In order to better capture the uncertainty of random fluctuations in traffic flow and improve the accuracy of short-term traffic flow prediction,an adaptive grey interval model is proposed in this paper for uncertain prediction of short-term traffic flow.The adaptive grey interval model consists of an adaptive grey model,a particle swarm optimization algorithm and a residual model.First,an adaptive grey model is con-structed to predict the mean of short-term traffic flow.The particle swarm optimization algorithm is used to obtain the optimal parameters of the adaptive grey model in real time.Then,the residual sequence is ob-tained by comparing the mean prediction result with the true value,the obtained residual sequence is pro-cessed by absolute value,and the residual model is used to process the residual sequence processed by abso-lute value.Finally,the mean prediction result is combined with the residual result to generate a prediction interval to realize the uncertainty quantification of short-term traffic flow.The performance of the proposed model is evaluated using traffic flow data collected from highways in Minnesota,USA.The prediction inter-val coverage probability,prediction interval width and comprehensive index are selected as uncertainty pre-diction performance evaluation indicators,and compared with the grey envelope model(GEPM),grey inter-val prediction model(GIPM)and linear grey interval model(LGIM).The results show that the proposed model can generate feasible traffic flow prediction intervals.By comparing the uncertain prediction perform-ance evaluation indicators,it is shown that the proposed model has better prediction accuracy and can pro-vide decision support for intelligent transportation systems.
维普
万方数据