基于粒子群优化的神经网络实时交通信号配时  

Neural Networks Method Based on Particle Swarm Optimizer for Traffic Real-time Signal Timings

作　　者：孙海英[1] 韦毅华[2] 

机构地区：[1]广东药学院数学教研室,广东广州510006 [2]新乡学院数学系,河南新乡453003

出　　处：《荆门职业技术学院学报》2008年第3期81-84,共4页Journal of Jingmen Technical College

摘　　要：由于人工神经网络具有表征复杂输入输出系统的强大功能,将其用于城市道路交叉口的实时信号配时,是极有应用前景的尝试。文章采用三层人工神经网络建立城市道路交叉口信号配时模型,并把粒子群优化算法作为神经网络的学习算法。该算法具有极高的全局优化形态与计算效率。这种方法建立的模型将有助于降低交叉口的总延误,提高通行能力,使神经网络发挥更好的控制效果。Because neural network can express complex in - out system , applying neural networks to build the signal timing model is a good try. The neural network is applied to build the signal timing model,which is trained with particle swarm optimization. The proposed algorithm demonstrates quite high capabilithm and computing effciency,and can help to shorten the total delay at the route intersection within the same signal .time period.

关 键 词：信号配时 神经网络 粒子群优化 学习算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]