毕业生知识库

随着我国经济的高速发展，我国机器人需求和销量迅猛增长，但我国机器人渗透率远低于发达国家，因此机器人行业前景广阔。其中，高速Delta并联机器人具有运动速度快、定位精确、成本低、效率高等优点，被广泛应用于医药分装、食品包装和电子产品等工业自动化生产或包装流水线的快速分拣、抓取、装配中。而国内的并联机器人起步晚，竞争力较薄弱，目前我国高速高精度Delta并联机器人系统主要依赖于国外的先进技术，因此加快Delta并联机器人关键技术研究及其产业化，能有效提高我国工业自动化水平，打破国外品牌的垄断地位及技术封锁，有利于并联机器人关键技术突破及国产化，强化基础制造装备竞争力。
本文围绕Delta并联机器人运动学及动力学建模、轨迹规划、系统控制算法、高速实时工业以太网等关键技术进行了深入研究，完成的主要工作及创新点如下：
(1) 针对Delta并联机器人的机械结构，计算得到其工作空间，并分析求解了运动学模型，包括Delta 机器人位置模型、速度模型、加速度模型。然后，利用虚功原理建立了Delta机器人的动力学模型。
(2) 针对Delta并联机器人难以在运动环境下有效分拣动态目标的问题，提出了基于费拉里法的Delta并联机器人动态目标抓取算法。首先介绍了修正梯形运动规律和3-4-5次多项式运动规律，之后对拾放操作进行了轨迹规划，并对轨迹的不同情况进行了分析。然后分别对不同运动规律下的运动环境中的目标抓取问题建立数学模型，针对该模型，提出了基于费拉里法的Delta并联机器人动态目标抓取算法来确定抓取点位置，实现了动态目标的准确快速抓取。算法测试与实验结果表明该算法能精准地计算抓取位置，速度快，能满足实时性要求，且不需设置初值，使用便捷。在此基础上，通过对两类工件的轨迹规划模型进行分析，提出了多类工件的轨迹规划算法。首先确定要排序的工件，然后再通过动态抓取算法计算拾取点位置。最后通过实验验证了上述算法的稳定性和有效性。
(3) 针对Delta并联机器人高速作业下笛卡儿空间轨迹不平滑的问题，提出一种基于毕达哥拉斯速端曲线（PH曲线）的轨迹规划方法。首先，在笛卡儿空间中利用PH曲线平滑竖直运动与水平运动间的直角过渡部分，确定拾放操作轨迹；然后，利用多项式运动规律对轨迹的一维曲线位移进行规划，确定运动轨迹插补点的位置；最后，以最小化拾放操作周期为目标优化PH曲线参数，得到平滑的运动轨迹。仿真和实验结果表明，该方法下的拾放操作具有较短的运动周期，轨迹平滑且具有良好的运动特性，实现了并联机器人的高速作业。
(4) 为了实现机器人的轨迹跟踪控制，提出了一种基于Delta机器人动力学模型的滑模控制方法。在给定了并联机器人期望运动轨迹后，采用滑模控制方法设计控制器，并通过求解并联机构的动力学反问题得到期望轨迹所需要的力/力矩，然后控制各主动关节的驱动装置输出相应的力矩，从而使并联平台按照期望轨迹运动。针对期望轨迹位置点数不够的情况，对轨迹生成的实现方法进行了介绍。仿真实验证明了该方法的有效性。
(5) 设计了Delta机器人控制系统，包括硬件结构和软件设计。在此基础上，针对Delta并联机器人的高速运行的需求，将EtherCAT工业以太网应用于该系统，设计了基于EtherCAT总线的Delta机器人系统，包括EtherCAT从站和主站的设计。在Delta机器人系统实验平台上，完成了基于拾放操作的高速分拣实验。

With the rapid development of China's economy, both the demand and sales of robots in China have increased rapidly. However, the penetration rate of robots in China is much lower than that in developed countries, so the robot industry has a promising and bright future. Among all the kinds of robots, the high-speed parallel Delta robots have the advantages of high velocity, accurate positioning, low cost, high efficiency, etc, which makes it widely used in industrial automated production or packaging lines of medicine, food and electronic products to complete rapid and accurate sorting, grabbing and assembly. While domestic parallel robots started late and have relatively weak competitiveness. At present, China's high-speed parallel Delta robot systems mainly rely on advanced foreign technologies. Therefore, speeding up the research of key technologies for Delta parallel robots and the industrialization of Delta parallel robots can effectively improve the level of industrial automation in China, break the monopoly position of foreign brands and technology blockade, facilitate the breakthrough and localization of key technologies for parallel robots, and strengthen the competitiveness of basic manufacturing equipment.
