Adaptive control of electrically driven space robots based on virtual decomposition

A systematic approach named virtual decomposition is presented for adaptive control of space robots incorporating motor dynamics. Virtual decompositio n imposes a modular structure on both control design and stability analysis . In the control design, the control problem of the complete system is conv erted into the control problem of each subsystem (rigid body or joint), whe reas the nonholonomic constraints are represented by a set of constraint eq uations imposed on the required acceleration. Two alternative joint control modes, namely, motor current control and motor voltage control, are consid ered. parameter adaptation can be carried out independently for each subsys tem, which makes decentralized parameter adaptation possible. In the stabil ity analysis, each subsystem (rigid body or joint) is assigned a nonnegativ e accompanying function. The dynamic interaction between every two physical ly connected subsystems is completely represented by a virtual power flow t hrough their connection. The system Lyapunov function is formed by merely a dding all nonnegative accompanying functions assigned to the subsystems. Ly apunov stability is ensured and computer simulations are conducted. The pro posed approach is a general one that can be extended to treat a variety of space robotic systems due to its modular structure.