论文部分内容阅读
为解决概率型随机用户均衡(SUE)配流问题,根据SUE条件(在平衡点,每条使用的路径理解阻抗相同)提出一种更为直观的SUE问题表示和计算方法。该方法通过简单的变换,将理解阻抗的正态分布形式变换成仅需处理一个与流量无关的随机变量的正态分布形式,然后通过MonteCarlo仿真对这个随机变量按其分布多次抽样取值,每次求解一个确定性用户均衡配流问题,多次求解得到的路段流量均值即为服从预定概率分布的SUE配流结果。研究结果表明:提出的方法与传统方法(相继平均算法)计算结果具有较好的一致性,且该方法易拓展到求解随机误差项与测定阻抗相关的SUE问题。
To solve the problem of probabilistic stochastic user equalization (SUE), a more intuitive representation and calculation method of SUE is proposed based on SUE conditions (the same impedance is understood by each used path at the equilibrium point). The method transforms the understanding of the form of normal distribution of impedance into a normal distribution which only needs to deal with a non-flow-related random variable through a simple transformation, and then uses MonteCarlo simulation to sample the value of the random variable according to its distribution multiple times, For each solution to a deterministic user equalization and distribution problem, the mean of the traffic of the road segment obtained through multiple solutions is the SUE distribution result subject to the predetermined probability distribution. The results show that the proposed method has good agreement with the traditional method (successive averaging algorithm), and the method is easy to be extended to solve the SUE problem related to the measurement of impedance with random error term.