应用行程时间信息反推OD矩阵的模型与算法研究

Study on models and methods of OD matrix estimation based on travel time information

为提高OD反推的精度,提出一种新的双层优化模型,在原来模型的目标函数上加入行程时间的误差项,使更多的信息源数据可以被利用在OD矩阵反推中.采用一种能近似获得全局最优解的模拟退火算法来求解这个新模型.实例验证表明引入行程时间信息的新模型反推得到的OD矩阵解比仅利用流量信息的原模型更准确.

In order to improve the OD estimation accuracy,this paper proposed a new bi-level programming model,of which the objective function has an additional term of travel time square errors,so more information source data can be used in estimating origin destination matrix,A method of simulated annealing is used to solve the global optimal solution of the new model.Numerical examples show that the solution of the new model that introduces travel time information is more accurate than that of the old models.