摘要
针对交通模型是一个非线性、不确定的复杂动力学系统,难以用精确模型来表达的问题,采用RBF神经网络建立交通流预测模型,具有较强的局部泛化能力,收敛速度快,克服了BP神经网络收敛速度慢、易陷入局部极小的缺点.实例仿真研究表明,该方法预测效果较好.
In the light of the problems of that the traffic model is a nonlinear, uncertain, complex dynamic system and hard to be described completely, the tragic flow forecasting model is successfully constructed by using RBF neural network. The Radial Basis Function neural network with the local generalization abilities and fast convergence speed can overcome the shortcomings of slow convergence speed and local minimum of BP network. The simulation experiment results illuminate that the application of RBF network is fairly effective.
出处
《天津工业大学学报》
CAS
2006年第2期71-73,共3页
Journal of Tiangong University
关键词
交通流
RBF神经网络
预测模型
高斯核函数
traffic flow
RBF neural network
forecasting model
gaussian kernel function