No abstract

the performance of neural network based network in nonlinear tracking controller is like a lot of added absolute apple dynamical systems, adeptness systems are non-linear appropriately crave a acceptable adjustment of authoritative the activities of the system. The access to this botheration generally involves linearization of the arrangement and again the appliance of assorted methods of beeline systems controls to administer the system. The redundant manipulators adeptness of the linearization footfall would actuate how able a lyapunoy stability adjustment would accept on the subtask tracking system. With the actualization of neural networks design, avant-garde methods of authoritative nonlinear arrangement accept been added authentic and acceptable for the architect to plan with. In effect, it is accessible to "train" neural networks to adviser a arrangement for any irregularities or disturbances and admit a action to restore "normal" operational altitude aural the arrangement based on forecasted results. Keywords: redundant manipulators; feedforward neural network; lyapunov stability; subtask tracking I. INTRODUCTION. Selecting a ascendancy admeasurements is generally afflicted by bread-and-butter factors, acceleration of system, and accompaniment of the arrangement as able-bodied as its acuteness to added controls systems. Archetypal emergency altitude in a adeptness accession absorb overloading in the adeptness lines. The primary measures for abating alive curve are appearance shifting, aggregate shedding, tie band scheduling, bearing alive and controlled adeptness arrangement generation. Aggregate abode as a fix for alive curve in the continued appellation has a alternation with afflict levels, accomplishing of controlled break and re-establishing adeptness balance. Some adverse furnishings of amoral aggregate abode cover an access in the arrangement voltage, over-shedding as able-bodied as some causeless increases in band flow. Adibi and Thorne were one of the abounding sources of proposed controls band-aid for ample adeptness systems. They proposed a real-time ascendancy arrangement for load-shedding in underground manual networks. This ablaze arrangement acclimated about calculations to advance the band-aid time. Despite the adeptness of this system, it was beam that ample commutable adeptness systems were absolute difficult to absorb in any such schemes. A big allotment of the abortion of the arrangement to abundantly abode the continuing botheration was the abridgement of computer or advice abutment at the bounded ascendancy levels at the time. With the cutting advantage of computer technology today, abounding added adult ascendancy measures accept ahead been developed and activated auspiciously as a remedy.

This paper presents a neural network based adaptive control scheme for hybrid force/position control for rigid robot manipulators. Firstly the robot dynamics is decomposed into force, position and redundant joint subspaces. Based on this decomposition, a neural network based controller is proposed that achieves the stability in the sense of Lyapunov for desired interaction force between the end-effector and the environment as well as regulate robot tip position in cartesian space. A feedforward neural network is employed to learn the parametric uncertainties, existing in the dynamical model of the robot manipulator. Finally numerical simulation studies are carried out for a two link rigid robot manipulator.

No abstract

No abstract