Abstract: To solve computationally intensive problems and low solution accuracy of robot arm in redundancy inverse solution,a solution method based on Glowworm Swarm Optimization algorithm of Time-varying (TVGSO) which evolved based on Glowworm swarm optimization algorithm (GSO) was proposed.Time-varying step size was occurred according to the changes of distance between fireflies and minimum distance of all fireflies individuals within the territory.The dynamic adaptability of the algorithm was enhanced based on golden segmentation method.Secondly,the introduction of the   selection mechanism was introduced to adjust the real-time direction and position of movements dynamically,and the selection pressure during the solving process was also selected.Four simulation experiments showed the simpleness and effectiveness of the proposed algorithm.In the application of inverse solution of the robot arm,the accuracy and stability were improved effectively.

Key words: glowworm swarm optimization algorithm of time-varying, golden section, simulated annealing, redundant robot arm

摘要： 为了解决冗余机器人手臂逆解中存在的计算量大和求解精度不高等问题,提出一种基于时变萤火虫群优化算法的解决方法,该算法基于萤火虫群算法进行演变,根据萤火虫与领域内所有萤火虫个体间的最小距离改变而时变步长,时变步长原理基于黄金比分割法,增强了算法的动态适应性。引入Boltzmann选择机制,实时动态调整搜索过程中移动方向与位置以及求解过程中的选择压力。通过对4个测试函数仿真实验证明,该算法简单有效、稳定性好、求解精度高,在机器人手臂逆解求解应用中,能够有效提高精度和稳定性。

关键词: 时变萤火虫群算法, 黄金比分割, 模拟退火, 冗余机器人手臂