Robust adaptive decentralized control strategies for robot manipulators based on nonlinear sliding mode

This paper proposed a new class of robust adaptive decentralized control strategies based on nonlinear sliding mode for trajectory tracking of robot manipulators with uncertainties. The controller, which consists of a decentralized nonlinear compensator, a decentralized nonlinear feedback term and a decentralized linear feedback portion, is rather simple. The information required in establishing the control strategies only includes the desired trajectories and output states of the single joint, while detailed description of the model is not required. The main features are that: 1. a class of nonlinear sliding mode and continuous nonlinear compensators are proposed based on some saturation-type functions; 2. some impractical assumptions in the existing literature are removed; 3. a smoother control activity is also achieved. It is shown by strict Lyapunov's theorem and computer simulation that the uncertain effects such as frictions and disturbances can be overcome and the global asymptotic stability is warranted.