Anti-windup adaptive control of concatenated closed-link hyper-redundant manipulators

This paper addresses the problem of controlling serial link hyper-redundant manipulators (HRM). Due to the large number of degrees of freedom (DOF) in such devices, the associated dynamics are relatively complicated, non-linear, and subject to a high level of parameter uncertainty. This makes the control problem particularly difficult to solve. An adaptive control scheme based on the minimum control synthesis (MCS) algorithm has therefore been proposed that has been successfully used for controlling similarly complex machines and devices across a number of diverse fields. MCS has a number of attractive features: e.g. no a priori knowledge of the robot structure is required, there is no need for parameter identification, no precise adjustment of control parameters is necessary, and stability proofs have been well established. After a brief summary of the basic MCS algorithm, an improved algorithm is introduced, which provides a more robust environment in terms of adaptive gain windup protection. A proof of stability, based upon the passivity criterion, is also provided. Simulation results are included, which demonstrate the control effectiveness of the new adaptive algorithm when applied to a manipulator consisting of serially connected, closed-link units.