Generation of initial guesses for optimal control problems with mixed integer dependent constraints

Aircraft trajectory optimization problems often include discrete decisions such as flaps settings. In order to use discrete controls in optimal control problems (so-called Mixed Integer Nonlinear Programs, MINLP) reformulation methods are necessary to create a continuously differentiable problem. One of the most promising methods is the Variable Time Transformation (VTT) which is able to find the optimal switching sequence through optimization. However, the optimal result highly depends on the initial guess provided, if it even converges. This is especially true if discrete controls are involved. The aim of this paper is to produce more suitable initial guesses, which help to solve the MINLP fast and numerically stable.