为了提高城市道路交通网络的行程时间可靠度,将交通规划专家依照其经验和综合判断能力提出的若干待建、改建道路作为被优化对象,建立了考虑城市路网改造中工程约束的有限离散路网设计双层规划模型。针对该模型设计了基于蒙特卡罗随机模拟技术并考虑约束条件的遗传算法进行求解,将该模型及其求解算法在广州老城区主干路网上进行了实际应用。结果表明:在路网改造成本约束下,该模型及其求解算法能有效提高行程时间可靠度,路网改造方案有较强的工程可行性;研究结果可为城市路网的改善和建设提供决策支持。
In order to improve travel time reliability of urban road traffic network, regarded future road and rebuilt road which were suggested by traffic planners with their experiences and integrative sense as optimal object, considering practical engineering restrictions in urban road network improvement, a discrete hi-level programming model of urban road network design was established. A genetic algorithm with restriction based on Monte-Carlo stochastic simulation was developed to solve the proposed model. The proposed hi-level model and algorithm were implemented on real road network in the aged zone in Guangzhou. The results show that the model and its solution can be used to improve the travel time reliability of urban road traffic network under the restriction of budget and the improvement scheme of road network has engineering feasibility. Outcomes of this research can provide decision support for the development and improvement of urban road network.