摘要
针对当前基于图像的视觉伺服(IBVS)方法难以处理系统约束以及局部渐进稳定等的问题,提出一种新的并行分布补偿(PDC)控制方法。首先,运用张量积(TP)模型变换将视觉伺服系统模型转换为线性时不变系统的凸组合形式;然后,根据并行分布补偿原理将视觉伺服系统的控制变量通过求解线性矩阵不等式的凸优化问题获得,其可行解保证视觉伺服系统的闭环渐进稳定性。该方法除了能够避免直接求解图像雅可比矩阵的逆而无需考虑图像奇异问题外,还易于处理系统约束,根据执行器的机械限制有效规划控制信号的强度。两自由度连杆系统的仿真结果验证了该方法的有效性。
To the problem that the current image-based visual servoing（IBVS） method is difficult to handle the system constraints and its local asymptotic stability, a parallel-distributed compensation（PDC） controller is presented in this paper. First, tensor product（TP） model transformation was applied, which is capable of transforming the model of visual servoing system into the convex combination of linear time-invariant（LTI） models. Then, according to PDC theorem, the parameters of the controller could be obtained by convex program ming techniques for LMIs. It＇ s proved that the feasible solutions of the LMIs ensure the closed-lood asymptotic stability of the visual servoing system. The proposed approach avoids the inverse of the image Jacobian matrix and hence no image singular problem exists. What＇ s more, the system constraints are easy to deal with, which can effectively plan the control signals according to the actuator me chanical limitations. The simulation results for two degrees-of-freedom visual servoing system demonstrate the effectiveness of the algo rithm.
出处
《控制工程》
CSCD
北大核心
2013年第2期334-338,共5页
Control Engineering of China
基金
国家自然科学基金资助(60804013)
