Adaptive Sliding Mode Neural Network Control and Flexible Vibration Suppression of a Flexible Spatial Parallel Robot

With the goal of creating a flexible spatial parallel robot system in which elastic deformation link causes rigid moving platform to produce small vibrations, we proposed an adaptive sliding mode control algorithm based on neural network. To improve calculation efficiency, finite element method was used discretize link, and then displacement field described floating frame reference coordinates, dynamic differential equation considering high-frequency vibrations established through Lagrange equation. This combined with displacements movable due deformation, highly nonlinear accurate model rigid–flexible coupling effect obtained. Based multi-body dynamics model, driving torque effects calculated advance for feedforward compensation, controller tracking performance system. The error examined determine network’s approximation trajectory errors X-, Y-, Z-directions were reduced by 12.1%, 38.8%, 50.34%, respectively. results showed that designed network met accuracy requirements, suppressed generated link.