| Summary: | There are two common methods for kinematic control of redundant manipulators. One is based on the use of the pseudoinverse of manipulator Jacobian matrix and a so-called null space solution to optimize a secondary criterion. The other way is to expand the rectangular Jacobian matrix to a square one by imposing enough additional equality constraints on the manipulator's motion. In most existing methods, there are some limitations. For example, only one secondary criterion or one set of constraints can be considered. Also, the resultant controller might not perform well when the environment configuration becomes more complex.
|
|---|