摘要
由于交通车流量预测存在不定性、周期性、非线性的特点,传统预测算法受到函数逼近能力的影响,容易陷入局部最优问题.该文在麻雀搜索算法(Sparrow Search Algorithm,SSA)和BP(Back Propagation)神经网络算法(BP Neural Network,BPNN)研究的基础上提出一种交通车流量区间预测优化算法,即SSA-BP预测算法.该算法采用SSA算法来优化BP神经网络算法的初始权值和阈值,利用SSA算法寻优能力强、收敛速度快、稳定性高的特点,在一定程度上解决了BP神经网络算法对初始值依赖度高,易陷入局部最优的问题.通过仿真实验,将改进算法的均方误差降至0.0092,拟合度值为0.9704,说明算法具有良好的泛化能力,能够更好地反映交通流量的变化.
Because the traffic flow prediction has the characteristics of uncertainty,periodicity and nonlinearity,the traditional prediction algorithm is easy to be affected by the function approximation ability and fall into the problem of local optimization.Based on the optimization research of sparrow search algorithm(SSA)and BP neural network(BPNN),this paper proposes an interval prediction algorithm of traffic flow,namely SSA-BP prediction model.The model uses SSA algorithm to optimize the initial weight and threshold of BP neural network model.Using the characteristics of strong optimization ability,fast convergence speed and high stability of SSA algorithm,the problem of high dependence on initial value and easy to fall into local optimization of BP neural network algorithm is solved.Through the simulation experiment,the mean square error of the improved model is reduced to 0.0092 and the fitting degree is 0.97042,which shows that the model has good generalization ability and can better reflect the change of traffic flow.
作者
姚洁
邱劲
YAO Jie;QIU Jin(School of Big Data,Fuzhou University of International Studies and Trade,Fuzhou 350202,China;Fuzhou Management Branch of Fujian Expressway Group Co.Ltd.,Fuzhou 350202,China)
出处
《西南大学学报(自然科学版)》
CAS
CSCD
北大核心
2022年第10期193-201,共9页
Journal of Southwest University(Natural Science Edition)
基金
福建省社会科学基金重大项目(FJ2021Z020).
关键词
车流量区间预测
SSA-BP预测算法
局部最优
泛化能力
interval prediction of traffic flow
SSA-BP prediction model
local optimum
generalization ability