To solve the difficulties of digital integral that exist in global optimization of redundant manipulators, this paper discussed how to build up dynamic equation, the inner relation between constraints and unconstraint optimum control problems, then a digi