| 姚卫红,方仁孝,张旭东.基于混合人工鱼群优化SVR的交通流量预测[J].,2015,55(6):632-637 |
| 基于混合人工鱼群优化SVR的交通流量预测 |
| Traffic flow forecasting based on optimized SVR with hybrid artificial fish swarm algorithm |
| |
| DOI：10.7511/dllgxb201506011 |
| 中文关键词: 交通流量预测 支持向量回归(SVR) 人工鱼群(AFS)算法 粒子群优化(PSO) 混沌机制 |
| 英文关键词: traffic flow forecasting support vector regression (SVR) artificial fish swarm (AFS) algorithm particle swarm optimization (PSO) chaotic mechanism |
| 基金项目:“八六三”国家高技术研究发展计划资助项目(2012AA111902-2). |
|
| 摘要点击次数: 1418 |
| 全文下载次数: 1867 |
| 中文摘要: |
| 实时准确的交通流量预测是智能交通系统(ITS)中的重要内容．支持向量回归(SVR)能够用于解决交通流量预测问题，针对SVR中存在的参数选择困难，提出一种混合人工鱼群(AFS)算法．该算法利用粒子群优化(PSO)算法公式改进AFS算法，减小AFS算法中步长因子的影响，并引入混沌初始化AFS机制，选取最优SVR参数，建立了基于混沌PSO-AFS优化SVR的交通流量预测模型．仿真结果表明，该交通流量预测模型具有更优的预测性能，证明了其可行性和有效性． |
| 英文摘要: |
| Real-time and accurate traffic flow forecasting is the important content of intelligent transportation systems (ITS). Support vector regression (SVR) can be applied to traffic flow forecasting problem. For the disadvantage of parameter selection of SVR, a hybrid artificial fish swarm (AFS) algorithm is constructed. In order to minimize the impact of the step factor of AFS algorithm the algorithm uses particle swarm optimization (PSO) formulation to improve AFS algorithm. In addition, chaotic initialization of AFS mechanism is introduced. Therefore, the optimal SVR parameters are selected. Then, an improved traffic flow forecasting model which combines the SVR with chaotic PSO-AFS (CPSOAFS-SVR) is constructed. Simulation results show that the traffic flow forecasting model has better prediction performance, and its feasibility and effectiveness are proved. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |