Application of two-point boundary problem with optimization of a candidate function

The paper deals with the control design of nonlinear systems. It presents the application of the presented methodology on a swing-up of a single inverted pendulum. It is aimed at planning the reference state trajectories and the appropriate feedforward control signal and formulated by a two-point boundary value problem. One of the complications during the solution tackles the problem of providing a good initial guess for state trajectories. The additional optimization procedure for dealing with this issue is described in this paper. There is an emphasis on the universality of the described solution of the boundary value problem because it can be applied for a variety of different nonlinear systems.