Energy-saving group scheduling is a significant challenge for multi-elevator system in building. Among various traffic patterns in building, most energy is consumed under up-peak traffic pattern. To save unnecessary energy consumption and realize optimal scheduling, an energy-saving scheduling optimization method for up-peak pattern is proposed. The optimization objective is minimizing the scheduling energy price of the group system. The up-peak energy price function is defined first, both of picking and traveling energy are considered in it. Based on that, scheduling Robust Optimization model is built to handle elevator scheduling under uncertain up-peak traffic flow. The key point of optimization is a dual energy-saving mechanism. One is adaptive configuration, choosing fewer but enough number's elevator to transport waiting passengers on the lobby floor. Another is optimization selection, choosing which elevators to serve according to minimum energy price after the number of dispatched elevators is decided. In practice, elevators in group system are dispatched by scheduling decision of the optimization to realize system energy-saving operation under up-peak pattern. In simulation, compared with the scheduling performances of other algorithms under different up-peak flows in several buildings, effectiveness of the method proposed is verified. (C) 2013 Elsevier B.V. All rights reserved.