PSO算法在交叉口OD矩阵推算中的应用

Application of PSO Algorithm in Intersection OD Matrix Estimation

针对由交叉口已知交通量推算OD矩阵时,当已知交通量数小于OD变量数时难以得到满意解的问题,提出了采用基于群智能技术改进的粒子群优化算法(Particle Swarm Optimization,简称PSO)来进行全局优化求解.论文设计了用粒子群算法求解交叉口OD矩阵推算模型的方法,确定了目标函数和终止条件,给出了具体的计算步骤及粒子群算子的选择,通过初始化粒子的速度和位置,不断迭代更新直到搜索出全局最优值.最后用Matlab语言编程进行了仿真试验.仿真结果表明,该方法具有较高的效率和准确性.

The paper for using the intersection existing traffic counts to estimate OD matrix, as the existing traffic counts are less than the OD variables, it is difficult to get the satisfied solution. So an improved particle swarm optimization algorithm （PSO） is presented to get the global optimal solution. The paper designs the method of solving intersection OD matrix calculated model with PSO, determines the objective function and the terminable conditions, gives specific calculated steps and choice of PSO operator, then initialize the position and velocity of particles, renew iteratively until to search the global optimum. Finally do the simulation test with Matlab programming. The simulation results show that the method has higher efficiency and accuracy.