采用内点法和图像反馈的机器人视觉伺服预测控制

IBVS Predictive Control for Robot Systems with Interior Point Method

针对具有控制量约束和可视性约束的六自由度机器人的视觉伺服系统,研究了基于图像空间局部模型的预测控制器的设计问题.首先对特征点的投影图像的运动学方程进行离散化,得到系统的误差预测模型.然后通过选取合适的性能指标函数,将视觉伺服控制器设计问题转化为一个具有控制量约束和可视性约束的最优化问题.进一步,利用对数障碍函数处理约束,得到系统的牛顿方程,获得控制量的迭代求解公式.最后,利用数值仿真和实验验证了所提方法的有效性.

This paper deals with the design of a predictive controller of a 6-DOF robot based on a local model in image space,subject to input constraints and visibility constraints.First,in order to obtain a systematic error prediction model,the kinematics equations of the feature points of the projected image are discretized.Then,a suitable cost function is selected,and the design of visual servo controller is transformed into a nonlinear optimization problem,in which the input constraints and visibility constraints on the system are considered.Furthermore,these constraints are processed by using the logarithmic barrier function.The Newton's equation of the system is obtained and solved for the iterative equations of control inputs.Finally,the effectiveness of the proposed method is verified by numerical simulation and experiments.