【作者】 蒋林；

【导师】 赵杰；

【作者基本信息】 哈尔滨工业大学 ， 机械电子工程， 2008， 博士

【摘要】 移动操作臂是集行走和操作功能于一体的复杂机器人系统,是机器人研究领域的一个热点,该机构有较好的操作灵活性和工作空间的广阔性。它和单个系统相比有广阔的应用前景,同时有着重要的理论研究意义。而目前在移动操作机器人的研究中存在定位不准确,无法建立移动操作机器人精确的动力学模型、难以找到有效的移动平台与机械臂间动力耦合辨识方法、难以得到一个较优的冗余度机器人运动规划方法等诸多不足。为解决移动操作机器人面临的难题,本文研制了全方位移动操作机器人(ODMM)及超声波绝对定位系统,并开展ODMM运动学动力学分析、全方位移动平台与操作臂协调运动规划、避障导航、路径规划等研究。首先,将全方位移动平台与操作臂结合,研制的ODMM具有全方位移动和可零半径回转的功能;为解决其定位精度难题,针对超声波传播的特点进行信息融合,提出基于冗余超声波信息融合的超声波绝对定位方法,提高超声波绝对定位的精度,并配合超声波避障模块与数字罗盘技术,实现ODMM的避障导航功能。本文提出了2种避障导航算法:一种为基于超声波绝对定位的ODMM导航算法,这个算法有避障与奔向目标这两个策略;另一个为基于超声波绝对定位的模糊避障导航算法。后一种避障导航算法中,模糊规则是根据人驾车的先验知识建立的,该避障导航算法提高了ODMM的环境适应能力与避障导航能力,仿真及实验证明避障导航算法的有效性和可行性。接着,研究ODMM运动学,并进行正逆运动学分析。由于ODMM是7-DOF冗余度机器人,在运动学逆解时,仅能得到一些量的约束关系而无法得到有效解。为获得有效解,必须添加一些约束方程或优化指标,这为后续运动规划及控制奠定了基础。并采用拉格朗日方法建立ODMM动力学模型,进行动力学仿真,在仿真中研究ODMM的运动速度、加速度与耦合力间的关系。然后对ODMM的全方位移动平台进行运动模式分析,并重点研究ODMM的全方位移动平台与操作臂协调运动的功能。针对ODMM的全方位移动平台与操作臂协调运动提出了2种运动规划方法:其一为基于绝对定位的实时运动规划与控制,该运动规划是在ODMM运动学逆解时,添加一项约束得到有效解;另一种采用种群范围自调整遗传算法,来优化ODMM关节角变化的协调运动规划,该方法很好地解决了冗余度机器人逆运动学求解问题,并充分利用冗余度机器人的运动灵活性进行运动优化。改进遗传算法在ODMM关节角变化优化的效率比传统的遗传算法高,仿真实验证明该协调运动规划方法的正确性与高效性。同时在基于改进遗传算法的协调运动规划的基础上研究了轨迹运动规划问题,并进行ODMM末端空间曲线圆运动规划仿真,仿真结果表明该协调运动规划方法是正确有效的。再次ODMM的路径规划也为本文研究的一个关键技术,它通常分为全局路径规划与局部路径规划。针对局部路径规划中人工势场法存在的不足,提出了基于遗传算法优化改进人工势场法的全方位移动操作机器人路径规划方法。该方法比传统的人工势场法的路径规划精度更好、更合理,随后的仿真实验结果证明该路径规划方法的有效性与合理性。最后,建立ODMM的硬件实验系统,进行ODMM典型轨迹运动实验与运动精度实验,全方位移动平台与操作臂协调运动规划实验,ODMM基于绝对定位的糊避障导航实验。实验结果证明了ODMM设计的合理性,验证了ODMM绝对定位的有效性,运动规划的合理性,避障导航算法的有效性。在结尾处指出了本文研究的不足之处,并对后续的工作作出了展望。本文研制ODMM及超声波绝对定位系统,建立ODMM运动学动力学模型,发展了ODMM定位、导航、运动规划和路径规划技术,为进一步深入研究ODMM和使ODMM进入实际应用奠定了基础。更多还原

