【作者】 董杰；

【导师】 姜力；

【作者基本信息】 哈尔滨工业大学 ， 机械电子工程， 2019， 硕士

【摘要】 随着医疗技术的发展,静脉采血化验已经成为诊断、治疗过程中必不可少的步骤,而目前临床上只能依靠熟练的医护人员手动采血完成,因此在体力和精神压力上均带来极大挑战。这些医护人员虽经过专门的培训,失败率依旧很高,亟需一种能够替代或协助医护人员完成自动、半自动采血的机器人系统。随着计算机视觉、传感和机器人技术的进步,开发自动静脉采血机器人系统的技术条件逐渐成熟。在巨大市场前景吸引和技术进步的推动下,静脉采血机器人的研究吸引了各国学者的广泛关注。本文针对静脉采血机器人系统的研制开展了深入研究,主要包括以下内容:首先在充分调研市场需求和人工采血工艺流程的基础上,完成了静脉采血机器人的总体方案设计。基于模块化设计理念,完成机器人系统功能模块划分,明确了个模块的技术设计指标。然后完成了静脉采血机器人系统的本体结构、硬件控制系统和传感系统的优化设计,采用改进的串联双平行四连杆机械臂和解耦式挑针设计完成机器人本体结构的优化设计,实现了各自由度的解耦控制,降低了控制难度,提升了开发效率;采用PC上位机-多轴运动控制器-伺服驱动器的架构完成了机器人控制系统的设计,具有人机界面友好、控制精度高、可迭代性好的特点;基于双目近红外相机完成静脉血管识别定位系统设计、基于薄膜压力传感器完成穿刺力实时检测系统设计,为自动实现静脉穿刺控制提供了硬件支撑。接着完成了静脉采血机器人软件系统的设计。人机交互模块采用基于QT的图形化用户界面方便用户操作;控制模块中采用自下而上,逐层封装的方式完成了电机单轴驱动,轴系驱动,机械臂驱动的控制程序;感知模块中采用基于SGBM的三维成像算法对静脉进行三维重构,并通过穿刺过程中的力变化判断血管壁是否穿透。最后在原理样机上进行了实验分析。着重分析了穿刺过程中的穿刺力变化,为实现更为准确的自动挑针提供实验数据,在此基础上完成了半自动的静脉穿刺实验,验证原理样机的功能性,分析性能特点并对下一代产品的设计提供参考。更多还原

【Abstract】 With the development of medical technology,venous blood sampling has become an indispensable step in the diagnosis and treatment process.At present,it is only clinically possible to rely on skilled medical personnel to manually collect blood,so it brings great physical and mental stress.Although these medical staff have been specially trained,the failure rate is still very high,and there is a need for a robotic system that can replace or assist medical personnel to complete automatic or semi-automatic blood collection.With the advancement of computer vision,sensing and robotics,the technical conditions for developing automatic venous blood collection robot systems have gradually matured.Driven by the attraction of huge market prospects and technological advancement,the research of venous blood collection robots has attracted wide attention from scholars from all over the world.This paper has carried out in-depth research on the development of venous blood collection robot system,including the following contents:Firstly,based on the full investigation of market demand and manual blood collection process,the overall design of the venous blood collection robot was completed.Based on the modular design concept,the functional module division of the robot system is completed,and the technical design indicators of the modules are clarified.Then the optimal design of the body structure,hardware control system and sensing system of the venous blood collection robot system is completed.The optimized series design of the robot body structure is realized by the improved series double parallel fourlink mechanical arm and decoupling pick-up design.The decoupling control of each degree of freedom reduces the control difficulty and improves the development efficiency.The design of the robot control system is completed by the architecture of PC-multi-axis motion controller-servo drive,which has friendly man-machine interface and high control precision.It can be iteratively good;the design of venous vessel identification and positioning system based on binocular near-infrared camera,and the real-time detection system design of puncture force based on membrane pressure sensor provide hardware support for automatic venipuncture control.Then the design of the venous blood collection robot software system was completed.The human-computer interaction module adopts QT-based graphical user interface to facilitate user operation;the control module uses a bottom-up,layer-by-layer package to complete the control program of motor single-axis drive,shaft drive,and robot arm drive;The SGBM-based three-dimensional imaging algorithm is used to reconstruct the vein three-dimensionally,and the blood vessel wall is judged by the force change during the puncture.Finally,an experimental analysis was carried out on the prototype.Focusing on the analysis of the puncture force changes during the puncture process,to provide experimental data for more accurate automatic needle picking,on this basis completed a semi-automatic venipuncture experiment,verifying the functionality of the prototype,analyzing performance characteristics and the next generation Product design provides a reference.更多还原