摘要
为研究机械臂的路径规划问题,将传统的退火算法与遗传算法相结合,提出了一种改进的模拟退火遗传算法。该算法不仅能自适应改变遗传过程中的交叉概率和变异概率,还加入了局部退火操作,对每代子种群进行退火处理以克服路径规划过程中陷入局部最优的问题。此外,引入整体退火观念,在整个进化过程中,随着温度的降低,不断拉伸适应度函数提高算法效率。将该算法应用于实验室自主研制的七自由度轻型冗余机械臂上,以时间冲击最优为目标进行轨迹规划。实验表明,该算法可以在保证各项约束的条件下获得最优路径。
By embedding the simulated annealing algorithm into the genetic algorithm,an improved simulated annealing genetic algorithm is introduced to research the trajectory planning of manipulator.The algorithm can change the probability of crossover and mutation in different conditions to improve the self-adaptability.Meanwhile,the spot annealing operation can be used in each generation of subpopulation to overcome the problem caused by local minimum.In addition,the concept of entire annealing is introduced.To increase the efficiency of the algorithm,the fitness function is stretched as temperature decreasing during the evolution.By taking the tim e-jerk optimization as the target,the algorithm is applied to trajectory planning of the 7-DOF manipulator.The experiment results show that the generated trajectories satisfy all the constraints.
作者
宗玉杰
崔建伟
ZONG Yu-jie;CUI Jian-wei(School of Instrument Science and Engineering,Southeast University,Nanjing 210096,China)
出处
《测控技术》
CSCD
2018年第3期1-5,共5页
Measurement & Control Technology
关键词
遗传算法
模拟退火算法
路径规划
时间冲击最优
genetic algorithm
simulated annealing algorithm
trajectory planning
time-jerk optimization