A convex approach to rocket ascent trajectory optimization

A convex programming approach for the ascent trajectory optimization of a multistage launch vehicle is presented. Since the problem cannot be readily solved as a convex optimization problem, a combination of lossless and successive convexification techniques is adopted to generate a sequence of convex subproblems having the same solution as the original one. Convergence towards the optimal solution is guaranteed in a limited, short, time by using highly efficient numerical algorithms. The requirement of free terminal time is explicitly accounted for in the optimization procedure. Numerical results are presented and compared with those provided by a differential evolution algorithm.