Adaptive Neural Control for Euler-Lagrange Systems with Unknown Control Direction

In this paper, the adaptive neural network control is proposed for Euler-Lagrange systems with unknown control direction. The command filters based virtual control signal, adaptive update law and error compensation mechanism are designed respectively under the Nussbaum type function, which can guarantee that the joint position tracking error reaches to a desired region when the control direction is unknown. The used finite-time command filter can guarantee that the output of the command filter fast approximate the differential of virtual signal.