一种新的三维测量机器人手眼标定方法

A New Method of Hand-eye Calibration for Robot Measuring System

摘要:

为了提高三维测量机器人的手眼标定精度以更好地满足工业需求,提出了一种采用动态改变变异率的自适应差异进化算法进行手眼标定的新方法；首先构建了由线结构光测头和机器人组成的视觉测量系统,利用齐次坐标变换的方法建立了测量系统的数学模型；然后针对数学模型中的手眼关系,使用半径已知的球体,采用定点变位姿的方法,进行粗略的估计,并对手眼关系求解的精度进行分析；最后采用动态改变变异率的自适应差异进化算法对手眼关系做进一步寻优；实验结果表明,算法具有较高的求解精度；该方法用于手眼标定是可行的、有效的,整个标定过程简单,便于实际应用。

Abstract:

In order to improve hand-eye calibration accuracy of the three-dimensional measuring robot to better meet industry needs,a new method of an adaptive differential evolution algorithm that adopts dynamic changes in mutation rate to conduct hand-eye calibration is proposed. Firstly,the vision measuring system is constructed,which consists of linear structured light measuring head and robot,a mathematic model for measuring system is constructed using homogenous coordinate transform principle. Then,for the hand-eye relationship in mathematical model,which is estimated roughly using a known radius of the sphere and the fixed displacement pose method,and the accuracy of solutions of hand-eye relation is analyzed; Finally,an adaptive differential evolution algorithm with dynamic changes in mutation rate is used to conduct further optimization on hand-eye relationship. Experimental results show that the algorithm has a high accuracy. The method for hand-eye calibration is feasible and effective,and the whole calibration process is simple and easy to the practical application.