Abstract: In this paper, some important stability of discret-time stochasic dynamic traffic assignment models is proved carefully by stochastic theory. It is found that the feasible time-varying route flows is a irreducible non-periodic homogeneous Markov process in the M-multiple traffic state space and deal with the alongrithm of the stable distribution and average value introduced.

Key words: stochastic dynamic assignment, dynamic stability multiple, Markov process, matheimatical model

摘要： 运用随机过程理论对离散时间随机动态交通分配模型的稳定性进行了严格论证,证明了可行路径流为m重交通状态空间(S_F^m)中的不可约非周期齐时马氏过程,并给出了其稳态分布及其均值表达式,为研究和应用该模型及算法提供了理论基础.

关键词: 随机动态交通分配, 动态稳定性, 多维马尔可夫过程, 数学模型