Earthquakes have often led to many damage points throughout the road-network in Japan and America. Therefore, mass travelers can't be efficiently conducted via the post-quake road-network. Since the road-network is important to maintain the basic life and daily transit after disaster, it is very necessary to develop available reconstruction strategies for the broken road-network. To aid the reconstruction decision for a post-quake road- network, the fuzzy multi-objective model is established as a simulation model of reconstruction scheduling with many work- troops. Furthermore, a technique of asymmetric traffic assignment is also employed as to measure the effectiveness of a reconstruction schedule. Study results show that a satisfying solution can be efficiently derived by our modified genetic algorithm. This solution not only can instruct the reconstruction order, but also can assign the appropriate reconstruction work to relevant work-troops. Nevertheless, this research can be a guideline for the pre-quake exercise.