【Abstract】 Mobile manipulator was a complex robot system which integrated the functions of walking and operation, now it is a hot subject in the robot research area. This mechanism had better operation flexibility and expands mobile robots working space. Compared with individual system, it has a relatively wide application prospect. Meanwhile it is very important for theory studies. However, there were some shortages: such as imprecise position, unable to set up accurate dynamics analysis model, efficient method to identify mobile manipulator and dynamic coupling, the difficulty in getting a motion planning method of redundant robot. In order to solve the problems which the mobile manipulator faced at present the subject group has developed omni-directional mobile manipulator(ODMM) and ultrasonic absolute positioning(UAP) system and developed the research of some key technology such as ODMM kinematics and dynamics analysis, omni-directional mobile platform and manipulator coordination motion plan, obstacle avoidance navigation, path planning. In summary, the goal of this thesis is to solve the following problems.First, combining omni-directional mobile part with manipulator, and developing ODMM to make ODMM have omni-directional mobile ability, since the turning radius of ODMM was zero, ODMM was able to move freely and flexibly when it work at narrow or crowded places; to solve the problem of position, the fusion method for information was used according to the spreading feature of ultrasound, and the UAP method based on the fusion method for redundancy ultrasonic information was suggested, accuracy of UAP was developed. At the same time, ultrasonic obstacles avoidance module and digital compass information added reliability to navigation. Obstacle avoidance navigation was the key technique in ODMM research. Two obstacle avoidance navigation algorithms were offered in the paper. One was ODMM navigation algorithm based on UAP, whick had two tactics avoiding obstacles and heading for the object. The other was fuzzy obstacle avoidance navigation algorithm based on UAP. Fuzzy rules set up according to one’s driving experience. The obstacle avoidance navigation algorithm improves the adoption and obstacle avoidance ability of the omni-directional mobile manipulator. The simulation and experiment has proved the validity and feasibility of obstacle avoidance algorithm.Next, the ODMM kinematics was studied. The forward and inverse kinematics were analyzed. The unique solution of the inverse kinematics for mobile manipulator could not be obtained due to ODMM was a 7-DOF redundant robot, thus only constraint relationships to some variables could be given, and not unique solution. To get effective solution, some constraint equations and optimization parameters were added based on these constraint relationships and the working conditions, then the satisfactory solution was obtained. The study results described above offer important technique foundations for the following motion planning designs. Meanwhile, ODMM modeling dynamics was set up with Lagrange approach, and research was done on dynamic coupling between omni-directional-motion model. The relationship for rate, acceleration and coupling was achieved via many simulation experiments under different conditions, the dynamics based on the relationship for ADAMS moving along elliptical track and sinusoid track were simulated.Then, different motion modules of omni-directional mobile platform were analyzed. To realize the function of motion coordination between omni-directional mobile platform and manipulator of ODMM, two motion planning methods were given in this paper: one was timing motion planning and control based on absolute position with an additive constraint equation to get control parameter in ODMM inverse value of the kinematics. The other was the improving Genetic Algorithm to optimize coordination motion planning of the change of ODMM joint angle. This method well solves the problem of redundant robot inverse kinematics, and makes full use of the motion flexibility of redundant robot to do motion optimization.The efficiency of improving Genetic Algorithm in ODMM joint angle change optimization was higher than traditional Genetic Algorithm. Simulation experiment proved the correctness and rationality of this coordination motion planning method. At the same time, trajectory tracking and control was analyzed based on improving Genetic Algorithm optimization ODMM joint angle change coordination motion planning. In trajectory tracking planning, the improving Genetic algorithm was used in order to optimize ODMM joint angle change coordination motion planning. The trajectory of 3D circle tracking and control were simulated. The simulation results proved that the coordination motion planning of optimizing ODMM joint angle change based on the improving Genetic Algorithm was correct and efficient.ODMM path planning was the key technique of this subject. It included complete path planning and partial path planning. Understanding the shortage of artificial potential field in partial path planning, the improving artificial potential field path planning method based on Genetic Algorithm optimization in complex environment was set up in the paper. This method was more accurate and rational than traditional artificial potential field path planning. The results of the following simulation experiments have proved the efficiency and rationality of this path planning method.Finally, hardware experiment system of ODMM was set up to do the coordination motion planning between omni-directional-mobile platform and manipulator experiment, ODMM fuzzy obstacle avoidance navigation experiment and ODMM path planning experiment which was based on improving artificial potential field. The result of the experiment had proved the efficiency of motion mobile manipulator absolute positioning, rationality of motion planning, and the efficiency of obstacle avoidance navigation algorithm and route planning algorithm. The shortage of this research is pointed out at the end of the article and the article looks into the follow-up work.This article had studied ODMM and ultrasonic absolute positioning system, set up ODMM kinematics dynamics analysis model, and developed the technology of ODMM positioning, navigation, motion planning and route planning, which has laid the foundation for further research on ODMM and ODMM entry into practical application.更多还原