包装生产线的操作臂轨迹规划的一种优化方法

An Optimization Method for Trajectory Planning of Manipulator in Packaging Production Line

中文关键词: 惯性权重  收敛因子  反向学习机制  优化的树形拓扑结构  时间最优轨迹规划

英文关键词: inertia weight  convergence factor  reverse learning mechanism  optimized tree topology  time optimal trajectory planning

中文摘要:

      目的 为提高仓库存储、包装生产领域操作臂运动效率，提出一种基于混合优化的鲸鱼算法(MAIWOA)来求解操作臂始末2点运动过程中时间最优值。方法 首先是通过混沌算法优化初始化种群;其次，提出一种对勾函数和包围收缩学习机制来帮助算法跳出局部收敛;然后利用改进的反向树形拓扑结构来提高种群探索的多样性;采取改进收敛因子和自适应权重机制来平衡前中期的全局探索能力和后期的收敛性。结果 该算法在收敛速度和收敛精度上均有可观的提升。将该算法应用到五次多项式插值关节运动时间最优求解上取得了很好的效果，关节1、2、3的运动时间由20 s分别缩减到了9.642 5、9.251 5、10.787 s，效率分别提高了51.79%、53.74%、46.07%。结论 将MAIWOA应用到一般性操作臂轨迹规划上，可提高生产线上操作臂的执行效率。

英文摘要:

      The work aims to propose a whale algorithm based on hybrid optimization (MAIWOA) to solve the time optimal value of the manipulator in the process of two-point movement, so as to improve the movement efficiency of the manipulator in warehouse storage, packaging and production. Firstly, the initial population was optimized by chaos algorithm; Secondly, a learning mechanism of tick function and bounding contraction was proposed to help the algorithm jump out of local convergence; Then an improved reverse tree topology was used to improve the diversity of population exploration; The improved convergence factor and adaptive weight mechanism were adopted to balance the global exploration ability in the early and middle stages and the convergence in the later stage. Results showed that the convergence speed and accuracy of this algorithm were improved significantly. The algorithm was applied to the optimal solution of joint motion time of quintic polynomial interpolation, and good results were obtained. The motion time of joints 1, 2 and 3 was reduced from 20 s to 9.642 5, 9.251 5 and 10.787 s respectively, and the efficiency was increased by 51.79%, 53.74% and 46.07% respectively. It is concluded that applying MAIWOA to the trajectory planning of general manipulator can improve the execution efficiency of manipulator in production line.