摘要
提出一种基于BP神经网络的遗传算法,分别利用其局部和全局寻优能力强的特点,综合为一种新的优化算法,并将改进的算法应用于交通流预测中。结果表明,改进方法的预测效果优于单一使用BP神经网络进行预测的效果。
The paper proposes a genetic algorithms based on BP neural networks . It uses their local and global characteristic to put forwards a new optimization algorithms. And uses the new algorithms in traffic flow forecasting. The results shows that the forecasting effect of the new algorithms is better than the BP neural networks.
出处
《交通科技与经济》
2009年第4期10-12,共3页
Technology & Economy in Areas of Communications
关键词
遗传算法
BP神经网络
交通流预测
genetic algorithms
BP neural networks
traffic flow forecasting