This dissertation researches deeply on the key technologies for Delta parallel robots, including kinematics and dynamics models, trajectory planning, control algorithm and high speed real time industrial bus. The main contributions and innovations of this thesis are as follows:
(1) For the mechanical structure of the Delta parallel robot, the working space is calculated. The kinematics model of Delta robot are analyzed and solved, including the position model, velocity model and acceleration model of Delta robot. Then the dynamic model of the Delta robot is established using the principle of virtual work.
(2) To solve the problem of picking up dynamic targets effectively for Delta parallel robot in the moving environment, a dynamic picking algorithm based on Ferrari’s method is proposed. First, the modified trapezoidal motion law and the 3-4-5 polynomial motionlaw are introduced. Afterwards, the trajectory planning for pick-and-place operation is performed, and the different conditions of pick-and-place trajectory are analyzed. Then the mathematical model for the target picking problem in the moving environment under different motion laws is established respectively. According to the mathematical model, the dynamic picking algorithm based on the Ferrari’s method is proposed to determine the picking position in order to achieve accurate and fast picking of dynamic targets. The test results and experimental results show that the algorithm is accurate and fast, which can meet the real-time requirements. And this algorithm does not need to set the initial value, less restrictive conditions make it more convenient. On this basis, by analyzing the trajectory planning models of two types of workpieces, the trajectory planning algorithm of sorting multiple types of workpieces is proposed. Firstly the workpiece to be sorted is determined, and then it will be picked up at the picking point which is calculated by dynamic picking algorithm. Finally, experiments prove the stability and effectiveness of the above algorithm.
(3) The trajectory planning method based on PH (Pythagorean Hodographs) curve is proposed to solve the problem that the trajectory of Delta parallel robot in Cartesian space is not smooth when performing high-speed operations. The pick-and-place operation trajectory is determined by applying the PH curve to smooth the transition portion between the vertical movement and the horizontal movement in Cartesian space. The position of the trajectory interpolation point is determined by the trajectory planning for the one-dimensional curve displacement based on the polynomial motion law. In order to minimize the movement cycle time of the pick-and-place operation, the PH curve parameters are optimized to smooth the motion trajectory. The simulation and experimental results show that the pick-and-place operation under the proposed method has a short motion cycle, smooth trajectory and good motion characteristics, and realizes high-speed operations of parallel robots.
(4) In order to realize the trajectory tracking control of the robot, the sliding mode control based on the dynamic model of the Delta robot is proposed. After giving the reference trajectory of the parallel robot, the controller could be designed by sliding mode control, and the force required for the reference trajectory is calculated by solving the inverse dynamic problem of the Delta robot. Then the motor driver of each active joint is controlled to output the corresponding torque so that the parallel robot could move along the reference trajectory. Afterwards, the trajectory generation method is introduced for the situation where the numbers of reference trajectory points are not enough. This thesis considers the circular trajectory as the the reference trajectory to be reached. Simulation experiments demonstrate the effectiveness of the method.
(5) The Delta robot control system is designed in this thesis, including hardware structure and software design. On this basis, EtherCAT is used for the robot system to satisfy the demand for high-speed operations of Delta parallel robots. The Delta robot system based on EtherCAT is designed, including the design of the EtherCAT slave and master. The high-speed pick-and-place operation experiments are completed on the Delta robot system platform.