摘要
利用人工神经网络的并行处理自适应学习能力可实现车流量预测,但网络权值的训练因搜索方法单一容易陷入局部极小值,不易得到全局最优解。将布谷鸟算法用于网络权值矩阵的训练,借助其搜索速度快、不易陷入局部最优解的优势,提出改进的网络学习方案。基于MATLAB软件对所得网络进行训练并将结果用于短时交通流实验,结果表明所得模型的预测结果误差小且具有收敛速度快、容易得到全局最优解的能力。
The parallel processing adaptive learning ability of artificial neural network can be used to predict the traffic flow.However,the training of network weights is easy to fall into local minimum because of the single search method,and it is difficult to get the global optimal solution.The algorithm of Cuckoo is applied to the training of network weight matrix.With the advantage of fast searching speed and being not easy to fall into local optimum,an improved network learning scheme is proposed.MATLAB software is used to train the network and apply the results to the experimental study of short-term traffic flow.The results show that the prediction error of the model is small,and it has the ability of fast convergence and is easy to get the global optimal solution.
作者
薛鹏
任鹏飞
刘守兵
XUE Peng;REN Pengfei;LIU Shoubing(College of Electrical Information Engineering,Henan University of Engineering,Zhengzhou 451191,China)
出处
《河南工程学院学报(自然科学版)》
2020年第4期68-71,共4页
Journal of Henan University of Engineering:Natural Science Edition
基金
河南省高等学校重点科研项目(18B413001)。
关键词
布谷鸟算法
神经网络
交通流量
预测
Cuckoo algorithm
neural network
traffic flow
prediction