A multivariate partial grey prediction model based on second-order traffic flow kinematics equation and its application

Abstract

The Intelligent Transport System (ITS) has been proven to be an effective way to solve urban traffic congestion and improve road capacity, the traffic guidance system is an important part of ITS, and short-time traffic flow prediction is the key issue for the traffic guidance system. In this paper, a second-order multivariate partial grey prediction model based on traffic flow kinematics equation is constructed from the traffic flow kinematics equations to study the spatio-temporal and partial grey prediction model mechanism of complex road networks. The structure of this new model has good interpretability and can capture some nonlinear features of the data, which can portray the dynamic evolution law of traffic flow in two-dimensional road networks. Meanwhile, the least squares technique is used to estimate the parameters of this model, and the model is solved by the Runge-Kutta formula, which solves the problem of solving the multivariate nonlinear system of equations and ensures the high efficiency and accuracy of the model computation. The spatiotemporal and cyclical nature of traffic flow data was considered, and traffic flow data from multiple road sections were selected by the grey correlation analysis method. Finally, the traffic flow data at the same time of different road sections and the traffic flow data at different times of the same road sections are selected to analyze the effectiveness of the new model using four cases, and it is illustrated through the experimental results that the new model has a higher fitting accuracy, which is better than the other five grey prediction models. At the same time, the new model effectively predicts the traffic flow of the two road sections in different periods, and can accurately insight into the trend of traffic flow, the results can provide real-time and accurate traffic flow data for the traffic guidance system, and can also improve the overall operational efficiency of the urban transport system.