CAJ | 학술논문

针舛仿人机器人机械连杆控制结构的动力学、运动学等特点，建立了25自由度仿人机器人家谱树彤结构，提供了连杆数据列表，分析了步态规划约束、步行倒立摆、奇异位姿控制、落地步态方向控制等特性，构建了行走步态模型，并采用链接图法对环境空间建模，引入种群规模自调整遗传算法实现路径规划。潋仿人机器人NAO为实验平台，采用链接图法进行全局建模，实验表明，改进的遗传算法在仿人机器人路径规划中具有更快的收敛速度与更好的适应性。
침천방인궤기인궤계련간공제결구적동역학、운동학등특점，건립료25자유도방인궤기인가보수동결구，제공료련간수거렬표，분석료보태규화약속、보행도립파、기이위자공제、락지보태방향공제등특성，구건료행주보태모형，병채용련접도법대배경공간건모，인입충군규모자조정유전산법실현로경규화。렴방인궤기인NAO위실험평태，채용련접도법진행전국건모，실험표명，개진적유전산법재방인궤기인로경규화중구유경쾌적수렴속도여경호적괄응성。
Aimed at the characteristics of control structure, dynamics and kinematics of the mechanical links in hu- manoid robot,authors set up the genealogy tree structure with 25 degrees of freedom,and the data list of links,ana- lyzed such characteristics as the step planning constraint, walking inverted pendulum,odd gesture control and landing direction of step control and constructed the step walking model. The linking graph was used to construct the envi- ronment space, and the genetic algorithm with self-adjusting group size was used to realize the path planning. The ex- periment proved that the improved genetic algorithm converged faster and was more adaptive than the traditional one by taking the humanoid robot NAO as the platform and using the linking graph to carry out the global modeling.