摘要
针对目前短时交通流的预测精度不够高这一问题,提出一种布谷鸟算法优化小波神经网络(CSWNN)的短时交通流预测模型。首先采用小波变换对数据进行降噪,并进行归一化处理,然后使用复自相关法对具有混沌特性的短时交通流进行相空间重构,将交通流数据拆分为训练数据组和测试数据组,使用布谷鸟算法优化小波神经网络的各项参数,并根据训练数据组来训练优化后的小波神经网络模型。最后使用测试数据组的数据对CS-WNN模型进行有效性验证。仿真结果表明,相比几种主流的优化预测模型,CS-WNN短时交通流预测模型具有更高的预测精度。
Aiming at the improvement of the prediction accuracy of current short-term traffic flow, a prediction model for short-term traffic flow based on cuckoo search algorithm-optimised wavelet neural network (CS-WNN) was presented. Firstly, wavelet transformation and normalisation were used for data noise reduction, and the phase space reconstruction of short-term traffic flow with chaotic characteristics was done to form training data set and test data set by using complex self-correlation algorithm. Then, the wavelet neural network whose parameters were first optimised by cuckoo search algorithm was trained with training data set. At last, test data set was used for validating the effectiveness of CS-WNN model. Simulation results show that compared with several mainstream optimised prediction models, the proposed CS- WNN model for short-term traffic flow prediction has higher prediction accuracy.
作者
黄晓慧
张翠芳
Huang Xiaohui Zhang Cuifang(College of Information Science and Technology,Southwest Jinotong University, Chengdu 611756 ,Siehuan, China)
出处
《计算机应用与软件》
2017年第3期238-242,共5页
Computer Applications and Software
关键词
短时交通流
复自相关
布谷鸟算法
小波神经网络
Short-term traffic flow Complex self-correlation Cuckoo search algorithm Wavelet neural network