粒子群算法在动态交通分配问题的应用本科毕业设计五邑大学.docVIP

五邑大学本科毕业设计 PAGE PAGE 23 摘 要 动态交通分配在交通控制与诱导中起重要作用，因其能充分考虑交通路网中的复杂性、时变性和随机性等典型交通流特性，比静态交通分配更有优势。传统的优化算法计算量大且容易使性能指标陷入局部最优，极大地限制算法在理论和实际上的应用，采用粒子群算法可以使求解变得简洁和方便。 本文主要研究动态系统最优模型，通过分析动态交通分配的特点，构建了动态交通分配模型，并运用引入惯性权重的改进粒子群算法，以总花费时间最小为优化目标对多路径交通网络进行研究。最后在多路径小型交通网络中进行MATLAB仿真，仿真结果体现了该模型的有效性和改进的粒子群算法的优越性，提高了模型的应用价值。 关键词 动态交通分配；粒子群算法；优化算法；多路径；仿真 Abstract Dynamic traffic assignment plays an significant role in traffic management and guidance. It has an advantage over the static traffic assignment, because it takes full consideration of the typical traffic flow characteristics such as complexity, time variation and probability, etc. Traditional optimization algorithms seriously restrict the application and development of the model in a large amount of calculation or easy to fall into local optimal value of performance index, but it can be simple and convenient to solve such problems via particle swarm optimization(PSO). This paper mainly studies the optimal model of dynamic system, which is based on the minimum of total spent time as the objective function. The dynamic traffic assignment model is built via to analysis its characteristics, and using improved particle swarm optimization with inertia weight factor. Finally, MATLAB simulation is carried out in a multipath traffic network, the simulation result shows the validity of dynamic traffic assignment model, the superiority of improved particle swarm optimization and the value of the model. Key words dynamic traffic assignment PSO algorithm optimization multipath simulation 目 录 TOC \o 1-3 \u 摘 要 PAGEREF _Toc357775731 \h I Abstract PAGEREF _Toc357775732 \h II 第1章 绪论 PAGEREF _Toc357775734 \h 1 1.1 引言 PAGEREF _Toc357775735 \h 1 1.2 智能交通系统概论 PAGEREF _Toc357775736 \h 1 1.2.1 智能交通系统研究现状 PAGEREF _Toc357775737 \h 1 1.2.2 智能交通系统的组成 PAGEREF _Toc357775738 \h 2 1.2.3 城市智能交通控制与管理系统 PAGEREF _Toc357775739 \h 3 1.3 动态交通分配的研究现状 PAGEREF _Toc357775740 \h 4 1.4 动态交通分