基于机器视觉的轴承盖外形轮廓分类方法

目的 为提高花卉类包装的自动化程度，解决运动冲击、低执行率等问题，以包装机械手为研究对象，设计一种花卉包装机械手时间和冲击最优轨迹规划方法。方法 采用5次B样条插值函数实现关节曲线构建。采用改进遗传算法实现机械手运动过程优化，不仅可以提高机械手执行效率，而且能够确保其运行过程中不会出现冲击问题。通过改进罚函数和自适应交叉变异概率来解决不等式约束和局部最优问题。另外，设计花卉包装机械手的运动控制系统，包括基于STM32芯片的控制器以及触摸屏控制界面。结果 通过仿真和实验验证了所述方法的有效性，结果表明包装机械手执行效率大约提高了10%，抓取速度可以达到60次/min;机械手抓取迅速、运转平稳，并没有出现明显的冲击问题。结论 所述轨迹规划方法和控制系统可以提高包装机械手控制性能，能够满足花卉包装、分拣要求。

Classification Method of Bearing Caps Contour Based on Machine Vision

The work aims to design an optimal trajectory planning method for time and impact of flower packaging manipulator with the packaging manipulator as the research object to improve the degree of automation of flower packaging and solve the problems of moving impact and low execution rate. The joint curve was constructed by using the quintic b-spline interpolation function. The improved genetic algorithm was used to optimize the manipulator''s motion process, which could not only improve the efficiency of the manipulator but also ensure that there would be no impact during the operation. By improving penalty function and adaptive crossover and mutation probability, inequality constraints and local optimal problems were solved. In addition, the motion control system of the flower packaging manipulator was designed, including the controller based on STM32 chip and the touch screen control interface. Through simulation and experiment to verify the effectiveness of the method, the results showed that:the execution efficiency of the packaging manipulator was increased by about 10%. The grasping speed could reach 60 times/min. The manipulator grasped fast, worked smoothly without obvious impact problem. The trajectory planning method and control system can improve the control performance of the packaging manipulator and meet the requirements of flower packaging and sorting.