基于粒子群的模糊神经网络的短时交通流量组合预测

为了进一步提高短时交通流量预测的精度,提出了一种粒子群算法的模糊神经网络组合预测模型,模糊神经网络融合了神经网络的学习机制和模糊系统的语言推理能力等优点,弥补各自不足,将自回归求和滑动平均(ARIMA)和灰色Verhulst模型进行初步预测,并将两种初步预测的结果作为模糊神经网络的输入,构建基于改进模神经网络的组合预测模型,在此基础上进行训练和预测,其中模糊神经网络的相关参数由改进粒子群来优化,利用本方法来对南京市汉中路短时交通流量进行预测,结论表明:方法充分发挥了单一模型的优势,比单一的预测模型更加精确,是短时交通流量预测的一个有效方法。

Combined Forecasting Model of Short-Term Traffic Flow Based on Modied Particle Swarm Optimizer and Fuzzy Neural Network

To improve the accuracy of the short-term traffic flow prediction,this paper puts forward a fuzzy neural network combination prediction model for the particle swarm optimization.The fuzzy neural network integrates the advantages of the learning system of the neural network and the language reasoning ability of the fuzzy system,and makes up for its own disadvantages.Besides,it carries out preliminary predication on the Autoregressive Integrated Moving Average(ARIMA) and grey Verhulst model,regards these two results of the preliminary prediction as the input of the fuzzy neural network,and establishes combination prediction model based on the improvement of the fuzzy neural network.On that basis,it proceeds with the training and prediction.The related parameters of the fuzzy neutral network are optimized through improving the particle swarm.This approach is adopted for predicting the short-term traffic flow of Hanzhong Road in Nanjing.It turns out that,this approach gives full play to the advantages of the single model;it is more accurate than the single prediction model;it is an effective way to predict the short-term traffic flow.