四自由度码垛机器人运动轨迹规划研究

发布时间：2018-12-14 04:42

【摘要】：在工业自动化和物流领域中,码垛机器人是提高生产效率的重要设备,而运动轨迹规划是机器人运动控制最核心的工作。本文以四自由度关节型码垛机器人的轨迹规划为研究课题,以时间最优、脉动最优、能耗最优等多目标最优为规划研究目标,重点进行了时间最优关节柔顺的轨迹二次规划方法研究和基于码垛工作环境因素的多目标最优规划算法研究。首先,对四自由度关节型码垛机器人进行运动学建模和分析。采用经典的D-H方法对各关节及连杆进行运动学分析,建立了机器人正逆解模型,同时通过Adams进行了仿真验证,为之后的轨迹规划研究建立基础。其次,进行时间最优关节柔顺的二次规划方法研究。分析了传统时间最优规划中存在的轴间等待问题,以时间最优为前提,采用柔性非对称S型曲线作为规划的基础曲线,结合采用线性插值法,进行关节加加速度柔和化,最终提出了一种适用的两步规划方法。再者,进行高适应性的多目标最优轨迹规划方法研究。增加考虑码垛工作环境因素,以时间、脉动、能耗最优等多目标为研究对象,以B样条曲线为基础规划曲线,采用模糊层次分析法对各目标对象的权重进行划分;通过NSGA-II算法求得多目标最优化问题的Pareto解集;并最终结合各目标权重,通过近距离理想解法(TOPSIS)和灰色关联度法相结合的改进TOPSIS法进行解集排序,并从中筛选出最优轨迹。最后,进行实验验证。为了对以上两种规划方法进行实验验证,进行了码垛机器人软件系统的设计与实现,并最终在码垛机器人实验平台下,通过重复定位精度实验和实际码垛实验,验证了两种规划方法的正确性和有效性。研究结果表明,时间最优关节柔顺的二次规划法能在不损失效率的情况下,最大限度地减少关节脉动;基于码垛工作环境因素的多目标规划法,灵活性大,适用于码垛工作环境多变和不同码垛需求的场合,且改进TOPSIS法比未改进前更能选出最佳轨迹。以上两种规划方法对于码垛机器人运动控制具有重要意义。
[Abstract]:In the field of industrial automation and logistics, palletizing robot is an important equipment to improve production efficiency, and motion trajectory planning is the core of robot motion control. In this paper, the trajectory planning of a four-degree-of-freedom joint palletizing robot is studied, and the multi-objective optimization, such as time optimization, pulsation optimization and energy consumption optimization, is taken as the research goal. This paper focuses on the research of the trajectory quadratic programming method of the time optimal joint compliance and the multi-objective optimal programming algorithm based on the palletizing working environment factor. Firstly, kinematics modeling and analysis of four-DOF joint palletizing robot are carried out. The classical D-H method is used to analyze the kinematics of joints and connecting rods, and the forward and inverse solution model of the robot is established. At the same time, the simulation by Adams is carried out to establish the foundation for the later trajectory planning research. Secondly, the quadratic programming method of time optimal joint compliance is studied. In this paper, the problem of inter-axis waiting in traditional time optimal programming is analyzed. Based on the premise of time optimization, flexible asymmetric S-shaped curve is used as the basic curve of programming, and linear interpolation method is used to soften the joint acceleration. Finally, a suitable two-step programming method is proposed. Furthermore, a multi-objective optimal trajectory planning method with high adaptability is studied. Considering the working environment factors of palletizing, taking time, pulsation and energy consumption as the research object, taking B-spline curve as the basic programming curve, using fuzzy analytic hierarchy process (FAHP) to divide the weight of each target object. The Pareto solution set of multi-objective optimization problem is obtained by NSGA-II algorithm. Finally, combined with the weight of each target, the improved TOPSIS method combined with the close distance ideal method (TOPSIS) and the grey correlation degree method was used to sort the set, and the optimal locus was screened out. Finally, experimental verification is carried out. In order to verify the above two planning methods, the software system of palletizing robot is designed and implemented. Finally, the experiment of positioning accuracy and actual palletizing is carried out on the platform of palletizing robot. The correctness and validity of the two programming methods are verified. The results show that the quadratic programming method of optimal time joint compliance can minimize the joint pulsation without loss of efficiency. The multi-objective programming method based on the working environment of palletizing is flexible and suitable for different palletizing environments and different palletizing requirements, and the improved TOPSIS method can select the best trajectory better than before. The above two planning methods are of great significance for the motion control of palletizing robots.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】