Approximate state-feedback linearization using spline functions (Q1371653)

A system \(\dot x=f(x)+ u\cdot g(x)\) is linearized in a neighborhood of the equilibrium manifold \(f(\overline x)+ \overline u\cdot g(\overline x)=0\) and begins with output function \(z_1=T_1(x)\). The first method extends a recursive algorithm to include splines, the second method is a compromise between gain-scheduling and approximate feedback linearization and uses the software Mathematica. The methods are demonstrated on a rotating inverted pendulum and acrobot-underactuated coplanar double pendulum examples.