Computational Solutions of Dynamic-Programming Processes — VI : On the Optimal-Trajectory Problem