The inverse kinematics (IK) problem of humanoid arm can be solved by using numerical optimization method, which is essentially an equivalent minimization problem. The harmony search (HS) is a meta-heuristic optimization method which mimics behaviors of music players in an improvisation process. To the best of our knowledge, there is very little research work on HS for inverse kinematics problem of robot manipulator. In this paper, an effective global harmony search algorithm (GHSA) based on the swarm intelligence is developed to solve the optimization problem. The GHSA combines the particle swarm optimization (PSO) with HS, and adopts the PSO operation to produce new improvisation instead of regular search rules of harmony search. The improvisation step based on the movement of harmony particles with self-cognitive and swarm behavior in GHSA makes the algorithm strive for a well balance between the global exploration and the local exploitation. Also, the GHSA performs the uniform mutation operation to get rid of the local optimum. The experimental results of benchmark functions show that the GHSA algorithm greatly improves both the global optimization performance and the reliability performance. Based on these, the 7-degree of freedom (7-DOF) redundant humanoid arm is used as an example, and the end-effector error (position and orientation) and the comfortable level of the humanoid arm constitute the fitness function of the GHSA. The proposed GHSA has been applied to solve the inverse kinematics problem of the 7-DOF redundant humanoid arm; numerical simulation results demonstrate the effectiveness of this algorithm.%仿人灵巧臂逆运动学(IK)问题可转化为等效的最小化问题,并采用数值优化方法求解.和声搜索(HS)是模拟乐师在音乐演奏中调整音调现象的一种启发式搜索方法,目前还尚未在机器人机械臂逆运动学问题中得到应用.本文提出一种基于粒子群体智能的全局和声搜索方法(GHSA),该方法在和声搜索算法中引入微粒群操作(PSO),采用粒子群策略替代常规和声搜索算法中的搜索法则创作新和声,通过粒子自身认知和群体知识更新和声变量位置信息平衡算法对解空间全局探索与局部开发间能力；同时算法还引入变异操作增强算法跳出局部最优解能力,基准函数测试表明该方法改善了全局搜索能力及求解可靠性.在此基础上以七自由度(7-DOF)冗余仿人灵巧臂为例,考虑以灵巧臂末端位姿误差和“舒适度”指标构建适应度函数并采用GHSA算法求解其逆运动学(IK)问题,数值仿真结果表明了该方法是解决仿人灵巧臂逆运动学问题的一种有效方法.
展开▼
