Stabilization of nonholonomic chained systems via model predictive control

This paper proposes a design mainly using of model predictive control for the chained non-holonomic systems. Reorganising the chained system into two subsystem, combine a exponential decaying term to deal with the uncontrollable which carry by the vanishing of u₁. After state scaling transformation, a model predictive control (MPC) method is developed for the chained systems. The MPC method iteratively solve a formulated quadratic programming (QP) problem by using a recurrent neural network(RNN) called general projection network(GPN) over a finite receding horizon to get the input vector for the system control. At the end of the paper, simulation results are provided to demonstrate the effectiveness of the proposed approach.