论文部分内容阅读
在非线性摩擦锥约束转换为对称矩阵正定线性约束的基础上，首先将抓取力摩擦锥约束进一步转换为线性矩阵不等式约束，使抓取力封闭判别问题转化为带线性矩阵不等式约束的可行解问题，完成抓取的力封闭条件判别，然后根据摩擦锥的正定线性约束矩阵和抓取力平衡约束矩阵的分块特性和稀疏特性，实现多指滚动操作实时抓取力优化计算。最后在多指手实验平台上实现并验证了算法。
Based on the transformation from the nonlinear friction cone constraint to the positive definite linear constraint of symmetric matrix, the constraint of the friction force cone is further transformed into the linear matrix inequality constraint, which transforms the problem of discriminating the grab force closure into a feasible solution with linear matrix inequality constraints Then, according to the block characteristics and sparse characteristics of the positive constraint linear constraint matrix and the grabbing force equilibrium constraint matrix of the friction cone, the real-time optimization of the multi-finger rolling operation can be realized. Finally, the algorithm is implemented and validated on multi-fingered experimental platform.