A global harmony search algorithm and its application to inverse kinematics problem for humanoid arm
全局和声搜索方法及其在仿人灵巧臂逆运动学求解中的应用

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 selfcognitive 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.