We investigate a goal-robust-optimization approach for solving open vehicle routing problem with demand uncertainty. The approach obtains an optimal solution that minimizes the weighted sum of undesirable deviations from a predetermined time window; for any realizations of the demand-uncertainty set, the solution enables the cumulative travel time for each route to finish within a predetermined time window as closely as possible. To improve the probability of finding exact solution for the robust-optimization model using a heuristic algorithm, we also propose a particle swarm optimization based on genetic algorithms (HPSO-GA) within the framework of hyper-heuristic to solve the goal-robust-optimization model. The computational results demonstrate that the optimal solution obtained by our goal-robust-optimization approach substantially reduced the penalty cost incurred by deviations from a predetermined time window.