Command filter‐based adaptive control of flexible‐joint manipulator with input saturation and output constraints

Abstract

Abstract This paper focuses on the topic of adaptive neural tracking control for flexible‐joint manipulator systems with output restrictions and input saturation. With the aid of the error compensation mechanism, command filter‐based adaptive neural control is proposed for robotic systems driven by permanent magnet synchronous machine (PMSM). An auxiliary signal produced by the first‐order linear system designed by the property of unmodeled dynamics is used to erase the dynamical uncertainties. The input saturation of system is estimated by ‐function with mathematical transformation. Using barrier Lyapunov function (BLF), each component constraint of output is tackled. During the virtual control design phase, radial basis function neural networks (RBFNNs) may reliably assess the unknown nonlinear continuous function. All the variables in the closed‐loop system are proved to be semiglobally uniform ultimate bounded (SGUUB) by integrating all compensation signals into the overall Lyapunov function and using the defined compact set in the stability analysis of dynamic surface control (DSC). An applied example on robotic system is used to verify the effectiveness of the constructed design idea.