基于多目标遗传算法的仿人机器人中枢神经运动控制器的设计  

GA BASED SELF-ORGANIZED STABLE HUMANOID ROBOT WALKING PATTERN GENERATORS DESIGN

作　　者：姜山[1] 程君实[1] 陈佳品[1] 包志军[2] 马培荪[2] 

机构地区：[1]上海交通大学信息存储研究中心,200030 [2]上海交通大学机械工程学院,200030

出　　处：《机器人》2001年第1期58-62,67,共6页Robot

摘　　要：针对多自由度仿人机器人的运动控制 ,从神经生理学和机器人学的角度研究了基于中枢模式生成器(CPGs)的仿人运动控制策略 .提出了一种将多目标遗传算法应用于 (CPGs)参数优化的方法 .首先构造用于仿人机器人运动控制的 (CPGs)的结构 ,其参数通过遗传算法按相应的评价函数得到优化 .This paper describes the design of CPGs for stable humanoid bipedal locomotion, using an evolutionary approach. In this research, each joint of the humanoid is driven by a neuron that consists of two coupled neural oscillators. Corresponding joint's neurons are connected by strength weight, to achieve more natural and robust walking pattern, an evolutionary-based multi-objective optimization algorithm is used to solve the weight optimization problem. The fitness functions are formulated based on ZMP and global attitude of the robot. In the algorthms, real value coding and tournament selection are applied, the crossover and mutation operators are chosen as heuristic crossover and boundary mutation respectively. Following evolving, the robot is able to walk in the given environment and a simulation shows the results.

关 键 词：仿人机器人 多目标遗传算法 中枢神经运动控制器 智能机器人 设计 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]