建的精度要求和机器人执行任务的难易程度也有所不同。本项目研究了深海环境焊接机器人任务空间智能分区策略。通过智能分区对任务空间的各区域区别对待，提高了机器人路径规划精度和效率。.深海环境下由于受到海流和水体静压力等因素的多重影响，焊接工艺对焊枪角度和焊枪摆动等参数提出了更加苛刻的要求。本项目针对深海环境机器人在执行一般运动任务和焊接任务是的精度要求，研究了深海环境焊接过程中人工遥控与局部自主相结合的路径规划策略。