Multistage approach for trajectory optimization for a wheeled inverted pendulum passing under an obstacle

Abstract A robotic system constructed as a wheeled inverted pendulum (WIP) serves an impressive demonstrator, since this kind of is inherently nonlinear, unstable, and nonminimum phase. These properties may pose several difficulties, when it comes to control trajectory planning. This paper shows method for deriving highly dynamic compliant with the dynamics by means solving optimal problem (OCP) using multiple shooting. The assumed task includes that WIP should pass height-restricting barrier. can be achieved leaning back or forth, in order reduce overall height WIP. constraints inherent definition are nonconvex due shape robot. constraint functions have local minimum infeasible region. cause problems initial guess within To overcome this, multistage approach proposed special OCP evade minimum. After four stages subsequent optimization problems, obtained used feedforward real system.