In this paper, an integrated rule-based meta-heuristic optimization approach is used to deal with a multi-level energy management system for a multi-source electric vehicle for sharing energy and power between two sources with different characteristics, namely one with high specific energy (battery) and other with high specific power (SuperCapacitors). A first (long-term) management level dynamically restricts the search space based on a set of rules (strategic decisions). A second (short-term) management level implements the optimization strategy based on a meta-heuristic technique (tactical decisions). The solutions to the optimal power sharing problem are be used to generate the power references for a lower (operational) level DC-DC converters controller. The Simulated Annealing meta-heuristic is used to define an optimized energy and power share without prior knowledge of power demand. The proposed scheme has been simulated in Matlab (R), with models of energy sources for several driving cycles. Illustrative results show the effectiveness of this multi-level energy management system allowing to fulfill the requested performance with better source usage and much lower installed capacities. (C) 2013 Elsevier Ltd. All rights reserved.