Vehicular ad-hoc networks (VANETs) become a popular communication paradigm for vehicles to share traffic information on roads. However, the special form of wireless network provide a frequently intermittent and constantly evolving connection among the VANETs, which may lead to the result that conventional routing protocols designed for mobile ad-hoc networks (MANETs) cannot work properly in some of these dynamic networks. Motivated by the urgent needs of frequently rerouting, quality-of-service (QoS) supporting and limited computational capabilities, we focus on the routing precomputation by making use of a foreseeable network topology under the circumstances of vehicles equipped with Global Position System (GPS) or rail transit network. In this paper, we present our study on designing a routing calculation and optimization strategies to adapt VANETs system. The novel routing strategies have a capability to rapidly response to the frequently change of network topology and ability of congestion avoidance and congestion controlling by a comprehensive employment of Evolving Graphs (EG) model and Ant Colony Optimization (ACO). Through the simulation, our proposed scheme significantly performs superior than the related protocols in the literature in term of end-to-end delay, routing reply ratio, route discovery time and delivery ratio.