A comparison of Filippov sliding vector fields in codimension 2 (Q2252361): Difference between revisions

The authors propose several methods for the choice of a Filippov sliding vector field on a codimension-2 surface \(\Sigma\). Specifically, they consider a piecewise-smooth dynamical system having two discontinuity boundaries (surfaces) of codimension-1, which locally divide the phase space into four regions, each of which assigned a smooth vector field describing the system behavior in the corresponding region. Therefore, the main question consists in how to define an appropriate sliding vector field on the surface \(\Sigma\) resulting from the intersection of the codimension-1 discontinuity boundaries. The study is restricted to the case when \(\Sigma\) is nodally attractive, i.e., trajectories near \(\Sigma\) are attracted to this surface and reach it in finite time. The key step in the paper is to construct a sliding vector field following the classical convexification method proposed by Filippov. By doing this, the problem is reduced to solving an underdetermined system of linear equations, and the main concern is how to find nonnegative solutions that vary smoothly with respect to the state variable. The rest of the paper is devoted to describing and comparing various techniques to construct such solutions based on analytical and geometric approaches, among which the method of the so-called barycentric coordinates seems to be the most effective one. The authors illustrate all proposed methods by examples.