1: \begin{abstract}

2: We propose a primal-dual interior-point (PDIP) method for solving

3: quadratic programming problems with linear inequality constraints that typically arise

4: form MPC applications. We show that the solver converges (locally) quadratically

5: to a suboptimal solution of the MPC problem. PDIP solvers rely on two phases:

6: the damped and the pure Newton phases. Compared to state-of-the-art PDIP

7: methods, our solver replaces the initial damped Newton phase (usually

8: used to compute a medium-accuracy solution) with a dual solver based on Nesterov's fast gradient scheme (DFG) 

9: that converges with a sublinear convergence rate of order

10: $\mathcal{O}(1/k^2)$ to a medium-accuracy solution.

11: The switching strategy to the pure Newton phase, compared to the state of the art, is computed in the dual space to exploit the

12: dual information provided by the DFG in the first phase.

13: Removing the damped Newton phase has the additional advantage that our solver

14: saves the computational effort required by backtracking line search. The effectiveness of the proposed solver is demonstrated on a 2-dimensional discrete-time unstable system and on an aerospace application.

15: 

16: \end{abstract}