基于改进粒子群算法优化小波神经网络的短时交通流预测
|
摘要
传统的小波神经网络预测模型,通常采用单向梯度下降法进行参数优化,但其存在收敛速度慢和局部最优等问题.为了提高城市道路短时交通流的预测精度,提出一种改进的粒子群算法优化小波神经网络预测模型.该算法可以调整惯性权重和学习因子,以改善粒子群算法后期收敛速度慢、局部搜索能力弱等缺点.最后将模型应用于短期交通流的实证研究,结果表明,与传统的小波神经网络和蚁群算法优化小波神经网络预测模型相比,提出的模型预测的结果误差更小,且具有较快的收敛速度和较好的非线性拟合能力.
Traditional wavelet neural network prediction models usually use one-way gradient descent method to optimize parameters,but there are some problems such as slow convergence speed and local optimum. In order to improve the prediction accuracy of short-term traffic flow on urban roads,an improved particle swarm optimization wavelet neural network prediction model is proposed. The model can adjust inertia weight and learning factor to improve particle swarm optimization. Finally,the model is applied to the empirical study of short-term traffic flow. The results show that compared with the traditional wavelet neural network and ant colony optimization wavelet neural network prediction model,the proposed model has smaller error,faster convergence speed and better non-linear fitting ability.
Traditional wavelet neural network prediction models usually use one-way gradient descent method to optimize parameters,but there are some problems such as slow convergence speed and local optimum. In order to improve the prediction accuracy of short-term traffic flow on urban roads,an improved particle swarm optimization wavelet neural network prediction model is proposed. The model can adjust inertia weight and learning factor to improve particle swarm optimization. Finally,the model is applied to the empirical study of short-term traffic flow. The results show that compared with the traditional wavelet neural network and ant colony optimization wavelet neural network prediction model,the proposed model has smaller error,faster convergence speed and better non-linear fitting ability.
引文
[1]Kumar K,Parida M,Katiyar V K.Short term traffic flow prediction in heterogeneous condition using artificial neural network transportation[J].Transport,2013,30(4):1-9.
[2]Wang L,Zeng Y,Chen T.Back propagation neural network with adaptive differential evolution algorithm for time series forecasting[J].Expert Syst.Appl,2015,42(2):855-863.
[3]Crovella ME,Bestavros A.Self-similarity in world wide web traffic:evidence and possible cause[J].IEEE/ACM Trans Netw(TON),1997,5(6):835-846.
[4]Jiang X,Adeli H.Dynamic wavelet neural network model for traffic flow forecasting[J].Transp.Eng,2014,131(10):771-779.
[5]韩超.基于混沌理论的短时交通流预测方法[J].物流工程与管理,2012,34(4):116-117.
[6]Yu H.Network traffic prediction based on wavelet analysis and hopfield neural network[J].Comput Appl Softw.2013,30(6):246-249.
[7]Li HY,Li D.Study on network traffic forecasting based on improved neural network[J].Comput Simul,2011,28(9):132-135.
[8]唐新来,李春贵,王萌,等.基于混沌粒子群算法的神经网络短时交通流预测[J].计算机测量与控制,2010,18(8):1893-1895.
[9]Sang A,Li S Q.A predictability analysis of network traffic[J].Comput Netw,2000,1(1):329-345.
[10]Hui M,Bai L,Li Y B,et al.Highway traffic flow nonlinear character analysis and prediction[J].Math.Prob.Eng,2015,25(8):1-7.
[2]Wang L,Zeng Y,Chen T.Back propagation neural network with adaptive differential evolution algorithm for time series forecasting[J].Expert Syst.Appl,2015,42(2):855-863.
[3]Crovella ME,Bestavros A.Self-similarity in world wide web traffic:evidence and possible cause[J].IEEE/ACM Trans Netw(TON),1997,5(6):835-846.
[4]Jiang X,Adeli H.Dynamic wavelet neural network model for traffic flow forecasting[J].Transp.Eng,2014,131(10):771-779.
[5]韩超.基于混沌理论的短时交通流预测方法[J].物流工程与管理,2012,34(4):116-117.
[6]Yu H.Network traffic prediction based on wavelet analysis and hopfield neural network[J].Comput Appl Softw.2013,30(6):246-249.
[7]Li HY,Li D.Study on network traffic forecasting based on improved neural network[J].Comput Simul,2011,28(9):132-135.
[8]唐新来,李春贵,王萌,等.基于混沌粒子群算法的神经网络短时交通流预测[J].计算机测量与控制,2010,18(8):1893-1895.
[9]Sang A,Li S Q.A predictability analysis of network traffic[J].Comput Netw,2000,1(1):329-345.
[10]Hui M,Bai L,Li Y B,et al.Highway traffic flow nonlinear character analysis and prediction[J].Math.Prob.Eng,2015,25(8):1-7.