【作者】 罗召成；

【导师】 朱世强；

【作者基本信息】 浙江大学 ， 机械电子工程， 2012， 硕士

【摘要】 在我国工业自动化水平不断提高的大背景下,工业机器人的装机总量也逐年增长,在广大工业生产制造现场得到了良好的应用。根据权威预测,国内机器人安装量将以每年25%的速度递增。运动控制器作为工业机器人运动控制系统的核心,在现代工业控制中的作用日益增大,其性能直接关系到被控对象的工作性能和工作效率。基于实验室自主研发的“钱江一号”6自由度弧焊机器人这一实验平台,以及刘华山、吴剑波等人关于多轴运动控制器的先期工作之上,本课题旨在引入工业现场总线这一通信方式来设计新一款多轴运动控制器,利用其高通用性和高可靠性的特点,实现现有的运动控制器向高开放性和模块化设计方向升级。不仅实现焊接机器人焊接过程中实时轨迹规划以及六个关节协调动作,还能够通过板卡叠加而向更多轴数控制方向的扩展。从长期来看,具有现场总线通信模块的多轴运动控制器可以胜任多机实时交互式通信,从而实现工业现场多机协同作业的功能。本文对于该新一代运动控制器,在充分调研国内外同产品通用设计架构的前提下,确定PC机+控制器的总体设计方案。控制器依然采用DSP作为主控单元,辅以CPLD最终完成多轴协调运动。控制过程中,依靠DSP强大的数据处理能力进行电机控制,CPLD进行相关的逻辑控制和I/O扩展。通过总线调研,选用具有高兼容性和高可靠性的CAN(Controller Area Network)现场总线作为多轴运动控制器板卡和PC机通信媒介,所有实时数据和控制指令通过CAN现场总线传输,完全能够满足其传输的速率和稳定性要求。伺服电机控制方面,论文在简要介绍伺服电机控制模式后,确定控制器采用速度控制模式。并就电机控制策略上,针对机器人系统在仅有位置传感、驱动器饱和、存在建模不确定性及干扰等条件下的轨迹跟踪控制问题,提出了一种新的自适应PID控制方案。通过仿真结果比较得出该方案在系统具有干扰和不确定性因素下具有良好的控制性能。软件上,完成对运动控制器功能的实现,包括CAN现场总线的通信协议实现；PC上位机操作界面的扩展性设计；以及运动控制上的单电机控制算法实现等。最后,运动控制器设计成功后硬件的调试,以及在基于此硬件系统的控制算法上机实验过程和出现的问题,均在本文中得以体现。并且为下一步的研究提供参考。更多还原

【Abstract】 With the development of industrial automation equipment, industrial robot (IR), as one of the most significant manufacturing facilities, has seen dramatically growing these years. According to authoritative estimate, the amount of domestic robot manipulators will increase by 25% every year. While being the core component of IR control system, multi-axis controllers are playing more and more important roles in industrial producing fields. The working performance and efficiencies of IR are largely depended on the quality of Multi-axis controllers. Based on the very experimental platform-- "Qianjiang-Ⅰ" 6-DOF Arc-welding IR and prior effective works by Huashan Liu and Jianbo Wu, this project aims at designing an update multi-axis motion controller(MAMC) by introducing in fielbus, in that case endows this control system more stability and make it a standardized module. The new edition of MAMC board has several advantages over previous and other kind of motion controllers, such as stronger capabilities to get it integrated into other systems and being more appropriate for self-expanding by connecting certain number of controllers together by fieldbus.The new MAMC system not only achieves real-time trajectory planning and coordination of 6 joints in welding process, but can be applied in the occasion where more than 6 joints controlling is needed. Before long, the new MAMC will be suitable for multi-robot communications in fully automatic manufacturing areas, because of fieldbus’ability of real-time data exchange.After full investigation into motion controllers both domestically and abroad, it is determined to adopt the PC+MAMC structure as the final solution. The MAMC system consists of two parts:main control board and port board. Main control board chooses DSP+CPLD as feature implementation chips, while port board functions as power supply isolator and signal filter. With the powerful data-processing capabilities of DSP and excellent logical control programming performance of CPLD, the new MAMC can be qualified to achieve motion controlling with high accuracy. Besides, CPLD can be used for I/O ports expand, which can receive more signals from servo units at one time.When it comes to fieldbus, CAN (Controller Area Network) fieldbus became the best choice due to its wide compatibility and reliability. All the instructions from PC to MAMC and feedback data are transferred through CAN bus.As to servo-motor control, this paper presents some algorithms for control strategy, after discussing control mode and introducing the methodology. The realization of these algorithms by programming together with debugging is also elaborated. Moreover, for trajectory tracking of robot manipulators with only position measurement, limited torques modeling uncertainties and disturbances, it proposed a novel adaptive PID control algorithm. Finally this algorithm is proved to be well performed in two-joint robot manipulator.The set of software achieves several functions:firstly, completing basic motor-controlling of MAMC, like CPLD and DSP instructions programming; then PC human-machine interaction interface and realization of CAN fieldbus communication protocols; and finally, the motion control algorithm.At the end of this paper, the problems during debugging both in hardware and software as well as respective solutions are demonstrated, for the sake of future improvement.更多还原