系统最优的动态交通网络流分配模型的分批并行算法

Batch Parallel Algorithm of Assignment Problem of Dynamic Transportation Network Flows for System Optimization

智能交通系统是一个复杂的综合系统 ,动态交通流分配是它的一个关键问题 ,是运输科学的一个前沿和热点。对于已经建立的系统最优的动态交通流分配模型 ,根据非线性规划问题的Kuhn Tucker条件求解出问题的解。由于交通网络是一个大系统 ,结合分批调度理论将交通网络进行分批处理后简化了配流过程中的寻径 ,最后设计出分批并行调度算法。

Intelligent transportation system is a complex integrated system. The study on dynamic transportation network assignment is one of its key techniques and is also the frontier and hot point of transportation science. In this paper, according to a model on the assignment problem of dynamic transportation network flows for system optimization, the solution to assignment problem is obtained based on Kuhn Tucker conditions of non linear programming. Because transportation network is a complex large system, the route searching steps for assignment are simplified by processing the transportation network flows using the batch scheduling theory, and a batch parallel scheduling algorithm of assignment problem is designed.